EP1558757A1 - Gas jet process and apparatus for high-speed amplification of dna - Google Patents

Abstract

Processes and apparatus for performing fast, precise, and reliable gas jet thermocycling on a sample using of pressurized gases delivered to a thermostated reaction chamber at high velocity using computer controlled electronic multiport valves are disclosed. Apparatus for achieving these processes are also disclosed.

Description

TITLE: GAS JET PROCESS AND APPARATUS FOR HIGH-SPEED AMPLIFICATION OF DNA

TECHNICAL FIELD OF THE INVENTION The present invention relates to methods and apparatus for perfonning high-speed amplification of a sample. More specifically, the present invention relates to methods and apparatus for performing pressurized jet polymerase chain reaction, wherein each cycle can be completed in as little as a few seconds.

BACKGROUND OF THE INVENTION

The Polymerase Chain Reaction. The Polymerase Chain Reaction (PCR) is one of the most widely used techniques in molecular biology (U.S. Pat. No 4,683,202 to Mullis; Saiki et al, 1985; Erlich, 1989; Mullis et al, 1994). PCR-amplified DNA can be used to diagnose mutations responsible for human genetic diseases (Saiki et al, 1985; Kogan et al, 1987), in blood and tissue typing (Saiki et al, 1989a), or to detect pathogens responsible for important infectious diseases (Persing et al, 1993; Nicoll et al, 2001).

In a typical PCR reaction, template DNA sequences lying between the ends of two defined oligonucleotide primers can be amplified in 1 to 2 hours (FIG. 4a). Three sequential steps are normally employed: (1) double-stranded DNA is denatured (D) to a single-stranded form at a high temperature (90°C to 95°C), (2) the resulting single-stranded DNA strands are annealed (A) to oligonucleotide primers at ~45°C to 65°C, and (3) primer: template complexes are elongated (E) using a thermostable DNA Polymerase such as Thermus aquaticus (Taq) Polymerase at ~72°C (Saiki, 1989b).

One cycle of these three steps (denaturation/annealing/elongation) results in a two- fold amplification of a DNA fragment whose 5 ' and 3 ' ends are defined by sequence- specific annealing of the oligonucleotide primers to the DNA template. Therefore, thirty perfectly efficient PCR cycles result in a 2³⁰-fold (~10⁹-fold) amplification of a particular DNA sequence. DNA is thus amplified from amounts of less than one picogram to microgram quantities, which can be detected by standard analytical methods, such as gel electrophoresis, DNA hybridization, or fluorescence-based optical detection.

Automated PCR Instruments. A variety of machines have been built which automate the three-step PCR amplification process (Oste, 1989; Oste, 1994; Newton, 1995; Johnson, 1998). Generally, these devices maybe classified into two categories: robotic devices which move the DNA samples to the heat; and thermocyclers which bring the heat to the samples.

Robotic devices such as Stratagene's ROBOCYCLER move tubes containing PCR reaction samples to and from a series of heat baths, which are thermostated at different temperatures. Although these devices maybe useful in certain research applications, they are incapable of high-speed PCR. They require >60 minutes for 30 cycles of amplification; equivalent to >2 minutes per PCR cycle.

Since the late 1980s, thermocyclers have become familiar devices in many biochemistry laboratories. Most commercially available PCR devices (Perkin-Ehner, MJ Research, Ericomp, Techne, Eppendorf, BioRad, Hybaid) are thermocyclers (Johnson, 1998). In general, two types of thennocyclers are employed: programmable heat blocks and forced hot-air thermocyclers.

Programmable Heat Blocks. Most thermocyclers are heat blocks with holes in them where plastic reaction tubes are heated and cooled under electronic control. Several such devices have been reviewed by Johnson (1998). The problem with programmable heat block designs is that most of the time is spent waiting for a bulky piece of metal to heat up or cool down. For example, in the MJ Research PTC-150 thermocycler (Watertown, MA), 14 seconds/cycle is lost in transition between D, A, and E temperatures (~94°C, ~55°C, and 72°C). A single heating/cooling cycle from 94°C to 55°C wastes 28 seconds per cycle; or -14 minutes during 30 PCR cycles.

Many commonly employed PCR protocols spend one minute at 94°C (denaturation), one minute at ~55°C (annealing), and one minute at 72°C (elongation). For example, in the original PCR method used by Cetus workers (Saiki et al, 1989b), a 536 b.p. human ^β-globin DNA fragment was amplified using 30 cycles of (1 min at 94°C, 1 min at 55°C, 1 min at 72°C). The active duty time for this thermocycling protocol is only ~3.5 minutes = 210 seconds. This time is the amount needed to enzymatically copy a 536 b.p. template 30 times at a Taq DNA Polymerase elongation rate of ~80 nt/sec (Innis et al, 1988; Gelfand and White, 1990). Commercially available heat block thermocyclers (Perkin-Ehner, Ericomp, MJ

Research, Eppendorf, Techne, BioRad, Snark Technologies) require 20 to 25 seconds in order to cool from 94°C to 55°C and another 14 to 20 seconds to heat from 55°C to 94°C (Johnson, 1998). Therefore, the "dead time" for each PCR cycle is another 40 + 5 seconds per cycle. As shown in FIG. 4a, commonly employed thennocycling protocols require (220 seconds/cycle x 30 cycles) = 6600 seconds = 110 minutes (Saiki et al, 1989b). Only ~3.5 minutes of this 110 minutes is productively focused on the PCR process.

Sobczak et al. (1995) have shown that faster PCR protocols can be employed in heat block thermocyclers when denaturation, annealing, and elongation steps are reduced to one to two seconds per cycle. For example, a 139 base pair long human p53 gene fragment was amplified through 35 cycles of (1 sec at 94°C, 1 sec at 58°C, 1 sec at 72°C). However, the heating/cooling time of the thermocycler was still >25 seconds/cycle. Therefore, the overall thermocycling time was ~30 seconds/cycle or -15 minutes for 30 PCR cycles.

Forced Hot- Air Thermocyclers. In order to overcome the long transitional dead times of heat blocks, forced hot-air thermocyclers have been constructed which allow 30 cycles of PCR amplification to be carried out in as little as -10 to 30 minutes. Wittwer and his colleagues have carried out considerable engineering groundwork to optimize rapid DNA amplification in hot-air PCR thermocyclers (Wittwer et al, 1989; Wittwer and Garling, 1991; Wittwer et al, 1994). The rate-limiting step in the three-step PCR reaction sequence (denaturation/annealing/elongation) is the rate of DNA Polymerase elongation. At an elongation rate of 80 nucleotides/sec by Taq Polymerase (Innis et al, 1988; Gelfand and White, 1990), theoretically one second per cycle is needed to amplify DNA fragments shorter than 100 b.p. using ~20-mer primers. For example, only about five seconds per cycle were needed to copy a 536 b.p. β-globin amplicon through 30 PCR cycles (Idaho Technology, 1995; cf. FIG. 4b). h commercial hot-air thermocyclers, first built by Idaho Technology (Idaho Falls, Idaho, USA), the reaction time needed for one PCR cycle of denaturation/annealing/ elongation was substantially reduced because: (1) the device had very low thermal mass; (2) gaseous phase heat transfer from hot air to the aqueous biochemical reaction sample was carried out in thin-walled capillary tubes; and (3) the denaturation and annealing times during the PCR cycle were minimized. For example, using a PCR protocol of 30 cycles of [0 sec 94°C (denaturation), 0 sec

55°C (annealing), 5 sec 72°C (elongation)], a 536 b.p. human β-globin DNA fragment was amplified in 9.9 minutes; or 19.8 seconds/cycle (Idaho Technology, 1995). This forced, hot-air thermocycling protocol was therefore -220/19.8 = 11 times faster than the original method of Saiki et al. (1989b) for PCR amplification of a 536 b.p. β-globin DNA fragment using a conventional heat block thermocycler. hi routine laboratory diagnostic PCR, the hot-air thermocycler is not quite so fast as described above. For example, Nicoll et al. (2001) amplified short (96 to 220 b.p. long) herpes virus DNA fragments through 50 PCR cycles of [0 sec 94°C (denaturation), 3 sec 58° to 67°C (annealing), 6 sec 72°C (elongation)] in -25 minutes; or about 30 seconds/cycle using a Roche LightCycler™. Hot-air Thermocyclers Rely upon Suboptimal Convective Heat Transfer. Although the design of forced hot-air thermocyclers (U.S. Pat. No. 5,455,175 to Wittwer et al; Wittwer et al, 1990; 1994; U.S. Pat. No. 5,576,218 to Zurek et al; U.S Pat. No. 6,200,781 to Tal et al.) is an improvement over block heaters, no attention has been paid to the important role of fluid mechanics in the heat transfer process. The inventors proposed "blowers" (Zurek et al. U.S. Pat. No. 5,576,218) and ambient air (U.S. Pat. No. 5,455,175 to Wittwer et al; Wittwer et al, 1990; 1994; U.S. Pat. No. 5,576,218 ) as the medium for convection. Blowers and fans move atmospheric air (compared to compressed gas sources used in the present invention) and hence the gas velocities are much lower than in compressed gas flows. Furthermore, previous inventions did not consider the use of chilled or cooled air or other chilled gases for the cooling part of the PCR cycle, relying solely on ambient air.

Since prior inventions did not use compressed gas flows, the velocities employed have not been very high (<3 m/s) and consequently they achieved low convective heat transfer rates, compared to what is attainable with high velocity turbulent gas flow. Turbulent flow at high velocity involves compressible, non-isothermal fluid dynamics. Forced-air heating and ambient air cooling at atmospheric pressure (P = 1 bar) are unique and suboptimal solutions to the fluid dynamics of biochemical reaction chambers. Neither Wittwer et al. (U.S. Pat. No. 5,455,175; Wittwer et al, 1990; 1994), Zurek et al. (U.S. Pat. No. 5,576,218), nor Tal et al. (U.S. Pat. No. 6,200,781) have used gases other than air, pressures other than atmospheric, cooling gases colder than 22°C, or time constants faster than one second, they have not employed flow conditioners to mix hot and cold gases and/or to increase gas pressure and velocity. They have not considered the possibility that different gases may be employed at different stages of the PCR process, and have not considered the role of fluid mechanics, the physics of heat transfer, or flow conditions for optimal heat transfer. For example, that "Air is an ideal heat transfer medium which can change temperature quickly because of its low density." (Wittwer et al. (1990)) is not only factually inconect but ignores the complexity of the problem: convective heat transfer in the gas phase depends upon a variety of factors (Welty et al, 1976; Chapman, 1984). The gaseous heat transfer conditions needed for high-speed amplification of DNA using hot/cold pressurized gas bursts is more accurately described as turbulent, non-isothermal compressible gas flow; and virtually none of the optimal gas parameters or flow conditions have previously been optimized. As outlined schematically in FIG. 2a and 2b, a pressurized gas jet thermocycler operates at gas pressures, velocities, turbulent non- isothermal flow conditions, and geometries which have not previously been considered — let alone implemented in high-speed DNA amplification experiments, as shown in FIG. 5 through 12.

In further reference to Wittwer et al. (1990), the heat transfer across a stagnant layer of air (for example, the boundary layer) is very poor. Such inefficient heat transfer explains why double-pane windows have such excellent insulation properties. The only reason why a forced hot-air thermocycler performs better than a block heater is that the flow of air makes the insulating boundary layer sunounding the DNA sample thin. Therefore, despite the low thennal conductivity coefficient for air, heat flux across the boundary layer is faster than through the block heater.

Comparison of Heat Transfer in Hot-air and Pressurized Gas Thermocyclers. The differences between the hot-air thermocycler and the pressurized gas thermocycler are best described by a brief analysis of heat transfer around a thin cylinder (capillary). The cylinder is chosen because both of these thermocyclers use capillaries for holding the DNA samples; but the analysis would be analogous for any other shape of sample holder.

Let d₀ and d_i denote the outer and inner diameters of the capillary tube. Natural convection heat transfer and internal thermal gradients in the reacting liquid mixture are negligible due to the dominant role of viscous forces and small inner diameter. The thermal energy balance of the aqueous solution inside the capillary is then given by

where p_m and Cp_m denote the mixture's density and specific heat capacity. This equation describes how fast the temperature of the sample in the capillary (denoted by T_s) changes when a gas at temperature T_g flows over the capillary. For example, suppose the sample is at 94°C and it must be cooled down to 56°C by flowing cold gas at 0°C over the capillary. Setting the time t = 0 when cold gas starts to flow, the time needed for the aqueous sample inside the capillary to cool down to can be obtained by solving equation (1):

Solving equation (2a) for time t, we obtain:

_{t =} _ P_mCp,Λ _ln 56 0M3p_uCp_Md_t

(2b) h eff 94

In a similar manner, any heating or cooling of biochemical reaction samples can be described. It is evident from equation (2b) that the time t is reduced when the effective heat transfer coefficient h_ef is maximized. This coefficient links the performance of the cycler to the heat transfer and the fluid mechanics. The coefficient h_eff includes the heat transfer through the capillary wall ( h_c ) and through the boundary layer between the gas and outer surface of the capillary,

The heat transfer coefficient for the capillary depends upon the thermal conductivity

2k coefficient of the capillary material ( k_c ) and its dimensions: h = - — . The heat

(d_o - d;) transfer coefficient through the boundary layer ( h_bl) is given by the following conelation, proposed by (1) Hsu and (2) Douglas and Churchill (both are referenced in Welty et al. (1976)). This coefficient is strongly influenced by the fluid mechanics. But first, define p _asd₀v two dimensionless numbers. The Reynolds number Re = ^as , and p_gas denote the mean gas velocity, dynamic viscosity of the gas and gas density. The h d Nusselt number Nu is defined as Nu - ^bl ° , where k_gas is the gas thermal conductivity ^kgas coefficient. The Nusselt number is defined for two flow regimes, laminar and turbulent, as follows.

Laminar flow:

For Re<500, Nu = 0.43 + 0.48 Re⁰⁵ (Hsu conelation).

Turbulent flow:

For Re>500, Nu = — ^— = 0.46Re⁰⁵ + 0.00128Re (Douglas and Churchill d₀k_gas conelation).

When the gas velocity is small (small Re) the Nusselt number Nu is small and h_bl will also be small. Consequently, if h_bl is small compared to h_c , then the effective heat transfer coefficient h_eff » h_bl . If the gas velocity increases to the point when h_bl » h_c , then the effective heat transfer coefficient approaches the limiting value h_eff * h_c .

The following typical conditions exist for the forced hot-air thermocycler: P = 1 bar, v = 2 m/sec and the density, dynamic viscosity and thermal conductivity are p = 0.98kg lm μ = 2.12 x l0^~5 kg f m - sec, k_gas = 0.03W ImK , respectively. Let the outer capillary diameter be d₀ = 0,001m . The Reynold's number is

0.98 x 0.001x 2

^Re = ^'„ _{n n}-_s — = ⁹²-⁵ • ^{Since Re} < ⁵⁰° » ^the conelation of Hsu (1963) is used: 2.12 x 10

Nu = 0.43 + 0.48 x Re^0,5 = 5 and the rate-limiting heat transfer coefficient is

The heat transfer coefficient through a glass capillary with thermal conductivity k_c = IΛW/mK and nm outer diameter and 0.8mm inner diameter is:

h = ^c = ^^ = 11000W /m²K (4b) ^c (d₀ - d_t) (0.001 - 0.0008)

Thus the effective heat transfer coefficient is h h. 11000 x 151 , ._{0 ft ™} , _{2 τ}, h _ff = — — = = l48.96W/m²K. (Ac) ^eff (h_c + h_bl) (11000 + 151)

This result can be compared with the pressurized air thermocycler using the following values: P = 3 bar, v = 20 m/sec. The pressurized gas density, dynamic viscosity and thermal conductivity are, respectively, p = 2.95kg I m μ = 2Λ2xlO^~5 kg /m - s∞,k_gas = O.OWImK . The outer capillary diameter is the same: d₀ = O.OO rø . The Reynold's number is then calculated to be

2.9S« O.001x «⁾

2.12xl0^~5

Since Re > 500 , the conelation of Douglas and Churchill is used: Nu = 0.00128Re+ 0.46 x Re⁰⁵ = 27.92 and the rate-limiting heat transfer coefficient is k_mNu h_bl = - ^. = 838 f lm²K . (5a)

The heat transfer through the capillary is the same and the effective heat transfer is:

By comparing Equation (4c) (hot-air thermocycler) and Equation (5b) (pressurized air thermocycler), it can be seen that the rate-limiting heat transfer coefficient through the boundary layer is more than five times larger for the pressurized air thermocycler. The cool-down time of the comparative study (Equation (2)) will be five times shorter for the pressurized air thermocycler compared to the hot-air thermocycler. Furthermore, this ratio will also be true for heating or cooling during any other step of the PCR process.

Changing the operating gas in the pressurized gas thermocycler to helium, the properties are p = O.Akg/m³,μ = 2.31 x l0^~5kg/m - sec, k_gas = 0Λ7W/mK . The Reynold's number is now _{Re =} 0.4x0.0Q₁x20 _{= 693 mi}

2.131 x l0^"5

Nu = 0.00128 x 693 + 0.46 x 693⁰'⁵ = 13.0. Using pressurized helium gas, the rate-limiting heat transfer coefficient is therefore k Nu h_bl = -£ϊ = 22\0W/m²K . (6a) d₀ The effective heat transfer coefficient is now: h h_bl 1 1000 x 2210 _{1 0} ,_ΛTT, , ₂ ~ h _ff = ^C-Z — = = l84QW/m²K. (6b)

( + h_u) (1 1000 + 2210)

By comparing Equation (4c) to Equations (5b) and (6b), there is an expected fivefold improvement in the pressurized air jet as compared to the forced hot-air thermocycler and over twelve-fold improvement using pressurized helium as an operating gas. This example demonstrates that superior heat transfer translates into faster PCR. Improved heat transfer is accomplished by using pressurized gases; preferably helium; however, even suboptimal gases like air perform far better than non-pressurized devices.

Advantages of a Pressurized Gas Jet Thermocycler. Based upon the above theoretical analysis, one expects that it will be possible to deliver hot/cold pressurized gas bursts at velocities in the range 3m/sec<v <30 m/sec (Mach 0.03 < v < Mach 0.30 to a biochemical reaction chamber, in order to rapidly change the temperature of enzymatic reactions from -50° C to -90° C in thin-walled vessels. However, the requisite geometry, gas dynamics, flow conditions, method for hot/cold gas mixing, and timing of the gas bursts are not trivial engineering problems. To those of ordinary skill in the art of fluid mechanics, solutions to such problems are solved using Navier-Stokes Equations and Computational Fluid Dynamic (CFD) modeling (Anderson, 1995).

The pressurized gas thermocycler differs from the hot-air thermocycler in six aspects:

(1) Higher-than-atmospheric pressure is used (increased gas density): P > 1.3 bar.

(2) Velocities are much higher (decreased boundary layer thickness): v > 5 m/sec. (3) Gases other than air can be used; for example, helium can be used for rapid heating and CO₂ can be used for superior thermal control and efficient cooling.

(4) Reactions to which one or more gas flow conditioners are attached, are employed for gas mixing.

(5) Digitally programmable electronic valves are used to deliver hot and cold gas bursts to the reaction chamber via conditioning chambers to regulate and mix the gas flows.

(6) DNA denaturation, primer.template annealing, and enzymatic elongation are programmable in time intervals of <100 milliseconds.

It is a well-known fact that convective heating/cooling of objects are far more efficient than thermal conduction. For example, wind chill temperatures indicate the additional cooling effect of wind motion, i.e., "convection". Efficiency in this sense means the time it takes an object to change it's temperature by a certain amount. Convective cooling/ heating rates depend on: (1) the temperature difference between the gas and the object; and (2) the thickness of the boundary layer around the object (the object is sunounded by a thin layer of gas that is stationary with respect to the object, termed the "boundary layer"). The thickness of this layer strongly conelates with the velocity of the gas relative to the object. At higher gas velocities, the boundary layer thins and heat transfer rate (termed "heat flux") improves. High velocities can be achieved by using pressurized gas: molecules will naturally flow from a region of high pressure to lower (atmospheric) pressure, as regulated by opening/closing of electronic and/or mechanical control valves. The transfer of thermal energy (heat) across the boundary layer is also strongly influenced by the type of gas that is used. Helium gas is particularly efficient for rapid heating of objects. Carbon dioxide (CO₂) gas cools to -24°C upon isenthalpic expansion and it is therefore an efficient, inexpensive cooling gas. This property depends on the polytropic coefficient of the gas - that is a thermodynamic property intrinsic and unique to each gas or gas mixture. When gas velocities approach values of Mach 0.3 (0.3 times the sound velocity of the gas) the gas becomes compressible. This compressibility implies a stronger coupling between gas temperatures (thermal internal energy) and kinetic energy (gas velocities) and consequently, turbulence develops. The rapid heat transfer needed for high-speed amplification of DNA is more accurately described as turbulent, non-isothermal compressible gas flow. Generally speaking, such reaction conditions are rarely used in biochemistry. It is therefore not obvious to those of ordinary skill in the art of biochemistry how to employ such complex gas flow conditions in order to carry out temperature- sensitive biochemical reactions.

Specialists in heat transfer always contrast the macroscopic nature of convective heat transfer with the microscopic nature of thermal conduction. Since heat fluxes in block heaters are of thermal conduction type (microscopic), the rate is described by Fourier's law:

Q_cond ~ ~k , where q_cond is the heat flux by thermal conduction in the x- direction, k is dx the thermal conductivity coefficient (a property of the material from which block heater is

made) and — is the temperature gradient in the x — direction. The rate is affected by (1) dx the distance over which the heat is conducted (of the order of centimeters), (2) temperature differences limited by material properties, (3) and the thermal conductivity coefficient. The basic laws governing heat transfer in the gas phase physical principles have been known since Fourier (1822) and Navier (1856), are taught in standard mechanical engineering textbooks (Bosworth, 1952; Azbel, 1984; Chapman, 1984). However, the fluid dynamics of turbulent, non-isothermal, compressible gas flow, as used in the novel process and device herein described, is a specialized subject not commonly known to those practiced in the art of biochemistry.

The thermal conductivity coefficient of helium was measured by Ubbink (1947) and is listed in tables published by Johnston and Grilly (1946), Bosworth (1952), Azbel (1984), and Chapman (1984). As shown in FIG. 1, helium has a thermal conductivity coefficient seven times that of air. Neon transfers heat about twice as fast as air. Therefore, hot-air thermocyclers have used the wrong gas for rapid heat transfer. Among the non- combustible gases, air is available; helium is optimal; in particular, the isotope of helium, ³He. U.S. Pat. No. 5,455,175 to Wittwer et al. assumes that the same gas (air) should be used for heating the reaction chamber, cooling the chamber, or holding its temperature at a fixed value. In fact, as shown in the temperature versus time profiles in FIG. 8b and 9b, gases which have a high thermal conductivity (high k gases), such as helium, are superior for heating/cooling the reaction chamber, whereas low k gases, such as air or CO₂, afford better thermal control when holding the temperature for several seconds during enzymatic elongation at ~72°C (see FIG. 5b, 9b, 10b). hi other words, different gases are optimal for different stages of the PCR process.

Secondly, U.S. Pat. No. 5,455,175 to Wittwer et al, U.S. Pat. No. 5,576,218 to Zurek et al, and U.S. Pat. No. 6,200,781 to Tal et al. specify the use of air (but no other gas) at atmospheric pressure (but no other pressure) for thermocycling. Not only is air a relatively poor gaseous heat transfer medium, but it does not need to be used at atmospheric pressure. PCR amplification of DNA using air at atmospheric pressure (P = 1 bar) is convenient. However, the process is demonstrably faster at elevated pressure (P > 1.3 bar) and velocity (v > 5 m s)(see FIGS. 4, 5b, 7b, 8b, 9b, and 12b).

In the forced hot-air thermocycling process of Zurek et al. described in U.S. Pat. No. 5,576,218, "the sample could be heated from 50°C to 85°C within 12 to 15 seconds . . . by injecting cooling air at a substantially lower temperature than the target temperature, for example 22° air, the sample was cooled from 85°C to 50°C in approximately 60 to 75 seconds." Therefore, the method of Zurek et al. described in U.S. Pat. No. 5,576,218 requires at least (12 + 60) = 72 seconds per cycle just to heat and cool the gas (air), regardless of whether any useful biochemistry has taken place. Zurek et al. provide no evidence that their process or device can actually be used to amplify DNA. Furthermore, Zurek et al. does not contemplate the possibility of employing gas pressures greater than atmospheric (P > 1 bar) because the gas pressure in their machine was never measured. In the present invention using pressurized gases at -2.5 bar, 30 cycles of PCR amplification are achieved in as little as 78 seconds — about the same time needed for one cycle using the method of Zurek et al. (U.S. Pat. No. 5,576,218).

Third, it has been assumed by Wittwer et al. (U.S. Pat. No. 5,455,175; Wittwer et al. 1990; 1994) that the heating chamber and reaction chamber are the same thing (see FIG. 2). In fact, much faster convective heat transfer is possible if the reaction chamber is separated from the source of hot gas by one or more electronic valves and a flow conditioner and/or gas accumulator chambers. Fourth, it has been assumed by Wittwer et al. (U.S. Pat. No. 5,455,175), Zurek et al. (U.S. Pat. No. 5,576,218), and Tal et al. (U.S. Pat. No.6,200,781) that air at ambient temperature is an optimal cooling medium. In fact, a biochemical reaction chamber can be cooled much more rapidly using gases other than air at temperatures less than 25°C. The temperature versus time profiles in FIG. 5b, 7b, 8b, and 9b demonstrate clearly that pressurized CO₂ gas (P > 2.3 bar, T = 5°C) is far superior as a gaseous cooling medium, as compared to ambient air (P = 1 bar, T = 23°C).

Theoretically, it may be possible to rapidly cool biochemical samples using pressurized gases, including air, at relatively high pressures (P > 5 bar) supplied by vortex tubes such as those described by Ranque (1934) and Hilsch (1947). However, neither Wittwer et al. (U.S. Pat. No. 5,455,175) nor Zurek et al. (U.S. Pat. No. 5,576,218) have considered the use of either pressurized gases or Ranque-Hilsch vortex tubes for cooling.

Fifth, it has been assumed by Wittwer et al. (U.S. Pat. No. 5,455,175; Wittwer et al. 1990; 1994), Zurek et al. (U.S. Pat. No. 5,576,218), and Tal et al. (U.S. Pat. No. 6,200,781) that the relevant time constants to be programmed for heating/cooling DNA reaction samples are integral values of 0, 1, 2, 3. . . seconds. They do not consider the possibility that much faster PCR is possible when the device is programmed in time intervals of fractions of a second. The present invention allows for much faster denaturation, annealing, and elongation because time parameters are programmable in tenth of a second (FIG. 13).

Altogether, for efficient gaseous heat transfer, forced hot-air thermocyclers utilize the wrong gases for heating and cooling, the wrong gas pressures, and the wrong configuration of heat chamber to reaction chamber. The time constants employed (>1 second) are too slow for high-speed thermocycling (see FIG. 2), forced hot-air thermocyclers gases lack optimal CFD-modeled gas dynamics and utilize no flow conditioners. Much faster thermocyclers can be designed using pressurized gases, particularly helium/air/CO₂ combinations, which are delivered to a thennostated reaction chamber using a flow conditioner (FIG. 3c) and electronically actuated valves. h order to put the speed and performance of the present invention in perspective, Chiou et al (2001) from the Massachusetts Institute of Technology have recently described a "state of the art" thermocycler. They state that "the objective of the present research was to produce a PCR machine to amplify a sample in as little time as possible..." (p. 2018). The device described by Chiou et al. (2001) allowed a 500 base pair long bacteriophage λ DNA fragment to be amplified through 30 PCR cycles in 23 minutes, starting with -1 nanogram (10^"9 g) of DNA; and the efficiency was 78%. For 30 PCR cycles, the amplification factor is therefore 1.58³⁰ = 6.0 x 10⁵ or 600,000-fold.

The present invention allows for high-speed amplification of viral, bacterial, and single-copy human gene fragments on a time scale which is over an order of magnitude faster than the method described by Chiou et al. (2001).

Furthermore, as shown in FIG. 8a, the efficiency of gas jet PCR is much higher, -93%. The present invention gives a DNA amplification factor of 1.86³⁰ = 1.35 x 10⁸-fold or 135 million-fold. Therefore, gas jet PCR is [1.35 x 1076.0 x 10⁵] = 225 times more sensitive than the method of Chiou et al. (2001). The pressurized gas thermocycler routinely amplifies viral and bacterial DNA fragments 85 to 2331 b.p. long through 30 PCR cycles in 1.3 to 10.3 minutes (78 to 617 seconds). Single copy human gene fragments in this size range are amplified through 35 PCR cycles to detectable quantities in 3.5 to 6.5 minutes (FIG. 5a, 9a, and 13).

Therefore, in light of the limitations of the above-described prior art, a novel gas jet amplification process and an apparatus for its automation using pressurized gases and electronic valves is disclosed herein.

BRIEF SUMMARY OF THE INVENTION

The present invention includes a high-speed process for amplifying DNA. A reaction chamber containing a biological sample, a DNA polymerase, oligonucleotide primers, and deoxynucleoti.de precursors is provided. The reaction chamber accepts the flow of one or more heat transfer gases. A first heat transfer gas is heated in a heating chamber which is physically separated from the reaction chamber. The heated gas is delivered to the reaction chamber at a pressure greater than or equal to 1.3 bar (P >20 p.s.i.).

Heat from the heated gas (>95°C) is utilized to denature DNA. A second heat transfer gas, which may be the same kind as, or a different kind of gas than, the first heat transfer gas is delivered to the reaction chamber. The second heat transfer gas (<20°C) cools the reaction chamber to a temperature low enough to allow the denatured DNA to anneal to the oligonucleotide primers. Finally, using the first hot (>95°C) heat transfer gas, the temperature of the reaction chamber is increased to a sufficient temperature (typically 68° to 75°C) to allow for enzymatic elongation of primer:template complexes. When hot (>95°C) and cold (<20°C) gases are delivered to a biochemical reaction chamber under the control of either two-way, three-way electronic valves, or multiport electronic valves, extremely fast thermocycling is possible. By delivering, alternately, bursts of hot and cold gas to a thermostated reaction chamber through a flow conditioner (or a set of flow conditioners), the temperature of an enzyme-catalysed reaction can be controlled on a time scale of tens of milliseconds. Ideally, the gas of choice is helium, but for convenience, other gases can also be employed.

An embodiment of the present invention also includes an apparatus for high-speed amplification of DNA. The apparatus includes a reaction chamber into which compressed hot (>95°C) and cold (<20°C) gas bursts are delivered on a time scale of fractions of a second. The overall configuration of the device consists of six parts: (1) a heater to raise the temperature of the heating gas to a pre-set value (> 95" C ), (2) one or more flow conditioning chambers for mixing cold and hot gases and regulate flow and attenuate pressure pulses, (3) an intake manifold with attached mixing chamber to mix hot/cold gases, (4) a gaseous reaction chamber, (5) an exhaust nozzle and (6) heat exchangers to recover heat from reactor outlet and hot gas by-pass. In certain respects, the pressurized gas thermocycler therefore resembles jet and/or rocket engines. However, instead of mixing fuel and oxidant (as in a jet or rocket), hot and cold non-combustible gases are mixed under the control of electronic valves; and no ignition occurs. The essential process described in the present invention is therefore properly termed "gas jet PCR" and the device is aptly named a "PCR jet". Fast, precise, reliable thermal control of enzyme-catalysed reactions, are the purpose of the PCR jet.

A heating chamber, which is physically separated from the reaction chamber, is fluidly connected to the reaction chamber. A first container, having a heating gas at a pressure greater than the reaction chamber pressure, is fluidly connected to the heating chamber. This connection may include a flow conditioning chamber for flow regulation and suppression of pressure pulses. A second container, having a cooling gas (which may be the same kind or different kind of gas as the heating gas), is fluidly connected to the reaction chamber. This connection may include a flow conditioning chamber for flow regulation and suppression of pressure pulses. One or more cooling gas inlet valves are positioned between the second container and the reaction chamber, and one or more heating gas inlet valves are positioned between the first container and the reaction chamber. A programmable controller, having electrical inputs and outputs, is used to control opening of the inlet valves. Temperature sensors, positioned in the reaction chamber, provide output to the controller. The controller opens and closes the inlet valves to reach or maintain a desired temperature within the reaction chamber. If the temperature of an optimal heat transfer gas, such as helium, is raised to >95°C in a chamber which is physically separated from the reaction chamber, and then delivered to the reaction chamber electronically actuated valves, then extremely fast thermocycling is possible. Even faster heat transfer is possible if cold (<20°C) gas is delivered to the chamber. For automated heating and cooling of the reaction chamber, preferably at least two electronic valves (hot gas valve and cold gas valve) and one or more mechanical relief valves are employed.

In practice, it is neither necessary nor optimal to employ pressurized helium gas for heating and cooling the reaction chamber. Bottled pressurized carbon dioxide (CO₂) gas expands and cools upon leaving its storage container (isenthalpic cooling). It is therefore convenient to use pressurized helium gas (or compressed air) to heat the reaction chamber and bottled CO₂ gas to cool the chamber.

For example, using hot (~170°C) pressurized helium gas and cold (~5°C) pressurized CO₂ gas, which are delivered to a thermostated reaction chamber using 6V or 24V electronic valves and digital and/or analog signals, it is possible to cyclically change the temperature from 92°C (DNA denaturation temperature) to 55°C (primer annealing temperature) to 72°C (DNA polymerase elongation temperature) in less than 4 seconds. Depending upon the length of the DNA fragment to be amplified, 30 cycles of 92°C/55°C/72°C can be carried out in -2.0 to 4.0 minutes when minimal (0 to 2 second) denaturation, annealing, and elongation times are used (FIG. 12). It is therefore an object of the present invention to provide a process for amplifying

DNA so that Polymerase Chain Reaction (PCR) can be carried out rapidly using pre-heated and/or pre-cooled gases which are delivered to a thermostated reaction chamber using electronic valves. In this process, hot (>95°C) or cold (<20°C) pressurized gases are delivered to a physically separate reaction chamber containing biological samples, in order to control the temperatures used for DNA denaturation, primer:template annealing, and polymerase-catalysed elongation.

It is a further object of the invention to provide for a process for subjecting biological samples to rapid thermocycling, by regulating the flow of hot or cold pressurized gases (P > 1.3 bar) to these samples.

It is another object of the present invention to provide an apparatus for carrying out amplification of DNA, using electronic valves to regulate the flow of hot and cold pressurized gas into a thermostated biochemical reaction chamber.

Yet another object of the of the present invention to use one or more chambers prior to the reaction chamber to regulate the hot and cold gas flows in order to mitigate pressure pulsations. Still a further object of the present invention to use a mixing chamber before the reaction chamber in order to achieve homogeneous temperature in the flow admitted into the reaction chamber.

It is also an object of the present invention to provide an apparatus which can subject a biological sample or samples to rapid thermal cycling, using one or more gases at pressures greater than or equal to 1.3 bar.

It is a further object of the present invention to provide an apparatus which can subject a biological sample or samples to rapid thermal cycling, in which pressurized air, helium, carbon dioxide, nitrogen, argon or mixtures thereof are employed as gaseous heat transfer media. The same gas or gas mixture need not be used for heating and cooling. Also an object of the present invention to provide an apparatus which can subject a biological sample or samples to rapid thermal cycling, in which the time parameters for DNA denaturation, primer annealing, and elongation are programmable in fractions of a second.

Finally, it is an object of the present invention to provide an apparatus in which a physically separated gas heating chamber, cooling chamber, and reaction chamber are employed. These and other features, aspects, and advantages of the present invention will become better understood with reference to the following drawings, description, and appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to better appreciate how the above-recited and other advantages and objects of the invention are obtained, a more particular description of the invention briefly described above will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings. Understanding that these drawings depict only a typical embodiment of the invention and are not therefore to be considered limiting of its scope. The invention will be described and explained with additional specificity and detail through the use of the accompanying drawings in which :

FIG. 1 is a table of Thennal Conductivity of Gases, with thermal conductivity units (k-values) expressed in Watts/meter-°C. FIG. 2 a is a schematic diagram which compares a hot-air thermocycler to a pressurized gas thennocycler.

FIG. 2b is a table which compares properties of a hot-air thermocycler to a pressurized gas jet thermocycler.

FIG. 3a is a perspective view of an embodiment of an apparatus to amplify DNA according to the present invention.

FIG. 3b is a perspective view of reaction chamber 38, flow conditioners 30 and 34, and the thermocouples 44, 46, 48, and 50.

FIG. 3c is a view of a flow conditioner used in the reaction chamber 38 of FIG. 3a. FIG. 3d shows a computational fluid dynamic mesh of a flow conditioner as shown in FIG. 3c.

FIG. 3e shows a perspective view of the reaction chamber 38 as shown in FIG. 3a.

FIG. 4 shows temperature versus time profiles of different types of thermocyclers.

Note that the pressurized gas thermocycler is -10X faster than a hot-air thermocycler; and

-100X faster than a conventional heat block thennocycler. FIG. 5a is a picture of a gel electropherogram, which shows high-speed gas jet PCR amplification of four different DNA templates, including (a) a 91 b.p. E.coli O157:H7 Stx2 amplicon, (b) a 333 b.p. bacteriophage λ 'D' gene fragment, (c) a 364 b.p. human Platelet Antigen HPA-4 allele gene fragment, and (d) a human β-globin 536 b.p. gene fragment.

FIG. 5b is a temperature versus time profile for gas jet amplification of the four DNA fragments shown in FIG. 5a. FIG. 6 is a picture of a gel electropherogram, which shows the reaction products of high-speed gas jet PCR amplification of an 85 b.p. E.coli O157:H7 Stx DNA fragment using 2.8 bar pressurized air with 2.9 bar CO₂ cooling.

FIG. 7a is a picture of a gel electropherogram, which shows He/CO₂ gas jet PCR amplification of an 85 b.p. E.coli O157:H7 Stx DNA amplicon through 30 high-speed PCR cycles in 85 seconds (lane a), 81 seconds (lane b), or 78 seconds (lane c).

FIG. 7b is a temperature versus time profile for the experiment shown in FIG. 7a (lane c). This experiment represents the fastest DNA amplification experiment which has been carried out.

FIG. 8a is a picture of a gel electropherogram, which shows the reaction products of high-speed gas jet PCR amplification of a 368 b.p. DNA fragment from Bacillus anthracis.

FIG. 8b is a temperature versus time profile for PCR jet amplification of the 368 b.p. B. anthracis DNA fragment shown in FIG. 8a (lane e). Total time for 30 high-speed PCR cycles was 233 seconds.

FIG. 9a is a picture of a gel electropherogram, wltich shows high-speed gas jet PCR amplification of (a) 191 b.p. human Accute Intermittant Porphyria (AIP) amplicon, (b) 147 b.p. human ABO blood group 'A-transferase' gene fragment, (c) 486 b.p. E.coli uidA gene fragment, (d) 297 b.p. E.coli uidA gene fragment, (e) 145 b.p. E.coli uidA gene fragment.

FIG. 9b is a temperature versus time profile of the PCR jet reaction used to amplify the 147 b.p. human ABO blood group A-transferase gene fragment shown in FIG. 9a. 35 PCR cycles were carried out in 228 seconds.

FIG. 10a is a picture of a gel electropherogram, which shows nested PCR amplification of an outer 486 b.p. and nested inner 186 b.p. E.coli uidA DNA fragment.

FIG. 10b is a diagram showing the DNA sequences and locations of primers used in the nested PCR experiment shown in FIG. 10a. FIG. 11a is a picture of a gel electropherogram, which shows gas jet PCR amplification of a 2206 b.p. long DNA fragment from bacteriophage λ.

FIG. lib is a picture of a gel electropherogram, which shows gas jet PCR amplification of a 2331 b.p. long DNA fragment from bacteriophage λ. FIG. 12a is a picture of gel electropherogram, which shows gas jet PCR amplification of a 297 b.p. long DNA fragment from the E.coli uidA (β-glucuronidase) gene.

FIG. 12b is a temperature versus time profile of the PCR jet reaction used to amplify a 297 b.p. long DNA fragment from the E.coli wt--4(β-glucuronidase) gene. FIG. 13 is a table summarizing high-speed PCR experiments carried out using the pressurized gas jet thermocycler.

FIG. 14 is a list of scientific nomenclature used herein. DETAILED DESCRIPTION OF THE INVENTION

The present invention will be described as it applies to a prefened embodiment. It is not intended that the present invention be limited to the described embodiment. It is intended that the invention cover all modifications and alternatives such as using two-way or three-way valves, which maybe included within the spirit and scope of the invention.

Five features of this pressurized gas thermocycler are important:

(1) It performs high-speed amplification of DNA with 30 cycles of PCR amplification of an 85 base pair (b.p.) DNA fragment carried out in 78 seconds. Longer (145 to 2331 b.p.) viral and bacterial DNA fragments were amplified ~10⁸-fold in approximately 2 to 10 minutes (FIG. 5 through 12).

(2) During amplification of DNA in a pressurized gas j et thermocycler, the time intervals used for DNA denaturation, primeπtemplate annealing, and enzymatic elongation are programmable in fractions of a second.

(3) The overall speed of DNA amplification is limited by the biochemistry rather than the heating/cooling time of the apparatus. In particular, the rate of Taq Polymerase elongation (-80 nucleotides/sec at 72°C) is rate-limiting. If faster (>200 nt/sec), thermostable DNA Polymerases can be found, then thermocycling times of less than 30 seconds for 30 PCR cycles are possible. (4) High-speed pressurized gas PCR is generally more accurate than slower methods, probably because false reaction products have so little time to anneal and/or elongate.

(5) The pressurized gas jet thermocycler is compatible with on-line, fluorescent dye- based DNA detection optics. The pressurized gas j et PCR process and thermocycler should prove especially useful in the diagnosis of life-threatening diseases where speed is essential. The present invention can be (and has been) practiced to carry out rapid (about 2 to 10 minute) DNA- based tests used in biomedical research, genetics, molecular medicine, agriculture, veterinary science, forensics, and the detection of biological warfare agents. A prefened embodiment of the invention is a pressurized gas thermocycler which carries out high-speed PCR amplification of DNA by injecting hot or cold gas into a thennostated reaction chamber under the control of fast (<20 msec/cycle) electronic valves, actuators which accept digital and/or analog signals, and a microprocessor controller. Ideally, this device would contain a minimum number of moving parts and also be compatible with optics used for detection of fluorescent dye-labeled DNA (Higuchi et al, 1992; Haugland, 1996; Spears et al, 1997).

It should be appreciated that in this embodiment the temperature can be rapidly changed by injecting hot and cold pressurized gas into a reaction chamber under electronic control, using electronically programmable "hot gas" and "cold gas" valves. The pressure in the reaction chamber is maintained at a relatively constant value by means of a mechanical relief valve, although the temperature can be changed upon demand. Reaction samples (10 to 25 microliters in volume) are sealed in thin- walled glass capillary tubes as described by Wittwer and Garling (1991), or similar thin- walled plastic vessels, as described by Tret'yakov et al. (1994). Therefore, although the pressure inside the reaction chamber varies over the range 1.3 to 6.0 bar (20 to 90 p.s.i.), as the temperature of the gas changes, the pressure in the capillary tubes remains unchanged near 1.0 bar (14.7 p.s.i.). A pressurized gas thermocycler differs from a hot-air thermocycler (U.S. Pat. No. 5,455,175 to Wittwer et al.) in several important respects, as shown in FIG. 2. In the hot- air thermocycler (FIG 2a), a single reaction chamber with a heat lamp is used to heat air and to carry out PCR reaction (heat chamber = reaction chamber). In the pressurized gas thermocycler (FIG. 2b), up to six separate chambers are used: a heating chamber, a cold gas supply chamber, a hot gas flow conditioner, a cold gas flow conditioner, a mixing chamber, and a reaction chamber. In order to heat samples in the reaction chamber, preheated gas (>95°C) from the heat chamber is delivered under pressure to the reaction chamber via the hot gas flow conditioner and mixing chamber by electronic valve VH (hot gas valve 28). hi order to cool samples in the reaction chamber, cold gas (<20°C) is delivered under pressure to the reaction chamber via the cold gas flow conditioner and the mixing chamber by electronic valve Nc (cold gas valve 36). A mechanical relief valve V (40 in FIG. 3 a) allows exit of pressurized gas from the reaction chamber. Efficiency is improved by recovery of heat from exit streams and electronic valve V_H when it is set to by-pass the hot gas flow conditioner.

Reference will now be made to the drawings wherein like structures will be provided with like reference designations. Schematic drawings of the pressurized gas thermocycler, flow conditioners, and its reaction chamber are shown in FIG. 3a, 3b, 3c, 3d and 3e. Fabrication and Optimization of a Pressurized Gas Thermocycler.

As seen in FIG. 3a, the thermal cycling device 10 includes a reaction chamber made of insulating (low k) material, generally designated at 38. Preferably, reaction chamber 38 has k value of less than 0.5 Watts/centimeter-degree K. Reaction chamber 38 can be comprised of a number of different materials, such as stainless steel or titanium. Polyurethane or polyimide plastics with ceramic binders, sold under trade names such as VESPEL, TORLOΝ, and BUTTERBOARD (which is manufactured by Golden West of California), can also be used as a reaction chamber 38 material. Cold gas supply 14 is separated from reaction chamber 38 by a pressure regulator 16, a one-way check valve 18 (available from Linweld of Lincoln, ΝE), and a two-way electronic valve 34 (a 38 VDC valve manufactured by Peter Paul Electronics of New Britain, CT) which is actuated by a digital relay operated by controller 42. Controller 42 can be a multifunction data acquisition board, such as a KPCI-3107 manufactured by Keithley of Cleveland, OH. Hot gas supply 12 is connected to a preheater 20 through a pressure regulator 16 and a one-way check valve 18. Preheater 20 is a heat exchanger which cools the exhaust gas from reaction chamber 38 and heats the incoming gas from gas supply 12. Preheated gas exiting preheater 22 then enters preheater 26 and then process heater 26 (500 Watt process heater manufactured by Hotwatt of Danvers, MA). Process heater 26 is actuated by a heavy-duty relay operated by controller 42. Hot gas exiting the heater 26 flows to a three-way electronic valve 28 (operated by controller 42). The process heater requires a continuous flow of gas; therefore electronic valve 28 exhausts gas into preheater 22 and finally through muffler 24 (manufactured by Exair of Cincinnati, OH) whenever the reaction chamber 38 is cooling. When reaction chamber 38 is heating, the hot gas flows from electronic valve 28 into flow conditioner 30. Flow conditioner 30 smoothes gas flow pulses into a steadier flow of hot gas. The conditioned hot gas then enters mixing chamber 32. Cold gas from electronic valve 36 also flows through a flow conditioner 34 and into mixing chamber 32. The mixed gas then flows into reaction chamber 38. Reaction chamber 38 is kept at an elevated pressure by mechanical relief valve 40. Mechanical relief valve 40 has a cracking pressure of 1.3 bar (20 p.s.i.). Exhaust gas from mechanical relief valve 40 cools in preheater 20 and exits through muffler 24.

As shown in FIG. 3b, electronic valves 28 and 36 can be opened or closed upon demand, using controller 42, based upon the temperatures in the mixing chamber 32, hot gas flow conditioner 30, and cold gas flow conditioner 34 measured by the thermocouples 44, 46, and 48 respectively. Thermocouple 50 is used by controller 42 to keep process heater 26 at a constant temperature.

As shown in FIG. 3e, the reaction chamber 38 contains an elliptical cavity 52, which contains reaction samples sealed in thin-walled capillary tubes. These tubes enter through staggered, sealed holes 54.

The pressurized gas thermocycler was tested with various combinations of hot and cold supply gases. Pressurized air is available, inexpensive, and easy to control. On the other hand, pressurized helium gas is a superior heat transfer gas (FIG. 1), but is more problematic with respect to thermal control (cf. FIG. 7b).

Operation of the pressurized gas thermocycler requires three operating modes, which are controlled by activating relays wired to the valves. The three operating modes are: 1) heating the reaction chamber, (2) cooling the reaction chamber, and (3) holding the temperature of the reaction chamber. In order to increase the temperature of the reaction chamber 38, the hot gas valve 28 is OPENED to flow conditioner 30, so that hot pressurized gas is delivered to the chamber. The cold gas valve 36 remains CLOSED.

In order to reduce the temperature of the chamber, the hot gas valve 28 is CLOSED to flow conditioner 30 and the cold gas valve 36 is OPEN.

During the elongation step of the PCR process, when the temperature is held near 72°C, the cold gas valve 36 and hot gas valve 28 flutter OPEN and CLOSED on demand based on the temperature measured by the thermal sensor located in the mixing chamber 32. This valve configuration provides superior thermal control for holding the temperature near a preset value.

The opening and closing of hot and cold valves is controlled by electrical signals from the system controller to electrical relays which actuate the valves. The system control software determines the opening and closing of these valves.

When different "hot" or "cold" supply gases are employed, modified software is utilized, since gas flow and mixing in the chamber varies, depending upon the gas mixtures chosen. It is convenient to use either pressurized air or helium as a "hot" gas; and to use air or CO₂ as a "cold" gas; but a variety of different gases can be employed. It is neither necessary nor optimal to use the same gas to heat, cool, or hold the temperature of the chamber. As shown below (FIG. 5a, 6, 7a, 8a, 9a, 1 la, b and 12a) high-speed gas jet amplification of DNA can be carried out using pressurized air (or helium) as a 'hot' gas, and a pressurized 'cold' gas (CO₂ or air). Thirty cycles of PCR amplification can be achieved in 78 to 617 seconds, depending upon the heat transfer gases chosen, the size of the thermal gradient, and the length of the DNA fragments amplified. In the examples described below, amplified DNA was detected using a relatively slow (-1 hour) process: gel electrophoresis followed by Ethidium bromide (EfBr) staining, a standard method of DNA detection known to those of ordinarily skill in the art.

However, pressurized gas jet PCR is compatible with on-line fluorescence-based detection (Higuchi et al, 1992) using dyes such as SYBR Green, which bind selectively to double-stranded DNA (Haugland, 1996). hi principle, the double-stranded DNA reaction products of high-speed gas phase PCR can be rapidly detected on-line using reporter dyes like SYBR Green and inexpensive photodiode detectors. For example, a pressurized gas thermocycler could be fitted with photodiode detectors and either a light-emitting diode (LED) or laser diode dye excitation source. Such optics, or equally inexpensive fluorescence polarization detection optics (Spears et al, 1997) and fluorescein-labeled primers would allow amplification/detection of DNA on an unprecedented time scale.

EXAMPLES

The following examples serve to better illustrate the invention described herein and are not intended to limit the invention in any way. Those skilled in the art will recognize that there are several different parameters which may be altered using routine experimentation and are intended to be within the scope of this invention.

High-Speed Amplification of DNA Using a Tri- valve Pressurized Gas Thermocycler.

Initial high-speed PCR experiments in a tri- valve gas jet thermocycler were carried out with compressed air as a "hot" gas (~170°C) and CO₂ as a "cold" gas (~5°C). This configuration was not the fastest possible, but was convenient and did not require active refrigeration of the "cold" gas.

EXAMPLE 1 30 PCR Cycles in 5:35 (335 seconds)

Gas jet DNA amplification experiments were carried out using pressurized air (2.4 bar = 36 p.s.i.) as a "hot" gas and pressurized CO₂ (2.6 bar = 40 p.s.i.) as a "cold" gas in the tri-valve machine (FIG. 3a). Four model DNA templates of different lengths were PCR- amplified in separate 10 μl reactions, which were carried out in thin- walled glass capillary tubes: (a) a 91 b.p. E.coli O157:H7 Stx amplicon, (b) a 333 b.p. λ 'D' gene amplicon, (c) a 364 b.p Human Platelet Antigen HPA-4 amplicon, and (d) a human β-globin 536 b.p. amplicon were amplified using 30 cycles of [0 sec 92°C/0 sec 55°C/5 sec 72°C] with pressurized air as a hot gas and CO cooling. Gas jet PCR reactions (10 μl) were carried out in thin-walled glass capillary tubes containing 50 mM Tris (pH 8.5 at 25°C), 250 μg/ml BSA, 3 mM MgCl₂, 0.2 mM dNTPs, 50 pmol of forward and reverse primers, 20 picograms of template DNA, and 5 U of Taq Polymerase (Promega, Madison, WI). After amplification, reaction products were separated on 3% MetaPhor agarose gels with EtBr staining. Molecular wt. markers were 67-501 b.p. long (pUC19/Msp/DNA fragments).

As shown in FIG. 5 a, all four amplicons were amplified in high yield using pressurized gas jet PCR. Starting with 100 pg of DNA, 30 PCR cycles in 335 seconds, using 2.5 U Taq Pol, in a 15 μl reaction volume. The expected 536 b.p. β-globin gene, 364 b.p. HPA-4 allele, 333 b.p. bacteriophage λ 'D' gene, and 91 b.p. E.coli O157:H7 Stx2 amplicons were amplified through 30 cycles of [0 sec 92°C (denaturation)/0 sec 55°C (annealing)/5 sec 72°C (elongation)]. A low background of non-specific amplification products was observed in all four reactions. In the case of the λ 'D' gene amplicon, an unusually high yield was also observed (lane b). A temperature vs. time profile of this 335 second gas phase PCR experiment showed considerable improvement in performance over any previously reported thermocycler or thermocycling method (FIG. 4 and 5b).

EXAMPLE 2 Software Refinements, 30 PCR Cycles in 2:48 (168 seconds)

The tri-valve apparatus is a functional thermocycler. Not only is it very fast, but it exhibits good thermal control (±1°C). However, the performance of the device was limited by its software, which was written in BASIC code. Every time it needed to execute commands in response to thermal changes in the reaction chamber or process heater, it needed to translate from BASIC (interpreted program) to assembly code (compiled program).

A substantial amount of time (-1 sec/cycle) was lost due to this software limitation. In particular, during 30 cycles of gas phase PCR amplification, over 30 seconds of time (> 1 second/cycle) were nonproductively lost. Consequently, the system control software was re-written in assembly code for faster operation. This modification resulted in much faster gas jet PCR.

At a Taq Polymerase elongation rate of >80 nt/sec, it was expected that -1 sec/cycle spent during the transitions A-E + E-D (FIG. 4a) would be more than sufficient to copy a short 85 b.p. DNA template using a 30-mer PCR primer (T_m ~65°C). Accordingly, experiments were carried out with pressurized air heating/CO₂ cooling, using a relatively short (85 b.p.) amplicon from the E.coli O157:H7 Shigatoxin gene (Stx). Gas jet PCR reactions (10 μl) were carried out with 2.4 bar (36 p.s.i.) of pressurized air and 2.6 bar (-40 p.s.i.) of CO₂ cooling in thin-walled glass capillary tubes containing 50 mM Tris (pH 8.5 at 25°C), 250 μg/ml BSA, 3 mM MgCl₂, 0.2 mM dNTPs, 50 pmol of forward and reverse primers, 20 picograms of template E.coli O157:H7 DNA, and 5 U of Taq Polymerase (Promega). After 30 cycles of amplification, DNA fragments were separated on 3% MetaPhor™ agarose followed by E Br staining. Molecular weight markers were 67-501 b.p. long pUC19/ -τ?7 digest.

FIG. 6 shows that a short 85 b.p E.coli O157:H7 Stx amplicon could be amplified through 30 PCR cycles in high yield in <2 minutes; and a considerable improvement in performance was achieved by writing instrument commands in assembly code. Based upon the high thermal conductivity coefficient of helium gas (Ubbink, 1947; Bosworth, 1952; FIG. 1), it was hypothesized that even faster thermocycling would be possible if helium was used rather than air as a "hot" gas.

EXAMPLE 3

30 PCR Cycles in 1:12 (78 Seconds)

A small (85 b.p.) amplicon from the E.coli O157:H7 Stx gene was PCR-amplified using "hot" pressurized helium gas and CO₂ cooling. In order to further reduce the thermocycling time, primer lengths were increased to 30 nt, so that higher annealing temperatures (62°C to 63°C) could be employed. In addition, the DNA denaturation temperatures were slightly reduced (86°C to 89°C) from those used in previous experiments.

Three different high-speed gas phase thermocycling protocols were employed: (a) [0 sec 89°C/0 sec 62°C/0 sec 72°C]; (b) [0 sec 87°C/0 sec 62°C/0 sec 72°C]; and (c) [0 sec 86°C/0 sec 63°C/0 sec 72°C]. PCR reactions (10 μl) were carried out in thin-walled glass capillary tubes containing 50 mM Tris (pH 8.5 at 25°C), 250 μg/ml BSA, 3 mM MgCl₂, 0.2 mM dNTPs, 50 pmol of forward and reverse primers, 20 picograms of E.coli O157:H7 DNA, and 5 U of Taq Polymerase (Promega, Madison, WI). After 30 cycles of [0 sec 89°C/0 sec 62°C/0 sec 72°C]; (b) [0 sec 87°C/0sec 62°C/0 sec 72°C]; and (c) [0 sec 86°C/0 sec 63°C/0 sec 72°C], reaction products were separated on 3% agarose gels with EfBr staining. Molecular weight markers were 67-501 b.p. long (pUC19/Msp/ digest). As shown in FIG. 7a, a high yield of the expected 85 b.p. Stx amplicon was seen in all three helium/CO₂ gas jet PCR reactions. FIG. 7a and 7b demonstrate the three fastest DNA amplification reactions which have ever been carried out. The expected 85 b.p. E.coli O157:H7 Stx amplicon was amplified in all three reactions: (a) 1:25 = 85 seconds, (b) 1:21 = 81 seconds, (c) 1:18 = 78 seconds. This experiment also shows that thermocycling was considerably faster when pressurized helium gas was used rather than air. High-speed gas jet PCR also resulted in a very low background of non-specific "haze" or false reaction products. Presumably, false priming or elongation are rare when such fast thermocycling parameters are used; there is simply no time for spurious reaction products to accumulate.

Further refinements in the high-speed PCR jet thennocycler resulted in substantial improvements in software, thermal control, optimization of gas operating pressures, noise reduction, and the ability to amplify DNA fragments of different lengths from a variety of sources. Examples of high-speed PCR experiments in which these improvements were implemented are shown in Examples 4 through 7, as described below.

EXAMPLE 4

PCR Amplification of 368 b.p. B. anthracis DNA Fragment. 30 PCR Cycles in <4 Minutes (233 Seconds) The ability to rapidly amplify DNA from life-threatening bacterial pathogens such as Bacillus anthracis, an agent of biological warfare, is especially important (Franz et al, 1997). As shown in FIG. 8a and 8b, high-speed jet PCR amplification of a 368 b.p. DNA fragment from Bacillus anthracis, a biological warfare agent, was carried out using 167°C pressurized helium gas at 2.9 bar (44 p.s.i.) with CO₂ gas cooling at 3.0 bar (45 p.s.i.). B. anthracis Sterne strain DNA, supplied by the U.S. Army Medical Research Institute for Infectious Diseases (USAMRIID), was used as a template for PCR jet amplification using primers BAN23MS1 and BAN23MA1 (21-mers). Thirty PCR cycles of [0 msec 92°C, 0 msec 55°C, 3.2 sec 72°C] were carried out using 24 μl reactions containing 10 pg of DNA, 1 μM primers, 5.5 mM MgCl₂, 100 mM dNTPs, 500 μg/ml BSA, and 1.25 U of Taq Polymerase. B. anthracis Sterne strain DNA (10 pg) was mixed with 25 pmol forward primer (BAN23MS1) and reverse primer (BAN23MA1) plus 200 μM dNTPs in a 24 μl reaction volume containing 1.25 Units Taq Pol in (50 mM Tris-Cl, 5.5 mM MgCl₂, pH 8.8, 500 μg/ml BSA). High-speed PCR was carried out in a Model 1.62 Heliac pressurized gas thennocycler, which was programmed in 100 msec time intervals, followed by gel electrophoresis on 2% agarose gels with ethidium bromide staining. Samples were as follows: [a] 100 b.p. ladder molecular weight markers, [b] 30 PCR cycles using 2.6 bar (38 p.s.i.) hot compressed air with (3.0 bar) CO₂ cooling (0 sec 92°C, 0 sec 55°C, 4.0 sec

72°C), [c] 30 PCR cycles using hot He gas (2.9 bar) with 3.0 bar CO₂ cooling (0 sec 92°C, 0 sec 55°C, 4.0 sec 72°C), [d] 30 PCR cycles using 2.9 bar hot He, 3.0 bar CO₂ cooling (0 sec 92°C, 0 sec 55°C, 3.6 sec 72°C), [e] 30 PCR cycles using 2.9 bar hot He, 3.0 bar CO₂ cooling (0 sec 92°C, 0 sec 55°C, 3.2 sec 72°C). PCR-amplified DNA was loaded onto 2% agarose gels, followed by electrophoresis for 45 minutes at 80V. Total elapsed times for 30 high-speed PCR cycles were: [b] 4 minutes, 28 seconds, [c] 4 minutes, 20 seconds, [d] 4 minutes, 10 seconds, [e] 3 minutes, 53 seconds.

As shown in FIG. 8a (lane e), 30 cycles of high-speed gas jet thermocycling resulted in ~10⁸-fold amplification of the expected 368 b.p. fragment in as little as 3 minutes, 53 seconds (233 seconds). FIG. 8b shows a companion temperature versus time profile.

The reaction yield may be calculated as follows. Starting with 10 pg of B. anthracis DNA, approximately 100 ng of 368 b.p. reaction product was formed (FIG. 8a). Since the genome of B. anthracis contains -5.0 x 10 base pairs, and the 368 b.p. DNA fragment represents only a small portion of the bacterial chromosome, one may calculate that the initial 10 pg of bacterial DNA contained -[368 / (5.0 x 10⁶) x 10^"11 g] = 0.74 x 10^"15 g of 368 b.p. amplicon. This 0.74 x 10^"15 grams (= 0.74 femtograms) of template DNA was amplified to -100 ng of 368 b.p. reaction product. Therefore, an amplification factor of (1.00 x 10^"7 g)/(0.74 x 10^"15 g) = 1.35 x 10⁸-fold was achieved using 30 PCR cycles. If the Polymerase Chain Reaction were 100% efficient, then the amount of DNA would be expected to double after each PCR cycle. Theoretically, after 30 cycles of DNA doubling, 2³⁰ = 1.07 x 10⁹-fold amplification would be achieved, hi the present experiments, 1.35 x 10⁸-fold amplification was observed after 30 high-speed PCR cycles. It can be calculated that 1.866³⁰ = 1.34 x 10⁸. Therefore, each PCR cycle resulted in 1.866- fold amplification of bacterial DNA, conesponding to an efficiency of 1.866-1 = 0.866. In other words, each high-speed PCR cycle was -86.6% efficient. This experiment demonstrates that it is possible to amplify DNA fragments from special pathogens such as B. anthracis in approximately <4 minutes, starting with about 10 picograms of DNA.

EXAMPLE 5

PCR Amplification of E.coli and Human DNAs 145 to 486 b.p. long. 30 to 35 PCR Cycles in 2.4 to 4.1 minutes (144 to 247 seconds)

Additional experiments were carried out to demonstrate the speed and versatility of the high-speed gas jet thermocycler. For example, bacterial gene fragments 145 to 486 b.p. long and single-copy human gene fragments 147 to 191 b.p. long were amplified through 30 to 35 PCR cycles in -2.4 to 4.1 minutes (144 to 247 seconds) using 20 pg of DNA, 1 μM primers, 4.5 to 5.5 mM MgCl₂, 100 mM dNTPs, 500 μg/ml BSA, and 1.25 U of Taq Pol.

FIG. 9a shows a picture of DNAs amplified from the human gene responsible for Accute Intermittant Porphyria (191 b.p.), ABO blood type A-transferase gene fragment

(147 b.p.), E.coli uidA gene fragment (486 b.p.), E.coli uidA gene fragment (297 b.p.), and E.coli uidA gene fragment (145 b.p.). As shown in the temperature versus time profile for FIG. 9a (lane a) (FIG. 9b), excellent thermal control was possible when pressurized air was used as a hot gas at 2.6 bar (42 p.s.i.) with CO₂ gas cooling at 2.8 bar (46 p.s.i.).

EXAMPLE 6

Nested PCR Amplification of 486/186 b.p. E.coli uidA DNA Fragments. 40 PCR Cycles in 4.8 minutes (285 seconds)

More sensitive DNA amplification is possible when "nested PCR" protocols are employed. In particular, Ruzicka et al. (1992) have described a nested PCR method in which, first, a pair of outer primers with an annealing temperature of 45° to 55 °C are used in -20 cycles of PCR; followed by amplification from a set of inner (nested) primers, whose annealing temperature is >60 °C (FIG. 10b). hi this fashion, the two stages of outer/inner PCR can be carried out in a single, closed reaction vessel. Using the primers of Juck et al. (1996), nested PCR amplification of an outer 486 b.p. and inner 186 b.p. E.coli uidA DNA fragment were carried out using: (1) 20 cycles of [200 msec 88°C, 200 msec 56°C, 2.8 sec 72°C], followed by (2) 20 cycles of [200 msec 88°C, 200 msec 65°C, 1.8 to 2.0 sec 72°C]. The first 20 cycles allow annealing of all 4 oligonucleotide primers at the pennissive 56°C annealing temperature using a relatively long 2.8 second Taq Pol elongation time. The second 20 cycles selectively amplify the inner 186 b.p. DNA fragment due to the higher 65°C annealing temperature and the shorter 1.8 to 2.0 second elongation step.

As shown in FIG. 10a PCR reactions (15 μl) were carried out in thin-walled glass capillary tubes containing 50 mM Tris (pH 8.5 at 25°C), 250 μg/ml BSA, 3 mM MgCl₂, 0.2 mM dNTPs, 50 pmol of forward and reverse primers, 4.5 pg of E.coli K12 DNA, and 1.5 U ofZ-Taq Pol (Takara Shuzo Ltd). After 20 cycles of (a) [100 msec 89°C/100 msec 65°C/2.5 sec 72°C] and then 20 cycles of [100 msec 90°C/100 msec 61°C/1.8 sec 72°C], reaction products were separated on 2.0% MetaPhor agarose gels with EfBr staining. M.W. markers were pUCl9/MspI digest fragments. The expected 186 b.p. uidA amplicon in lanes (a) and (b) ran just below the 190 b.p. M.W. marker. Total time for 40 high-speed PCR cycles was 4 minutes and 45 seconds (285 seconds).

As shown in FIG. 10a (lane a), starting with 1 picogram of E.coli O157:H7 DNA, the expected 186 b.p. inner amplicon could be amplified through 40 cycles of two-step nested PCR in less than 5 minutes when the second 20 cycles were [200 msec 88°C, 200 msec 65°C, 2.0 sec 72°C]. As shown in FIG. 10a (lane b), if a shorter 1.8 second/cycle Taq Pol elongation was employed, 40 cycles of nested PCR was achieved in 4 minutes, 45 seconds (total PCR time for 40 cycles = 285 seconds). The amplification factor may be calculated as follows: the E.coli genome contains -4.5 x 10⁶ b.p. The 186 b.p. E.coli DNA fragment represents (1.86 x 10²/ 4.5 x 10⁶) = 4.1 x 10^"5 of the bacterial genome. An initial 1 picogram (10^"12 g) of bacterial DNA was amplified to -50 ng (5 x 10^"8 g) quantities (FIG. 10a). Therefore, the 186 b.p. E.coli uidA DNA fragment was amplified (1/ 4.1 x 10^"5 ) x (5 x 10^"8 / 10^"12) = 1.2 x 10ⁿ-fold in a total nested PCR time of 285 seconds.

EXAMPLE 7

PCR Amplification of Bacteriophage λ DNAs 2206 to 2331 b.p. long. 30 PCR Cycles in 10.3 minutes (617 seconds)

FIG. 11a shows a picture of a 2206 b.p. bacteriophage λ DNA fragment which was amplified using 20 pg of DNA, 1 μM primers, 4.5 mM MgCl₂, 100 mM dNTPs, 500 μg/ml BSA, and 1.5 U oϊZ-Taq Pol (Takara Shuzo Ltd) and 30 cycles of [800 msec 89° to 91°C, 800 msec 61°C, 13 sec 72°C] in a gas jet thermocycler, using 2.6 bar (42 p.s.i.) compressed air (from a pneumatic airtool compressor) and 2.8 bar (46 p.s.i.) CO₂ cooling. Lane (a) shows the expected 2206 b.p. reaction product using a denaturation temperature of 91°C (total PCR time = 10 minutes, 49 seconds). Lane (b) shows the 2206 b.p. reaction product using a denaturation temperature of 90°C (total PCR time = 10 minutes, 28 seconds). Lane (c) shows the 2206 b.p. reaction product using a denaturation temperature of 89°C (total time for 30 PCR cycles: 10 minutes, 3 seconds = 603 seconds). Molecular weight markers were bacteriophage λ Hindrfl digest fragments.

FIG. lib shows a picture of a 2331 b.p. bacteriophage λ DNA fragment which was amplified using 20 pg of DNA, 1 μM primers, 4.5 mM MgCl₂, 100 mM dNTPs, 500 μg/ml BSA, and 1.5 O Z-Taq Pol (Takara Shuzo Ltd) and 30 cycles of [500 msec 91°C, 500 msec 61°C, 11 to 14 sec 72°C] in a gas jet thermocycler, using 2.6 bar (42 p.s.i.) compressed air (from a pneumatic airtool compressor) and 2.8 bar (46 p.s.i.) CO₂ cooling. Lanes (a) through (d) show the expected 2331 b.p. reaction product using an elongation time of: (a) 14 seconds at 72°C (total PCR time = 10 minutes, 50 seconds); (b) 13 seconds at 72°C (total PCR time = 10 minutes, 17 seconds); (c) 12 seconds at 72°C (total PCR time = 9 minutes, 42 seconds); (d) 11 seconds at 72°C (total PCR time = 9 minutes, 8 seconds). The reaction yield was clearly higher when elongation times of >12 seconds were employed, as shown in lanes (a) and (b). This result indicated that, for phage λ DNA fragments 2206 to 2331 b.p. long, Z-Taq Pol had an apparent elongation rate of >194 nt/sec.

FIG. 11a and lib demonstrate that DNA fragments 2206 to 2331 b.p. long have been successfully amplified in high-yield using the pressurized gas jet thermocycler. Figure 13 shows a summary of the high-speed DNA amplification processes using an embodiment of the invention which employs the tri- valve gas jet thermocycler.

Example 8

PCR Amplification of E.coli uidA DNA Fragments in a Bi-valve Jet Thermocycler.

FIG. 13 shows a picture of a 297 b.p. E. coli uidA DNA fragment which was amplified using 1 ng DNA, 150 pM primers, 3 mM MgSO4, 200 μM dNTP, 400 μg/ml BSA, and 0.625 U KOD Hot Start polymerase. Lane 1 was 35 cycles of [30 sec 90°C (hot start), 0 sec 90°C; 0.2 sec 58°C; 2 sec 72°C], Lane 2 was 40 cycles of [30 sec 90°C (hot start), 0 sec 90°C; 0.2 sec 58°C; 2 sec 72°C], and Lane 3 was 30 cycles of [30 sec 90°C (hot start), 0 sec 90°C; 0.2 sec 58°C; 2 sec 72°C] in a bi-valve gas jet thennocycler, using 2.4 bar (35 p.s.i.) compressed air. 25 μl of sample were placed in a Roche glass capillary tube. This result indicated that high-speed jet PCR could be carried out in reaction vessels which are suitable for on-line fluorescence-based DNA detection. Additionally, superior thermal control (plus or minus 0.7°C) is possible using a 2-stage flow conditioner and CFD-optimized chamber geometry.

Some of the major advantages of the invention are as follows: First, the invention decreases the time needed for amplification of DNA using the Polymerase Chain Reaction by one to two orders of magnitude over any previously described process or device (see FIG. 12). Second, the gas jet process is inherently more accurate than slower procedures, since false reaction products have practically no time to anneal and/or elongate (see FIG. 5a, 6, 7a, 8a, 9a, 11a and lib). Third, the pressurized gas jet thennocycler (FIG. 3a) has superior timing control as compared to heat block or hot-air devices; the gas jet device is programmable in time intervals of 100 milliseconds. Fourth, high-speed gas jet PCR is a very reliable process. Except for the electromechanical relays and valves, no moving parts are employed. Fifth, the high-speed gas jet PCR process is fully compatible with on-line detection optics, so that timely DNA amplification/detection can be carried out. The invention therefore allows fast, accurate, and reliable DNA amplification/detection to be achieved on an unprecedented time scale.

Comparison to 'Ultrafast' PCR Thermocyclers. The fastest experimental thermocyclers built by Lawrence Livermore National

Laboratory (Northrup et al, 1998), the University of Pittsburgh (Oda et al, 1998), the

University of Washington (Friedman and Meldrum, 1998), and the Massachusetts Institute of Technology (Chiou et al, 2001) required 8.5 to 23 minutes for 30 PCR cycles.

For example, Oda et al. (1998) described "Ultrafast PCR" in which "cycle times as fast as 17 seconds could be achieved." Unfortunately, the infrared heating method used by

Oda et al. is incompatible with fluorescent optical detection using >450 nm dyes and photodiode detectors.

Chiou et al. (2001) stated that "Thirty-cycle PCR runs have been conducted which result in up to 78% efficiency for 30-cycle PCR of a 500-bp target from genomic λ DNA in 23 min." (p. 2021). Thirty PCR cycles at this efficiency results in 1.56³⁰ = 6.2 x 10⁵-fold amplification at a speed of 23 minutes/30 cycles = 46 seconds/cycle.

The pressurized gas device shown in FIG. 3a requires less than 2.6 seconds/cycle; and the reaction yield is over 200-fold greater (see FIG. 5a, 6, 7a, 8a, 9a, 11a, and lib).

For example, as shown in FIG. 8a, a single copy bacterial gene fragment 368 b.p. long was amplified 1.35 x 10⁸-fold in a total of 233 seconds for 30 PCR cycles.

Kopp et al. (1998) have described a miniature continuous-flow PCR device was able to amplify a 176 b.p. DNA fragment through 20 cycles in 3 to 4 minutes. However,

-10⁸ copies of template DNA were required as starting material (Zorbas, 1999).

Extrapolating from the data of Kopp et al. (1998), 30 cycles of PCR amplification would require 4.5 to 6 minutes; and the yield of amplified DNA is 10³- to 10⁴-fold lower than that obtained using the pressurized gas device shown in FIG. 3 a.

Comparison to State-of-the-Art Commercial Devices.

The fastest commercially available thermocycler is manufactured by the

Boehringer-Mannheim Division of Roche Modular Systems, a division of Roche, under license from Idaho Technology (U.S. Pat. No. 5,455,175 to Wittwer et al). This hot-air thermocycler requires about 9.5 minutes for 30 cycles of amplification of the 536 base pair β-globin DNA fragment. However, with its on-line detection optics attached, the Lightcycler™ requires -30 minutes for 30 PCR cycles.

As shown in FIG. 7a and 7b, a pressurized gas PCR process can be used to amplify DNA to detectable quantities in as little as 78 seconds. The high-speed pressurized gas jet process is therefore >15 times faster than the Roche machine and is compatible with online fluorescent dye-based DNA detection optics.

Higher-yield amplification (>10⁸-fold) has been achieved with longer elongation times (3.2 sec/cycle at 72°C). For example, using a 368 b.p. B. anthracis amplicon, 30 PCR cycles were achieved in less than 4 minutes (FIG. 8a and 8b). Shorter (145 to 486 b.p.) bacterial DNA fragments were amplified ~10⁸-fold in 144 to 280 seconds (FIG. 5a, 9a and 12). A general description of the present invention as well as a prefened embodiment has been set forth above. Those skilled in the art will recognize and be able to practice additional variations in the methods and devices described which fall within the teachings of this invention. Accordingly, all such modifications and additions are deemed to be within the scope of the invention which is to be limited only by the claims appended hereto.

REFERENCES CITED

U.S. PATENT DOCUMENTS

1,952,281 x/1934 Ranque.

4,683,202 7/1987 Mullis.

5,187,084 2/1993 Hallsby.

5,455,175 10/1995 Wittwer et al. .

5,576,218 11/1996 Zurek et al. .

5,779,977 7/1998 Haffet α/. .

5,783,439 7/1998 Reichler et al. .

6,200,781 3/2001 Tal et al. .

OTHER PUBLICATIONS

Anderson, JD, Jr. (1995) Computational Fluid Dynamics: The Basics with Applications.

McGraw -Hill, Inc., New York. Azbel D (1984) Fundamentals of Heat Transfer for Process Engineering, Noyes Publications, Park Ridge, New Jersey, pp. 12-20. Bosworth RCL (1952) Heat Transfer Phenomena: the Flow of Heat in Physical Systems, John Wiley & Sons, Inc., New York, Chapter II, "The Thermal Conductivity of Gases." Chapman AJ (1984) Heat Transfer. Fourth Edition, The Macmillan Company, New York. Chiou J, Matsudaira P, Sonin A, and Ehrlich D (2001) "A Closed-cycle Capillary Polymerase Chain Reaction Machine." Anal. Chem. 73: 2018-2021. Erlich HA, ed. (1989) PCR Technology: Principles and Applications for DNA Amplification, Stockton Press, New York. Fourier JBJ (1822) Theorie Analytique de la Chaleur. Fermin Didot, Paris. Translated by A. Freeman (1988) as The Analytical Theory of Heat, Dover Paperback Books, New York. Friedman NA and Meldrum DR (1998) "Capillary tube resistive thermal cycling." Anal. Chem. 70: 2997-3002. Franz DR, Jahrling PB, Friedlander AM, McClain DJ, Hoover DL, Bryne RW, Pavlin JA, Christopher GW, and Eitzen EE (1997) "Clinical Recognition and Management of

Patients Exposed to Biological Warfare Agents." J. Am. Med. Assoc. 278: 399-411. Gelfand DH and White TJ (1990) "Thermostable DNA Polymerases." In: MA h is, DH

Gelfand, JJ Sninsky, and TJ White, eds., PCR Protocols: a Guide to Methods and Applications, Academic Press, hie, San Diego.

Haugland RP (1996) "Nucleic Acid Stains," in Handbook of Fluorescent Probes and

Research Chemicals, Sixth Edition, Molecular Probes, Inc., Eugene, OR, pp.144-150. Higuchi R, Dollinger G, Walsh SP, and Griffith R (1992) "Simultaneous amplification and detection of specific DNA sequences." Bio/Technology 10: 413-417. Hilsch R (1947) "The Use of the Expansion of Gases in a Centrifugal Field as Cooling

Process." Ann. Rev. Scientific histruments 18(2): 108-113. Idaho Technology, Inc. (1995) The Rapid Cyclist 3(1): 16, Idaho Falls, Idaho, inis, MA, Myambo KB, Gelfand DH, and Brow MAD (1988) "DNA sequencing with

Thermus aquaticus DNA polymerase and direct sequencing of polymerase chain- reaction amplified DNA. " Proc. Nat. Acad. Sciences USA 85 : 9436-9440.

Johnson B (1998) "The Competition Heats Up. The annual review of thermal cyclers takes a sneak peak at the new products for 1998." The Scientist, 12 (24): "Thermal Block

Table." Johnston HL and Grilly ER (1946) "The thermal conductivities of eight common gases between 80° and 380°K," J. Chem. Physics 14: 233-238.

Juck D, Ingram J, Prevost M, Coallier J, and Greer C (1996) "Nested PCR protocol for the rapid detection of Escherichia coli in potable water." Can. J. Microbiol. 42: 862-866. Kenebrock JL (1969) Aircraft Engines and Turbines, MIT Press, Cambridge, MA. Kogan SC, Doherty M, and Gitschier J (1987) "An improved method for prenatal diagnosis of genetic diseases by analysis of amplified DNA sequences." New England J. Med.

317: 985-990. Kopp MU, Mello AJ, and Manz A (1998) "Chemical amplification: continuous-flow PCR on a chip." Science 280: 1046-1048. Mullis K, Fene F, and Gibbs RA, eds. (1994) The Polymerase Chain Reaction, Birkhauser, Boston. Navier CLMH (1856) Resume des Lecons d' analyse donnees a "1"' Ecole Polytechnique,

Suivi de notes par MJ. Liouville, Deuxieme edition revue par M. Ernest Liouville,

Victor Dalmont, Paris. Newton CR (1995) "PCR Instruments," h : CR Newton, Essential PCR Data. John Wiley & Sons, Chichester, England, pp. 12-23.

Nicoll S, Brass A, and Cubie HA (2001) "Detection of heφes viruses in clinical samples using real-time PCR." J. Virol. Methods 96: 25-31. Northrup MA, Benett B, Hadley D, Landre P, Lehew S, Richards J, and Stratton P (1998)

"A miniature analytical instrument for nucleic acids based on micromachined silicon reaction chambers." Anal. Chem 70: 918-922.

Oda RP, Strausbauch MA, Huhmer AF, Jurrens SR, Craighead J, Wettstein PJ, Eckloff B,

Kline B, and Landers JP (1998) "Infrared-mediated thermocycling for ultrafast polymerase chain reaction amplification of DNA." Anal. Chem. 70: 4361-4368. Oste CC (1989) "PCR Automation." In: Erlich HA, ed., PCR Technology: Principles and Applications for DNA Amplification, Stockton Press, New York, pp. 23-30.

Oste CC (1994) "PCR histrumentation: Where Do We Stand?" In: K Mullis, F Fene, and

RA Gibbs, eds. (1994) The Polymerase Chain Reaction, Birkhauser, Boston, pp.165-

173. Persing DH, Smith TF, Tenover FC, and White TJ, eds. (1993) Diagnostic Molecular Biology: Principles and Applications, ASM Press, Washington, D.C.

Ruzicka V, Marz W, Russ A, and Gross W (1992) "Automatic separation of two PCRs in one tube by annealing temperature." Trends in Genetics 11: 127-128. Saiki RK, Scharf S, Faloona F, Mullis KB, Horn GT, Erlich HA and Arnheim N (1985)

"Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia." Science 230: 1350-1354.

Saiki RK, Walsh PS, Levenson PH, and Erlich HA (1989a) "Genetic analysis of amplified

DNA with immobilized sequence-specific oligonucleotide probes." Proc. Nat. Acad.

Sciences USA 86: 6230-6234. Saiki RK (1989b) "The design and optimization of the PCR." hi: PCR Technology: Principles and Applications for DNA Amplification. HA Erlich, ed., Stockton Press,

New York, pp. 17-22. Sobczak K, Kozlowski P, and Krzysosiak WJ (1995) "Faster and cheaper PCR on a standard thennocycler." Acta Biochimica Polonica 42: 363-366. Spears PA, Linn CP, Woodard DL, and Walker GT (1997) "Simultaneous Strand Displacement Amplification and Fluorescence Polarization Detection of Chlamydia trachomatis DNA." Anal. Biochem. 247: 130-137.

Tret'yakov AN, Gel'fand VM, Pantina RA, Shevstov SP, and Bulat SA (1994) "Rapid

DNA Amplification in Pipettor Tips." Molecular Biology (Moscow) 28: 441-443. Ubbink JB (1947) "Thermal conductivity of gaseous helium." Physica 13: 629-634; 659- 668. Welty JR, Wicks CE, and Wilson RE (1976) Fundamentals of Momentum, Heat, and Mass Transfer, Second Edition, John Wiley & Sons, New York. Wittwer CT, Filmore GC, and Hillyard DR (1989) "Automated polymerase chain reaction in capillary tubes with hot air." Nucleic Acids Research 17: 4353. Wittwer CT, Filmore GC, and Garling DJ (1990) "Minimizing the time required for DNA amplification by efficient heat transfer to small samples." Anal. Biochem. 186: 328-

331. Wittwer CT and Garling DJ (1991) "Rapid cycle amplification: Time and temperature optimization." BioTechniques 10: 76-83. Wittwer CT, Reed GB, and Ririe KM (1994) "Rapid Cycle DNA Amplification." In: K Mullis, F Fene and RA Gibbs, eds., The Polymerase Chain Reaction, Birkhauser,

Boston, pp.174-181. Zorbas H (1999) "Miniature Continuous-Flow Polymerase Chain Reaction: A Breakthrough Angew. Chem. frit. Ed. 38: 1055-1058.

Claims

What is claimed is:

1. A high-speed process for amplifying DNA, comprising:

(a) providing a reaction chamber containing a biological sample having DNA, a DNA polymerase, oligonucleotide primers, and deoxynucleotide precursors, the reaction chamber accepting flow of one or more heat transfer gases;

(b) heating a first heat transfer gas in a heat chamber that is physically separated from the reaction chamber;

(c) delivering the first heat transfer gas to the reaction chamber at a pressure greater than that of the reaction chamber, wherein heat is fransfened from the first heat transfer gas to the DNA in order to denature the DNA;

(d) delivering a second heat transfer gas to the reaction chamber to cool the reaction chamber to a temperature where the denatured DNA is annealed to the oligonucleotide primers; and

(e) increasing the temperature of the reaction chamber to a temperature sufficient to allow for enzyme-catalysed elongation of primer:template complexes.

2. The process of claim 1, further comprising: repeating steps (c) through (e) until a desired quantity of amplified DNA has been produced.

3. The process of claim 1 , further comprising regulating the flow of the first heat transfer gas from the heat chamber to the reaction chamber using at least one multiport valve.

4. The process of claim 1, wherein the first heat transfer gas and the second heat transfer gas are the same type of gas.

5. The process of claim 1, wherein the first heat transfer gas and the second heat transfer gas are different gases.

6. The process of claim 1, wherein the first heat transfer gas is at a pressure greater than or equal to 1.3 bar (> 20 pounds per square inch).

7. The process of claim 1, wherein the second heat transfer gas is at a pressure greater than or equal to 1.3 bar (> 20 pounds per square inch).

8. The process of claim 1, wherein the first heat transfer gas is air.

9. The process of claim 1, wherein the first heat transfer gas is helium or its isotopic mixtures thereof, including helium-3.

10. The process of claim 1, wherein the first heat transfer gas is selected from the group consisting of hydrogen, neon, methane, nitrogen, argon, krypton, carbon dioxide, and mixtures thereof.

11. The process of claim 1 , wherein the second heat transfer gas is carbon dioxide.

12. The process of claim 1, wherein the second heat transfer gas is air cooled to less than 20°C by a Ranque-Hilsch vortex tube before delivery to the reaction chamber.

13. The process of claim 1, wherein the second heat transfer gas is selected from the group consisting of air, hydrogen, neon, methane, argon, krypton, carbon dioxide, and mixtures thereof.

14. The process of claim 1, wherein the second heat transfer gas is cooled to a temperature less than 20°C prior to entering the reaction chamber.

15. The process of claim 1 , further comprising regulating the flow of the first and second heat transfer gases into the reaction chamber using one or more electronic multiport valves.

16. The process of claim 1, further comprising using one or more flow conditioner to suppress pressure pulsations and temperature fluctuations.

17. The process of claim 1, wherein the first and second heat transfer gases flow from their respective flow conditioners into a mixing chamber to achieve a uniform temperature.

18. The process of claim 16, further comprising actuating the electronic multiport valves via electrical signals.

19. The process of claim 1, further comprising mixing the first and second heat transfer gases within the reaction chamber using one or more gas flow conditioners.

20. The process of claim 1 , wherein a holding time for DNA denaturation, a holding time for primer annealing, and a holding time for enzymatic elongation are programmable in time intervals of fractions of a second.

21. An apparatus for high-speed amplification of DNA, comprising: (a) a reaction chamber having an operating pressure greater than or equal to 1.3 bar;

(b) a heating chamber, physically separated from the reaction chamber, fluidly connected to the reaction chamber, the heating chamber comprising a first container, having a heating gas at a pressure greater than the reaction chamber pressure, fluidly connected to the heating chamber;

(c) a flow conditioner, fluidly connected to the heating chamber and the mixing chamber, to suppress pressure pulsations;

(d) a second container, having a cooling gas, fluidly connected to the reaction chamber; (e) a flow conditioner, fluidly connected to the cold gas supply chamber and the mixing chamber, to suppress pressure pulsations in the cold gas flow;

(f) a heating gas inlet valve positioned between the heating chamber and the hot gas flow conditioner;

(g) a cooling gas inlet valve positioned between the second container and the cold gas flow conditioner; and

(h) a programmable controller having inputs and outputs, wherein outputs are connected to the heating gas inlet valve and the cooling gas inlet valve to control opening and closing of the valves; and a first temperature sensor coupled to the reaction chamber, the first temperature sensor connected to an input of the programmable controller, whereby the controller opens and closes the heating gas inlet valve and the cooling gas inlet valve in response to the temperature sensor to maintain the temperature of the reaction chamber at a desired setting.

22. The apparatus of claim 21, wherein the first heat transfer gas pressure is greater than or equal to 1.3 bar.

23. The apparatus of claim 21, wherein the first heat transfer gas is air.

24. The apparatus of claim 21, wherein the first heat transfer gas and the second heat transfer gas are the same kind of gas.

25. The apparatus of claim 21, wherein the first heat transfer gas and the second heat transfer gas are different gases.

26. The apparatus of claim 21, wherein the first heat transfer gas is helium and the second heat transfer gas is carbon dioxide.

27. The apparatus of claim 21, wherein the first heat transfer gas is air and the second heat transfer gas is carbon dioxide.

28. The apparatus of claim 21, wherein the first heat transfer is air and the second heat transfer gas is air cooled using a Ranque-Hilsch vortex tube.

29. The apparatus of claim 21, further comprising a flow conditioner connected to a mixing chamber to mix hot gas and cold gas.

30. The apparatus of claim 21, further comprising at least one electronic valve to control pressure and gas flow into the reaction chamber.

31. The apparatus of claim 21 , further comprising a noise-reducing muffler assembly operatively connected to an exhaust gas outlet.

32. The apparatus of claim 21, further wherein at least one output of the programmable controller is connected to a heating element of the heating chamber to turn the heating element on and off.

33. The apparatus of claim 21 , wherein the reaction chamber has a heat transfer coefficient of less than 0.5 Watts/centimeter-degree K.

34. The apparatus of claim 21 , wherein the reaction chamber comprises a polyurethane plastic.

35. The apparatus of claim 21, further comprising an optical detection device to facilitate fluorescence-based DNA detection.

36. The apparatus of claim 35, wherein the optical detection device comprises a light emitting diode, a collimating lens, a filter, and a photodiode.

37. The apparatus of claim 35, wherein the optical detection device comprises a laser diode, a collimating lens, a filter, and a photodiode.

38. A high-speed process for amplifying DNA, comprising:

(a) providing a reaction chamber containing a biological sample having DNA, a DNA polymerase, oligonucleotide primers, and deoxynucleotide precursors, the reaction chamber having a pressure;

(b) heating helium gas in a heat chamber that is physically separated from the reaction chamber;

(c) delivering the heated helium gas to the reaction chamber, wherein heat is transferred from the helium gas to the DNA in order to denature the DNA; (d) delivering a cooling gas to the reaction chamber to cool the reaction chamber to a temperature where the denatured DNA amieals to the oligonucleotide primers; and

(e) increasing the temperature of the reaction chamber to a temperature sufficient to allow for elongation of primer template complexes.

39. The method of claim 38, wherein the cooling gas is carbon dioxide.

40. The method of claim 38, wherein the second heat transfer gas is air, which has been cooled to a temperature of less than 20°C by a Ranque-Hilsch vortex tube before delivery to the reaction chamber.