Conventionally, DNA concentration is determined spectrophotometrically, a method which relies on the characteristic absorption of ultra-violet light (ca.

26on) by the nucleotide ring structure of DNA molecules. To determine molecular weight, DNA molecules are normally size-separated by agarose gel electrophoresis and then visualised using the dye ethidium bromide. DNA can also be quantified in agarose gels by comparing with known standards by measuring the fluorescence emitted following excitation by ultra-violet light (ca. 300nm). Both methods have inherent disadvantages. UV spectrophotometers are expensive pieces of equipment requiring the use of costly quartz curvettes and rely on the 'destructive' processing of relatively large sample volumes. The use of ethidium bromide for DNA visualisation and quantifi- cation, although a cheaper alternative, is also undesirable due to its extremely toxic and carcinogenic nature.

The present invention comprises a method for the estimation of a property or parameter of a nucleic acid material, or of a process in which a nucleic acid is modified by a chemical or biochemical reaction, said property or parameter being one to which the electrical conductivIty of the nucleic acid material is related, which method comprises measuring the electrical conduc- tivity of the nucleic acid material, or of a reaction mixture containing said material, and estimating from said measurement the property or parameter of the material or process by reference to a predetermined relationship between electrical conductivity and said property or parameter.

The present invention is based on the discovery that there are certain important properties of nucleic acids, the quantitative determination of which is frequently desirable, which can be assessed by measure- ment of the conductivity of a solution of the nucleic acid or acids. Changes in such properties may be reflected in corresponding changes of electrical conductivity. The concentration of nucleic acid in solution and the molecular weight of a species of nucleic acid are examples of important property which may be determined in accordance with the present inven- tion. Changes of molecular weight occurring in the course of an enzymatic processing of DNA or other nucleic acids may therefore be determined by monitoring changes in electrical conductivity of the solution or reaction mixture.

For the purposes of this invention electrical conduc- tivity may be conveniently measured as the electrical current flowing through a solution of the nucleic acid material.

The present invention is of primary interest for the determination of properties or changes in properties in a single species of nucleic acid. However, absolute purity of the material is not always necessary and the invention is applicable to nucleic acids containing minor amounts or other materials, including other species of nucleic acid.

The concentration of DNA may be determined by measur- ing the crrent/conductivity at a known alternating current frequency. The current recorded at any one fixed frequency has been found to be proportional to the DNA concentration. For example Figure 1 shows how the currer/conductivity recorded at 2KHz varies with DNA concentration. This relationship applies not only to a single size of nucleic acid but is true for a range of different sized molecules. In designing apparatus for carrying out the method of this invention a conductivity meter may be readily adapted and calibrated in accordance with the predetermined relationship between current flow and concentration of nucleic acid. In practice it will usually be desirable to calibrate the instrument to deal with homogeneous DNA species but for a range of molecular weights. Thus the sample will normally first be "sized" following which the appropriate nucleic setting for concentration determination will be chosen.

The molecular weight of a DNA species is determined by the response of these molecules to varying frequen- cies applied across the electrodes. Plotting the current/conductivity recorded (Table 1) as percentage of the maximum current/conductivity response ( response), versus the frequency of the a.c. signal applied between the two electrodes gives characteristic curves which differ for the molecular weights of the molecules concerned (Figure 2).

Table 1 current (uA) freauencv 25 mer pucl8 lambda 2.00E+03 6 62 93 4.00E+03 6 63 96 6.00E+03 6 64 98 8.00E+03 6 64 99 1.00E+04 7 64 99 2.OOE+04 8 65 100 4.00E+04 11 65 102 6.00E+04 15 66 102 8.00E+04 19 67 102 1.00E+05 24 68 104 2.00E+05 45 78 109 4.00E+05 80 99 120 6.00E+05 98 109 122 For a single DNA species the molecular weight can be determined by firstly calculating the gradient of the response versus frequency curve (over the frequency range 0 - 5x105 Hz). The gradient value can then be compared with a calibration curve (figure 3) of log molecular weight plotted against gradient. For all DNA molecules tested, the gradient varies with molecular weight such that the larger the gradient value, the lower the molecular weight. The basis of this relationship is presumed to be that the mobility of the DNA molecule changes as the frequency changes, as the frequency increases, large molecules are less respon- sive to changes compared with smaller molecules.

Methods of DNA quantification and molecular weight determination in accordance with the invention circum- vent the problems associated with the known methodolo- gies and offer a number of distinct advantages over those conventional methods. Thus, the invention allows rapid and accurate determination of both molecular weight ant. concentration; it is more sensitive and accurate, and it requires small sample volumes.

The characterisation of DNA outlined above has been achieved using an a.c. signal, of modulating frequency, between two thin wire platinum electrodes (alternative metals such as copper, stainless steel would also be adequate) through a solution containing DNA. Generally a fixed a.c. signal of between 0 to 10 volts is applied between the two thin wire conductive electrodes. A range of frequencies of this a.c. signal, between 0 and 1MHz, is applied across these electrodes and the correspondIng conductivity recorded at each frequency in turn as the current passing through the solution.

For example the sample DNA is dissolved in water or TE buffer in a volume of at least 10 41. For convenience a standard 500 41 plastic tube can be used. Two plati- num thin wire electrodes are placed in the solution and connected to a function generator operating at 1 V a.c.

A range of frequencies from 0-1 MHz are passed through the solution and the resulting current measured as mA using an ammeter. The DNA concentration is calculated by comparing the current passing through the solution at a frequency of 2 KHz with a standard curve (relating DNA concentration to current at a fixed frequency).

Alternative frequencies can also be used.

Specific Example Confirmation that the molecular weight of DNA molecules in solution can be determined by conductivity methods was achieved by preparing four different DNA solutions in both milli Q water and Tris EDTA (TE) buffer (25 base oligonucleotide, 700 base pair fragment, pUC 18 [=2,690 base pair] and lambda [=50,000 base pair]) and a fifth solution, containing DNA of unknown molecular weight (UNI. The conductivity of these solutions was measured i A over the range of frequencies from 2 x 104 Hz to 105 Hz. There was no significant difference between mlli Q and TE buffer indicating that DNA molecules zan be accurately sized in the most common buffer used to store DNA.

To confirm the relationship between molecular weight and conductIvity, solutions were prepared in triplicate and tested. Table 2 provides this confirmation and shows that interreplicate variation is slight. Table 3 shows the mean conductivity in uA and also as a W of the maximum conductivity. The main reason for express- ing data in terms of a response is that although within experiment variation is low (as shown in Table 1) the fragility of the current probe system causes signifi- cant variation between experiments in terms of uA readings. However the trends of conductivity changes (as expressed in t response) is consistent between experiments.

Figure 4 shows the mean W response date from table 3 plotted as a graph. The most obvious difference between the DNA solutions is the gradient of the slopes. When this gradient is calculated from the conductivity at 2 x 104 and 4 x 105 a near linear relationship is observed between gradient and log molecular weight (figure 5). Thus gradient can be used to calculate molecular weight.

The molecular weight of sample UN found from Fig 5 is about 1047 bps, which corresponds well with estimates made from agarose gel sizing of the same fragment.

Table 2: Conductivity (pA)(A) (pA)(A) of DNA solutions over a range of frequencies replicates mean (A) frequency size 1 2 3 I zPI 21( 22 zi 2122 21 22 22 36 36 35 54 56) 56 1 55 80 86( 82 82 111 115 116 114 32 ' ' 35351 35351 33 33 35 37 38 36 40 40 43 41 52 --- --- 52 52 54 55 53 63 .69 69 ~ ~ 67 90 | | 96 93 93 115 118 119 117 51 53 53 52 55 ~ ~ 58 55 56 60. 60 60. 60 64 62 62 65ES 65ES 71 68 68 69 75 75 73 94 100 100 98 120 130 125 125 68 c c 68 72 70 69 75 70 71 80 79 78 75 80 79 78~75 ~ 80 79 ~ 78 75 80 79 78~75 ~ 80 79 ~ 78 76 77 80 77 78 81 - 82 80 102 104 111 106 135 ~ ~ 145 ~ ~ 139 139 341 1371 138 136 138 142 152 144 139 147 150 145 142 ~ ~ 143 ~ ~ 153 146 145T 14S 145T 14S 154 146 7 7 ~ ~ 148 151T lSl 151T lSl 154 1637163 1637163 156 162 169 169 166 2.00E + 04 4.00E + 04 6.00E + 04 25b 8.00E + 04 1.00E + 05 2.00E + 05 4.00E + 05 2.00E + 04 4.00E + 04 6.00E + 04 700 bp 8.00E + 04 l+ l+ OOE ooze OOE ooze + + 03 2.00E + 05+ OS + 05+ OS d.OOEu.OOE d.OOEu.OOE + 05 2.00E + 04 4.00E + 04 6.00E + 04 un 8.0or 8.0or + 04 l.OCE1.002- l.OCE1.002- + 05+ OS + 05+ OS 2.00EHz 2.00EHz + 05 4.OGE4.OGRE 4.OGE4.OGRE +05t 05 +05t 05 2.OCE2.owe 2.OCE2.owe + 04 4.0cm 4.0cm + 04 6.00 +6.00E + 6.00 +6.00E + 04 8.3cm 8.3cm + 04 l.OGEloge l.OGEloge + 05 2,690 bp 2.00 + 05 4.OGEA.0O2- 4.OGEA.0O2- + 05 2.002-2.0C~ 2.002-2.0C~ + 04 4.OGE4.0GE 4.OGE4.0GE + 04 6.or3 6.or3 + 04 8.00E8.005 8.00E8.005 + 04 1.OCE + OSOR OSOR 50,000 bp <BR> <BR> 2.0on 2.0on + 03 <BR> <BR> <BR> <BR> 4.0or 4.0or + 05+ OS + 05+ OS Table 3. Conductivity (as pA and % response) of DNA solutions over a range of frequencies.

25b 700 bp un 2,690 bp 50,000 bp 21121 21121 33 52 70 136 7P1 136 70 136 7P1 136 Z1 33 52 70 t 36 36 I I 56 71 144 22 41 62 78 145 36 53 68 77 146 55 67 73 80 148 82 93 98 106 156 114 117 125 139 166 114 117 125 139 166 18 28 42 50 82 18 31 45 51 87 19 35 50 56 87 32 45 54 55 88 48 57 58 58 89 72 79 78 - - - - 76 94 100 100 100 100 100 2.00E + 04 4.00E + 04 6.00E + 04 pA 8.00E8.OOE 8.00E8.OOE + 04 l.00El.OOE l.00El.OOE + 05+ OS + 05+ OS 2.00E + 05 4.00E +4.ooze + 4.00E +4.ooze + 05 100% 2.00E + 04 4.00E4.OOE 4.00E4.OOE + 04+ OA + 04+ OA 6.00E5.OOE 6.00E5.OOE + 04 8.0CE8.OCE 8.0CE8.OCE + Ca04 Ca04 X% X% response 1.OOE + 05 2.0CE2.0 2.0CE2.0 + 05+ OS + 05+ OS a.OO£ + 05 The above relationships apply to characterising DNA under statIc conditions. A further novel application of these relationships is to monitor, in real time, the progress of reactions and processes where DNA is enzy- matically modified, for example the polymerase chain reaction ?CR), reverse transcription, DNA-DNA liga- tion, or where DNA interacts with non-enzymatic proteins, for example DNA-protein interactions. There are a number of advantages which this process facili- tates, namely the determination of end-point of the reaction or process, identification of the key stages of the reaction or process, trouble-shooting and auto- mation. For example, PCR is a major molecular biologi- cal tool both in academic research and in medical diagnostics, where PCR is used to distinguish individu- als who carry specific genetic traits. One of the major bottle-necks encountered in medical diagnostics is the analysis of the PCR products. This usually involves a single pure DNA product which needs to be sized accurately. This is normally carried out by laborious gel electrophoresis.

Measurements of conductivity have also been applied to PCR. A fixed frequency a.c. signal (0 and lMHz - O to 10 volts) may be passed between two thin wire platinum electrodes as above and the current/conductivity meas- ured through the PCR reaction mix as the reaction proceeds. Measurements may be taken at fixed points during the reaction, or constantly. Conductivity fluctuations during the PCR follow distinct patterns which allow real-time assessment of the reaction.

Initially there is a drop in conductivity, probably caused by dissociation and denaturation of the genomic template DNA. Successful reactions show a rapid rise in conductivity over the proceeding cycles which reach a plateau which signifies that the end of the reaction has been reached (Figure 4). Unsuccessful reactions show no such increase in conductivity over the proceed- ing cycles. Experimental details and results obtained are described below.

Real-time monitoring of PCR reactions gave similar results. For reactions which gave poor yields the curve shown in Figure 7 was typical. Initially over the first couple of cycles there is a slight increase in overall conductance (a), this is followed by a decrease in conductance (b) then an increase (c) which levels off to a plateau during the final stages of the reaction (d).

This increase in conductivity during the final cycles of the reaction can be attributed to an increase in the yield of the reaction. The decrease in conductivity during the initial cycles may be accounted for by changes in the concentration of small conducting species such as primers, nucleotides and inorganic ions, whose mobilities contribute to a greater extent than large DNA molecules at the high ac frequencies used. As the concentration of product increases it begins to contribute to the overall conductivity of the solution, thus at high DNA concentrations the conduc- tivity of the reaction mixture increases. When the reaction reaches its end-point the concentration of DNA begins to level off and the conductance of the solution reaches a plateau.

Addition of single reaction components at the end point of the reaction should therefore increase the yield of the reaction further and this would be shown as an increase in the conductivity of the solution. Figure 8 shows how the addition of Taq polymerase (e) increases both the conductivity of the reaction and the overall yield of the reaction. This would suggest that Taq could be limiting. Further experiments at establishing limiting components should include addition of primer, nucleotide and buffer.

It will be recognised from the foregoing that there is provided a method for the estimation of a property or parameter of a nucleic acid material, or of a process in which a nucleic acid is modified by a chemical or biochemical reaction, said property or parameter being one to which the electrical conductivity of the nucleic acid material is related, which method comprises measuring the electrical conductivity of the nucleic acid material, or of a reaction mixture containing said material, and estimating from said measurement the property or parameter of the material or process by reference to a predetermined relationship between electrical conductivity and said property or parameter.

The parameter estimated may be the concentration of a simple species of nucleic acid in a solution thereof. Alternatively, the property estimted my be the molecular weight of a nucleic acid, estimated from the predetermined relationship between molecular weight and the characteristic curve of electrical current/frequency response of a part thereof. The relationship may be between molecular weight and the area under the characteristic curve of frequency response. In a further alternative the property estimated is an overall indicator of molecular weight for a range of nucleic acid species.

The measurement may be carried out with apparatus which has been precalibrated in accordance with the predetermined relationship, which may be both the molecular weight/conductivity relationship and the concentrationlconductivity relationship.

If a process parameter is estimated. said parameter may be the extent to which an enzymatic processing of nucleic acid has proceeded, such as a polymerase chain reaction, a reverse transcription, or a nucleic acid ligation.

ApsendlxAtendix ApsendlxAtendix Data a^quis--:onacuistn a^quis--:onacuistn and system desiqn Measurement ^- cm ^- cm PCR conductivity was achieved by passing a pulsed, 2:.Hz 1V alternating current through the reaction mixture (Figure 1.0). This was amplIfied through a variable gain amplifier (1-25) and this signal was lugged lugged via a PC. The frequency of signal pulse and data acquisition were also controlled via the PC. An experimental scheme for measuring PCR conduc- tivity is shown in Fig 9.

Digital to analogue control of the switching circuit, and analogue to digital measurements were controlled using a singe I/O card (PCL-812PG,(PCL-8l2PG, (PCL-812PG,(PCL-8l2PG, Semaphore Systems Limited, London) . Initially acquisition of conductivi- ty data was at fixed time intervals, ut a significant recent modificationmodifIcation modificationmodifIcation (requiring the design of an addi- tional electronic circuit board) has been to regulate acquisition by temperature such that as the temperature reached typical annealing values the rate of acquisi- tion was increased. At high denaturattondenatura*:on denaturattondenatura*:on and extention temperatures acquisition times were Increased (Table 4)4). 4)4).

Table 4: PCL-PlD-GPGL-8l^?G PCL-PlD-GPGL-8l^?G I/O card control Op Fn Parmi Parm2 Parm3 Remark 01 21 1 2000 Set D/A :~large :~large to 2V (Switch probe on) 02 51 1 Wait 1 second 03 27 229 4 Set internal card gain 4 04 01 Read A/D conductivity probe 05 27 229 0 Set initialinitIal initialinitIal card gain 0 06 01 Read A/D conductivity temperature probe 07 21 1 0 Set D/A voltage to OV (Switch probe off) 08 51 1 Wait 1 second 09 51 $0 4 0.69 If temperature s annealing temperature then 10 If not then 13 10 57 1 Wait 1 second 11 52 1 Goto Ol 12 58 End of statement 13 15 wait 15 seconds 14 00 End