GB2180941A - System for determining sizes of biological macromolecules - Google Patents

Abstract

An electrophoresis system for determining the sizes of radiolabelled biological macromolecules comprising a cell 1 containing an electrophoresis gel 7 and having at least one lane, a voltage source connected across the gel for effecting the movement of macromolecules in the lane, a detector 19 fixed relative to the moving molecules for generating electrical pulses responsive to signals emitted by the radiolabelled molecules; a pulse processor 23-31 for counting the pulse rate, and a computational device for comparing the pulse rate to a predetermined value. <IMAGE>

Description

SPECIFICATION System for determining sizes of biological macromolecules Deoxyribonucleic acid (DNA) is the cellular macromolecule responsibleforthetransmission ofgenetic information, and is composed ofthefournuc- leotides, adenosine (A), thymine (T), cytidine (C) and guanosine (G). The history of technology for deter- mining the sequence of DNA began with the 1970 discovery of restriction enzymes which cut doublestranded DNA at specific sequences. (1,2).

DNA-sequence determination requires a method for defining nucleotide pairs at each position along the double helix. For experimental purposes, since each nucleotide pairs with its Watson-Crick comple- ment in the oppposite duplex strand, the analytical problem requires the definition ofthe sequence of nucleotides A,T,G, and C along one (or both) ofthe strands ofthe double helix.Asecond breakthrough was the development of the "ladder" methodforthe determination of DNA sequence using either en zymatic(3) or chemical (4) methods.

Both major rapid DNA sequence determination methods rely on thin denaturing polyacrylamide slab gels to fractionate single-stranded DNA reaction products.

(5) These techniques share many features as follows: 1. They are based on the creation of collections of radiolabeled DNA fragments randomly extending from one end fixed at a specific location to one ofthe four possible bases in the DNA chain.

2.Asimilarcollection offragmentsiscreatedes- sentiallyforeach ofthefourbases as a chainterminus.

3. The fragments are size fractionated by highresolution gel electrophoresis, using a high voltage field as the driving force.

4. After resolution, the gel is exposed to x-ray film for several hours, and then the DNA sequence can be read directlyfrom the autoradiograph of these fragments. (6, 7, 8) Gel electrophoresis is one ofthe most widely used analytical techniques in molecular biology and genetics. This technique physically separates molecules according to the different speeds at which they are forced to travel down a gel underthe influence of an electric field. Typically, the separation requires sev- eral hours and the gel is then removed and analyzed forthe location of different molecular groups on the surface ofthe gel. Although many markers can be used to highlight the molecular groups, radioisotope labels and fluorescent labels arethetwo most common tracers.

Gel electrophoresis in polyacrylamide or agarose gels is used particularly for separation of DNA (De oxyribonucleic Acid), RNA (Ribonucleic Acid) and proteins in one (9-14) and two (15,16) dimensional formats. Several types of radiolabels are used, but the most common are phosphorous 32 (P32) and sulfur 35 (S35) which both emit beta rays (11,17).

Upon the application of a bias voltage of 5 to 45 volts/cm across the length of the gel, the molecules are propelled down the gel by the effect of the electrical field on their inonic charge. Theirvelocity isan inverse function of molecular weight, often stated as: R = k/(log Mr) where R = Rate of travel at standard electric field; k = retardation constant related to the gel polymer concentration; Mr=molecularweight(11,18,19,20) Thus over a period oftime the molecules separate according to theirweight, with smallerfragments reaching the bottom of the gel more quickly.

While proteins conform well to this logarithmic function, it is notwidelyunderstoodthat DNAand RNA exhibitdifferent behavior in thatthetimefor migrationoverafairlywide range of molecular weights is observed to be inversely proportional to molecularweight. This direct relation of migration speed to molecularweight (not log Mr) was noted first by Southern. (21,22) In the size range of interest in DNA or RNAsequencing, smallerfragments less than 50 b.p. move slowerthan this relationship and will exhibit band diffusion at the bottom ofthe gel. In the key decade 50 to 500 b.p., migration rate is closely proportional to the inverse of molecularweight.

Each channel in the sequencing gel contains a band that corresponds to a particular base located at a particular position in the DNA sequence. In reality, the position of the band reflects the distance between some fixed point in the DNA sequence (the 5' or3' end ofa restriction fragment) and a particular base in that sequence. Since the resolving power of the gel ena bles frag ments that differ in length by one nucleotideto be clearly resolved from one another, it is possible to read the sequence of the next longest fragment. The development of very thin (0.1-0.4mm) urea-containing polyacrylamide gels has improved spatial resolution in autoradiogramsofsuch high resolution separations.

To determine a sufficient length of sequence from an experiment, it is usual to load aliquotsfrom a reaction onto two or more gels run forvarioustimes. Gel strength is also varied, since lower acrylamide contentwill enable largerfragmentsto migrate faster down the gel and be spatically resolved in less time.

Itwould be advantageous to be able to obtain more sequence data from each gel becausethiswould allow either more total sequenceto be obtained, or fewer gels to be run per reaction. One of the main limits on the length of sequence obtainable from an autoradiograph is the progressively poorerseparation of the bands corresponding to longer DNA molecules. On the same gel,the spacing between the shorterDNAfragments iswiderthan is required for correct interpretation of the sequence pattern. A buffer gradient method has been developed which pro gressively lowers the electrical resistance per unit length in the lower gel.Sincethe current is constant, throughout the gel, by Ohm's lawthevoltage drop per unit length also decreases toward the anode (bottom). Since the voltage potential difference across the gel drives the polynucleotide migration, the spacing between smallerfragments is reduced. By running the gel for a longer time, or by increasing the length ofthe gel, spatial resolution of the longerfragments is increased. (23, 24) The partial sequences are read from the family of gels run for the selected DNA fragment, recorded and compared in order to reconstruct the entire sequence. Two sources oftrivial error are associated with this stage: one involves careless reading ofthe gel (for example, mistaking channels or skipping a band); the other involves errors introduced when manually recording data.Errors due to manually recording the x-ray film exposed to the gel have been reduced by automatically transferring data directly from the original autoradiograph into a computer. A digitizing tablet allows the sequenceto be read dir- ectly into computer memory. This tablet operates by sending the location of any point on the surface of the pad once this point has been touched by the signal pen. The location is represented in the form of a digitized set of X and Ycoordinates. By repeating the reading process, the readings can be compared and discrepancies highlighted by an interactive computer, thus allowing a data entry check ofthe appropriate region of the gel.

After reading primary gel data, the next step is to order these short segments to find overlapping stret chesofcommon nucleotidesandfinallyjointhem together into one long block of sequence that repre sents the primary structure of the original molecule.

Various computer programs have been written to aid in this process. These programs provide three essential functions. The first is to direct each piece of sequence data into a master archive which can be used for further reference. The second function allows the identification of those blocks of sequence thatcontain homologousorcomplementarystret- ches -the overlaps. The third function is that of melding anytwo strings of nucleotides that contain overlapping sequences. If discrepancies occur at a point, diacritical marks are placed above the nucleotide concerned. This serves to identify those positions in a sequence where either new data are needed or some reevaluation ofthe old data is required.As the sequencing project continues, the original stretches of short sequences gradually become melded into longer and longer blocks until the reconstruction of the original macromolecule is accomplished.

From the foregoing, it can be seen that both the efficiency and technical effectiveness of DNA sequencing could be improved by the ability to read longer gels. Besides the use of ultra-thin gels and the grad- ientbuffermethod, use of S35 asthe radiolabel increases the sharpness ofthe autoradiograph bands, but with a penalty of longer exposure time. Although both S35 and P32 are beta particle emitters, the peak strength of S35 is only one tenth that of P32.When the gel is exposed to x-ray film overnight by direct contact, much Iessofthefilm is exposed byangular emission of beta particles from the gel with S35 than with the stronger P32, which causes a dimensional expansion oftheband image on thex-rayfilm. Elimination ofthis diffusion pattern by using S35 reduces the image size ofthe band and increases the spatial resolution of adjacent bands, thus increasing the length of sequence that can be read. (9) All ofthese methods require a delay in reading data, a manual interface in reading data, and relatively short DNA fragments for analysis compared to the DNA sequ ence ofinterestforatypical researcher.

The band pattern can be read by noting their relative positions or with the aid of a digitizing system available from several companies (e.g. DNASTAR, Madison, Wi; IntelliGenetics, Palo Alto, Ca). Tykva has previously experimented with an off-line method of scanning gels using a silicon surface barrier detector. (25, 26, 27) These detectors were not suitablefor phosphorous 32 and were not used in a system with any on-line capabilities. In a latter review including Tykva's work, the potential of using semiconductor detectors was included but without possible methodologyfor use in an on-line systems. (28) An object of the present invention is to provide an improved apparatus, and improved process, for det- ermining the nucleotide sequence in nucleic acid molecules.

Another object ofthe invention is to provide an improved apparatus for, and an improved method of, gel electrophoresisforsequencing DNA, RNAand protein molecules.

Another object isto provide an improved apparatus for, and an improved method of, gel electrophoresisforsequencing nucleic acid moleculesata fast rate while yielding precisely defined results.

It is an additional object of the invention to provide apparatusanda methodforsequencinglonger sequences of DNA and RNAwith reduced effort by the user.

Still another objective is to provide an improved gel electrophoresis system and method wherein radiation quanta emitted by radio-labels on nucleic acid molecules are read and converted to electrical signalsforfurtherconversion into an intelligible output.

An additional object ofthe invention isto provide apparatus of the foregoing types which is efficient and effective in use, economical to manufacture and operate, and capable of rugged construction and reliable performance. Other objects will be apparent from the description to follow and from the appended claims.

The foregoing objects are achieved according to the preferred embodiment ofthe invention by means of a system comprising a tubular electrophoresis gel cell, a detectorforthe on-line measurement of radioactive beta emissions of radio-labelled molecular fragmentswhile providingfortemperatureuni- formity across multiple gel lanes, and electronic and computer modules for processing the detector signals and generating an intelligible output which does not require the manual interpretation of gel band positions.

Figure lisa schematic diagram of a system according to the invention showing a slab gel, and a detector and associated electronic circuitry in general terms.

Figures2and 3show in schematic form alternate apparatus according to the system of Figure 1.

Figure 4 is a cross-sectional side view of an electrophoresis cell according to an aspect of the invention, and Figure 5 is a partial front view ofthe cell shown in Figure 4.

Figure 6is a perspective view of a tubular gel and an annular detector according to the invention.

Figures 7and 8are perspective views of two dif ferenttypes of annular detectors as shown in Figure 5.

Figure 9is a schematic diagram of detection and read-out circuitry incorporated in an embodiment of the invention.

As an alternative to the standard autoradiog ra ph ic technique in which the relative positions of the macromolecules are determined by exposing an x-ray film in contact with a gel,the invention in its preferred forms determines relative nucleotide positions by detecting the macromolecules as they pass a fixed nuclear detector. The detection is carried out by rad iationdetectorswhich respond to beta orgamma radiation emitted by radio-labelled fragments. The detectorsare placed closetothe gel with theirfield of view restricted so that they only detect radiation coming from a narrow horizontal section ofthegel.

The outputs of the detectors are coupled to electronic signal conditioning and data acquisition equipment.

Referring to Figure 1, a system according to the invention is shown in general terms. This system includes an electrophoresis cell 1. Cell 1 includes a pair of vertical, preferably glass plates 3 and 5 between which is disposed a polyacrylamide or agarose gel 7.

The plates and gel extend between an upper buffer reservoir9 and a lower buffer reservoir 11. A power supply 13 is connected to a negative electrode 15 in upper reservoir9 and to a positive electrode 17 in lower reservoir 11. A radiation detector 19 is positioned adjacent plate 5, and it is electrically connected at its output to a preamplifier 21. The output of a detector bias supply 23 is also connected to preamp 21. The output ofthe latter device is connected to a single channel analyzer 25, which is in turn connected to a counter 27. A preset timer 20 is connected at its output to both counter 27 and to a multi-channel storage 31, to which the output of counter 27 is also connected. The output ports of storage 31 are con nected respectively to a printer33 and to a display monitor35.

Although the speed ofmigrationofbiomolecules in gel electrophoresis is affected by voltage and tem perature, the separation between segments that vary in size by one nucleotide (n vs. ne1) depends only on molecular weight. Our laboratory data confirms Southern's observation that the migration rate of segments is inversely directly proportional to molec ularweight.

An embodiment of the system of Figure 1 is shown in Figure 2. This system includes an electrophoresis cell 100 having parallel, vertical glass plates 103 and 105 which are dimensioned to contain between them a standard 10 cm. long gel 107 made with 10% acry lamide. Cell 100 has a negatively charged upper buf fer reservoir 109 and a positively charged lower buf fer reservoir 111. scintillator 119, which more specifically is a sodium iodide crystal, is disposed nearthe base of gel 107. Scintillator 119 has an entr ancewindow 120 cut into a lead sheet 122 defining a horizontal slit and functioning as a collimator. A photomultipliertube ("PMT") 124 detects that light output ofscintillator 119, and it is energized by a high voltage supply 123.The output of PMT 124 is connec ted to a preamplifier 121. Preamplifier 121 is connected to a single channel analyzer 125 of a multichannel analyzer ("MCA") 126. MCA 126 also includes a counter 127 having input ports connected re spectivelytotheoutputofsinglechannel analyzer 125 and to a preset timer 129. The counter output is connected to a monitorl35whichdisplaysaccumu- lated counts.

The system of Figure 2was used in run Aasfol- lows. Markers of pBR322 were cut with restriction enzyme Mspl, labelled with radioactive iodine (1-125, and loaded onto gel 107 in a conventional fashion. A voltage of 1000 volts was applied between buffer tanks 109 and 111.Gamma rays emitted bythe radioactive iodine passed through the glass plate, butonly those level with slit 120 in lead shield 122 passed to the scintillation detector and gave rise to output pulses from PMT 124 of detector 119. Pulses from preamplifier 121 were transmitted to single channel analyzer 125where low level noisepulseswereex- cluded.The absence of a band in the collimator window appeared as a low count or background accumulation in monitor 135, whereasthe presence of a band appeared as a high count accumulation. The total counts were registered by the MCA at 30 second intervals. At the ends ofthese intervals, the cumulative counts registered by MCA 126 were read and noted by hand, and the counterwas reset to zero.

Peaks corresponding to the expected 26 markers werefound when the count readingswere plotted out by hand.

Another specific embodiment ofthe system illustrated in Figure 1 is shown in Figure 3. This system includes an electrophoresis cell identified by the number 200 which comprises three parallel, vertical, glass plates 203,204 and 205. Plates 203 and 204are separated buy a horizontal gap of 3 mm, and are lined with a 10 mil polyester film 206 so that beta rays emitted by P-32 isotope labels may passthrough the horizontal gap between plates 203 and 204. Upper and lower buffer reservoirs 209 and 211 which are charged as described previously are in communication with gel 207.Aleadcollimator222likethatoftheem bodimentshown in Figure 2 is provided in conjunction with a Cadmium Telluride detector 219.The latter device can be an experimental model made by Radiation Monitoring Devices, Inc. Detector 219 is connected to a preamplifier 221 through which it is also connected to a detector bias supply 223. An appropriate bias supply is a Radiation Monitoring Device Model PSP-1. A multi-channel analyzer ("MCA") 226 is connected to preamp 221. This unit functions as a multi-channel pulse height analyzer, and includes an ADC input 227 connected to preamp 221. A Can berra Model 40 has been found to be an effective MCA. The input of a multi-channel storage 231 is connected to ADC 226, and the output of device 231 is connected to a display monitor 235 which displays pulse height distribution.

Attest was run on the system of Figure 3 to prove that nucleotide fragments labelled with different isotopes, in this case P-32and 1-125, could be distinguished. As each group or band of mac romolecules passed the gap and collimator, some of the radiation passed through and hitthedetector.

The beta radiation from P-32 consists of electrons with most oftheir energies distributed about 600 KeV (maximum 1700 KeV), whereas the 1-125 gamma ray energy is distributed sharply near 30 KeV. A beta or gamma rayofgiven energywhich impinges on the sensitive part ofthe detector may, depending on the efficiency ofthe detectorforthat energy, give an output pulse from the detector. In this case Cadmium Telluride detector 219 gave different characteristic puise heights forthe gamma ray emitted by the iod- ine and the beta rays emitted by the phosphorus.Use of MCA 226 permitted the operatorto distinguish between pulses originating from P-32 vs. 1-125. As in run A on the system of Figure 2, in this run Bathe sample was loaded atthetop ofthe cell, the voltage for electrophoresis applied and every 30 seconds the number of counts registered by the MCA in the preceding 30 seconds read and noted. The sample consisted of 14 oligonucleotides from 1 to 14 base pairs in length with 1 and 14 being base pairs in length with 1 and 14 being labelled with 5P-32, the rest with 1-125.

Segments 1 and 14were detected and identified as being labelled with P-32; ofthe rest 2,3,4,5,8,9,10, 11, 12, and 13 were clearly seen as identified as being labelled with 1-125. These oligonucleotides resulted from an actual DNA sequencing "A+G" Maxam Gilbertchemical degradation.

Referring next to Figu res 4 and 5, a preferred form of electrophoresis cell 300 and detector319 is illustrated. This system was built primarily to improve sensitivity and resolution so that a sample of mu It- iple fragments, labelled only with P-32, extending up to and beyond 1000 base pairs could be detected. Cell 300 includes vertical, parallel, opposing glass plates 303,304 and 305. Plates 303 and 304 have a sheet or film 306 of 10 mil polyester bound on theirinterior face and a gel 307 is disposed between the polyester film 306 and plate 305. Film 306 provides sidewall supportforgel 307 and permits beta radiation to pass through to the detector (discussed below) with insignificant radiation absorption of lessthan 20%.Plates 303 and 304 are spaced apart, this space being sufficient so that plates 303 and 304 can receive between them a detector crystal 310 of detector 319. An acrylic collimator322 is mounted between detector319and sheet 306 to improve the resolution ofthe detector.

Turning to Figure 5, a pair of spacers 308 are shown which comprise plastic strips which are preferably 0.5 mm thick and 2.5cm wide at either side. Another set of spacer strips 312 having the same thickness as that of spacers 308 are placed between each lane of a set of lanes 314,316,318 and 320 of cell 300, on either side of each of a set of wells 322. Spacer strips 312 extend from the top to the bottom of gel 307. When cell 300 is assembled, pressure is applied to glass plates 303,304 and 305 to pinch the spacers between plate 305 and film 306. Spacers 308 and 312 function both to maintain a precise and constant gel thickness, and to ensure thatthe track of macro- molecules passes directly in front of detector 319.

Preamplifier321 amplifies pulses from detector 319 sufficiently so that they can be counted for fixed periods oftime in standard electronic counters. Such counterscan,forexampie, bescalers in a multichannel analyzer (as shown) orthe counters in an Ortec quad timer/counter. Both reservoirs are large relative to the amountofgel 307 to enable long runs while maintaining gel and buffer integrity. The electrolytes are charged respectively by a negative electrode 315 and by a positive electrode 317 which are surrounded by the electrolyte in which they are disposed. Detector 319 is connected to its preamplifier by a coaxial cable.Detector319 is specifically a Model A101/P4cadmium telluride unit manufactured by Radiation Monitoring Devices, and its entrance window is 2mm high and 1 Omm wide.

Two types of radiation detectors have been found to be quite effective, namely scintillation detectors and solid state detectors. Scintillation detectors such as detector 119 in Figure 2 convert incident radiation to light which is collected with a photomultipliertube designed to limit noise. Three types of scintillation detectors are sodium iodide crystal modified with thallium, calcium fluoride crystal modified with europium, and plastic. The Nal type is made by Har shawChemicalCompanyand by Bicron Inc. Itis highly sensitive to both isotopes (1-125 and P-32) and operates without degradation at 550C. However, the Nal type is hygroscopic and therefore must be sealed. Hence, there is going to be some attenuation in the housing window of the lower energy betas.

The Nal type measures both beta and gamma rays.

The CaF type is also made by Harshaw Chemical Company. It is less sensitive than sodium iodide, but is sensitive to all of the beta and gamma rays of interest. The CaFtype ofscintillation detector operates without degradation at 55"C and is non-hygroscopic.

It therefore achieves a high degree of collection efficiency even with low energy beta and gamma rays when placed close to the gel since it does not require moisture sealing.

Athirdtypeofscintillation detector is the plastic type. This type is usually anthracene mixed in a poly vinyltoluene base. It is extremely rugged and, ifthick enough, has a high collection efficiency for beta rays.

Onetypeoftheforegoing detector is manufactured by Nuclear Enterprises Ltd. of the United Kingdom.

Solid state detectors such as detector 219 in Figure 3 convert incident radiation to electrons and act as a semi-conductorto collect the electrons. Two types of solid state detectors are the cadmium telluride crystal (CdTe) type and the silicon surface barrier type. The CdTe type is manufactured by Radiation Monitoring Devices and by Dositec Inc. The CdTe type has excellent theoretical sensitivity, especially for low energy beta radiation. It is suitable for detecting low energy gamma radiation, such asthatemit- ted by 1-125.

The othertype of solid state detector is the silicon surface barriertype. Forthistype of detector, a layer of gold is placed on a silicon or germanium base, and forms an active region within the semi-conductor that responds linearly to the energy of an incident particle. These detectors are favored because oftheir compact size as well as because oftheirfast, linear response. This type of detector comprises a disc of high purity silicon mounted in a housing with a thin layerofgold deposited on top ofthe silicon. In operation, reverse bias polarity is applied through the output connection to the back contact to produce a region of high electric field, the depletion region, in the silicon underthe gold layer.The penetration depth of incident beta rays is typically only a few mic rometers and hence they lose all their energy within the depletion region. Because the gold layer is thin, the dominant energy loss is within the silicon and the number of charged carriers produced is proportional to the energy of the alpha particles. The electrons and holes created within the silicon depletion layer are swept apart bythe electricfield withinthe depletion layer. The motion ofthe carriers produce a current pulse which appears as a fast rising voltage pulse acrosy the output load resistor.The height ofthevol- tage pulse is directly proportional to the energy of the incident particle, enabling pulse height analyzers to process and display the energy spectrum ofthe particles as described earlier. This type of detector cannot measure gamma rays.

Theoutputofthe multi-channel analyzer(MCA) 226 can be connected to a Centronics printer. The MCA has been run in the multi-scaling mode in which the input pulses were first routed through a window discriminator, counted for a fixed period oftime and then stored in successive data channels. By this means a time histogram ofthe count rate from the detector was stored for subsequent analysis. This time histogram was displayed on the MCA screen allowing the operatorto monitorthe progress ofthe experiment.

In a proving run Con the system of Figure 3, standard slab gel electrophoresis procedures as described in the literature were followed. The gel was 4.5% polyacrylamide. The sample to be analyzed was a mixture of Hinc II and Hae Ill digests of phi-X174RF DNA labelled with 32P-dATP. This digest contained 24 DNA fragments ranging from 72 to 1353 base pairs. Three microliters of a solution of the above fragments in formamidewere loaded intothewell, 600 volts were applied across the gel and data accumulation commenced in the MCA. The window discriminatorinthe MCAwassettoexcludethe majority of low amplitude noise pulses without sign ificantly reducing the desired signal count rate. The presetcountertimewas setto 30 seconds.Asthe fragments passed the detector the accumulated count in each time period increased giving peaks in the time histogram.

At the end of run Cthe data stored in the MCAwas printed out on the Centronics printerwith one dot per channel representingthecountaccumulationforthat channel.A line connecting each dot was drawn by hand, revealing a series of 12 well-defined peaks.

The foregoing embodiments can be modified in different respects. Thus, considering electrophoresis cells, 100,200 and 300, and the detectors cooperating with them, the two glass plates 303 and 304 on the detector side may be moved closer together, so that they are within a millimeter or so of touching, and the detector moved back, so that the plates themselves form the collimator. Also the glass plates can be replaced with 1/8 inch acrylic plates in which the radiation exitwindow/collimator is machined directly to make a horizontal slot, 1 mm high by 1 Omm wide, and machined to a depth of approximately 0.1 mm from thefarsurface (orfrom the gel). This leaves an acrylic window thin enough to permit penetration ofthe majority of the beta rays emitted by P-32 labelled samples.The acrylic plate can be bonded to two pieces of 1/8 inch thich aluminum, one above and one below the detectors, to give both rigidity and heat conduction for uniform temperature across the gel. This configuration eliminates the polyesterfilm and eases the assembly of the cartridge; however it has proven difficult to keep such plate gels from distorting upon prolonged use in an aqueous environment.

Afurthermodification oftheforegoing embodiments is depicted in Figure 6. The gel for each electrophoresis lane is drawn into a tube 407 ortubes connected between an upper electrolyte reservoir 409 and a lower electrolyte reservoir41 1, and to a detector419 which may be an annular detector as discussed below. Tube 407 should have a small bore (1 to 3 mm) and a thin wall (0.2 to 0.3 mm), and be made ofglass or plastic. The tubes can be contained within a single largertube 408 having a heatexchange medium which can be circulated by a pump 410 to assure isothermal gel operation.Thus, when the apparatus of Figure 6 is used with a cell having four lanes of linked analysis as shown in Figure 5, the fourtubes can be contained in a single larger column which also holds a circulating liquid with mass 10100 times greaterthan the total mass ofthe gel.The advantages ofthis geometry include: (a) long gel lengths are easily obtainable which permit greater separation of bands; (b) multipletubes, oneforeach lane, can be run together in a closely controlled temperature environ ment,thus ensuring closely matched characteristics between one lane and another; (c) lane wandering and band skewing are eliminated; (d)the inherent strength of a tubular section per mitsextremelythin glasswall tubing to be used, maximizing transmission of beta rays; (e)the annular detector configuration can improve detection efficiency by 2 to 3 times;; (f) lower current consumption is obtainable, by a factor of up to five times, for a given voltage gradient becauseofthesmallercross-sectional area; and (g) gels may be pre-drawn and the tubes sealed at either end which permits long term storage ofthe gel.

A detail of an annulardetector419 is shown in Figure 7. A silicon surface barrier detector assembly 519 is illustrated which comprises a gel tube 507, as discussed immediately above, and an array of four, equiangularlydisposed silicon surface barrierdet- ectors 520 surround tube 507 for receiving radiation energy from the radioisotope labelled fragments of macromolecules in tube 507.

Atypeofscintillation detector 419 is shown in Figure 8. Detector 619 in Figure 8 is an annular fluoride scintillation detector in which a plastic gel tube (ND 120) 607 passes through a plastic scintillator 620 attached to a photomultipliertube 622. Radiation from the radioisotope-labelled macromolecules in tube 607 effect the generation of light by scintillator620, and that light is sensed by PMT 622 which generates electrical signals in responsethereto.

Figure 9 illustrates another embodiment of detector and readout circuitry associated with electrophoresis cells as discussed above. Accordingly, a silicon surface barrier detector 719 which cooperates with an electrophoresis cell is connected to a pream plifier72i A detector bias supply 723 is connected to another input of preamp 721. The output of preamp 721 is connected to an amplifier/singlechannel amplifier (SCA) 722, whose output is in turn connected to a counter 727. Preamplifier 721, SCA 722, det- ector bias supply 723 and counter727 collectively comprise a pulse processing means. A bus 728 con nectsthecounterto a laboratorycomputer("SCA") 736.

An experimental version of the electrophoresis system of Figure 6 and the circuitry of Figure 9 was builtusing a glass capillarytubefrom Wilmad Glass Co. as tube 407. This tube was 23 cm long with an internal diameter of 1.5 mm and wall thickness of 0.15 mm. In a run D,thetube was filled with 6% de- naturing gel by holding it vertically and injecting the gel from a syringe connected to the bottom ofthe tube. The top of the acrylamide was overlayered with isoamyl alcohol to produce a flat gel surface. After allowing two hours for complete polymerization of the gel the syringe was removed. A collimator com prisingtwo brass tubes, each 3 mm O.D., 1.5 mm was used.This brass collimatorwas slipped overthe glass tube such that the brass tubes were 0.5 mm apartas awindow, andthe glasstube connectedto the top and bottom reservoirs. Two surface barrier detectors were mounted close to the collimator window, one on either side, so that their sensitive surfaces were directly facing the gap between the brass tubes. The reservoirs were filled with electro lyte and a voltage of 700 volts applied for two hours.

An aliquot of DNA fragments was cut from 4N (a derivative of pBR 322) with Hinfi. It was labelled in standard mannerwith a radioactivetriphosphate having a specific activity of 400 curies/millimole, taken up in formamide and fully denatured by heat ing at 65 deg Cforten minutes. After gentlywashing the surface ofthe gel with buffer by means of a syringe pipette, the sample was loaded onto the gel surface. A voltage of approximately 750 volts was applied from a constant current supply to give a cur rent of 0.4 ma.

The pulse processing means (members 721,722, 723 and 727 in Figure 9) received and counted pulses over succesive ten second periods. Atthe end of each ten second period, the countertotal was transferred to individual memory locations of the HP86 micro computer. The computational means used the data stored in memoryto provide a hard copy histogram of count rate vs. time. Peaks were observed on the histogram atthose points in time when radiolabelled macromolecules passed the detector.

After running for ten hours, radiation peaks cor responding to fragments of length 72,75,83,110, 154,220,221,296,310, 346,398,506,517 bases were clearly identified, with only the doublet at 200 and 210 bases being unresolved as a single peak. In run D radiation detectors 719 were a pair of Ortec surface barrier models # BR-XX-3X25-300-S, preamplifier 721 an Ortec model 1421H, amplifier/single channel analyzer 722 an Ortec model 590A, detector bias supply 723 an Ortec model 428 (set to 150 V), and counter 727 an Ortec model 874.

Asimilarexperimentwas made usingthesame equipment setup as in run D, except that the tube was 25 cm long and the detectors were mounted 21.5 cm below the gel surface. This run E was used to prove thatthe system could be used to identify the components of a dideoxy reaction mix as normally used in oneofthefourlane of DNAsequencing. In run E the sample used was produced by a dideoxy G reaction on a piece of unknown DNA. The radioactive triphosphate used in this reaction has a specific activity of 400 curies/millimole. Gel preparation and prerunning was the same as in run D. A 1 ul aliquotofthe sample was fully denatured by heating at 70 deg Cfor ten minutes and immediately loaded onto the gel.

The voltage was set to approximately 750 volts for a current of 0.4 ma. Pulse processing and computational means were the same as in run D,and after six hours running the printer output showed several peaks. In the range of 40 to 150 bases 23 peaks were clearly identified. In orderto confirm these peaks, a second gel was prepared in exactlythe same manner as prerun. A larger sample, 2.5 ui,was denatured by warming and then loaded onto the gel as run F. The gel was run at approximately 800 volts at a constant current of 0.4 ma. Radiation peaks from run F confirmed all peaks from rum E and, because of greater signal strength, revealed five more making a total of 28 peaks. This isthe number of peaks one would expect to find on average for a single reaction over the 110 base rangefrom 40 to 150 bases.Measurement of dideoxyA, T, and C would run in separate lanes in the same manner. Run E confirmsthe basic capability ofthe system to generate the information necessary to determine a DNA/RNA sequence.

The data processing unit used in the systems of Figures 1,2 and 3 is a multi-channel analyzer(MCA).

An MCA isthe functional equivalent of many (upto 1,000) single channel analyzers in which the energy discrimination range ofthe analyzers are set in a continuum of steps from the lowest energy of interestto the highest energy of interest. That is, each analyzer has its unique energy window. Its output pulses are counted and the total displayed on the MCAvideo display as a histogram versus energy. This plot shows the energy distribution of incoming pulses over a period oftime. In practice pulse heightanalysis is done by a high speed analog-to-digital converterwhich, on determining the pulse/height (energy) adds a count into the appropriate memory address of the MCA. In the systems of Figures 2 and 3, MCAs 126 and 226 are used in thisfashion.

The MCA has a multi-scaling mode in which the energy resolution is performed by a conventional single channel analyzer and the counts are routed into successive memory addresses at given time increments. This process is also displayed on the video screen as a time histogram of the pulse rate.

Although the MCA is a convenient tool for making these arrangements, it is unwieldy and notflexible. A better and more economical arrangement utilizes a single channel analyzer(ortwofordual isotope measurements) connected to a counter interfaced with a microcomputer. The system of Figure 9 is of this type where the microcomputerwas a Hewlett Packard HP-86. The software for the system of Figure 9 provides the following functions: (a) a realtime video plot of signal intensity vs. time, (b) a hardcopy printout of same, (c) facilities for averaging out statistical variations in the signal.

The unit's computer needs are preferably fully compatible with other microcomputers such as the IBM model PC/XT or compatible unit. Appropriate software to handle DNA/RNA sequencing, protein analysis, restriction sites and mapping, and comparisons to private data or genetic databases is avaiiable.

The invention has been described in detail with particular emphasis on the preferred embodiments thereof, but it should be understood that variations and modifications within the spirit and scope of the invention may occurto those skilled intheartto which the invention pertains.

REFERENCES 1. H.O. Smith and K.W. Wilson; J. Mol. Biol. 51:379391(1970).

2.T.J. KellyJr. and H.O. Smith; J. Mol. Biol. 51:379- 391(1970).

3. Sanger, F., Nicklen, S. & Coulson, A.R.; Proc. Nat.

Acad. Sci. USAVol. 74, pp.5463-5467(1977).

4. Maxam, A.M. & ilbert,A.R.; Proc. Nat. Acad. Sci.

USA Vol.74,560-564(1977).

5. Sanger, F. & Coulson, A.R.; FEBS Lett. Vol.87, pp.

107-110(1978).

6. Maxam, A.M. & Gilbert, W. in Methods of Enzymology (Grossman, L & Moldave, K. eds.), Vol.65, pp. 499-560, Academic Press, New York (1980).

7. Smith,A.J.H. ibid, pp.560-580.

8. Deininger, P.L.; Approaches to Rapid DNA Sequence Analysis, Analytical Biochemistry Vol. 135, pp.247-263(1983).

9. Rickwood, D. and Hames, B.D. eds. (1982) Gel Electrophoresis ofNucleicAcids, IRL Press Ltd., Oxford.

10. Maniatis,J.,etal. (1975) Biochemistry 14:3787.

11. McDonnell, N.W., etal. (1977)J. Molecular Biology 110:119.

12. Spohr, G., et al. (1976) European J. Biochemistry 62:313.

13. McMaster, G. and Carmichael, G. (1977) Proc. Nat.

Acad. Sciences 74:4835.

14. Laemmli, U.K. (1970) Nature277:680.

15. O'Farrell, P.H. (1975)J. Biol. Chem.250:4007.

16. Derynck, R. and Fiers,W. (1977)J. Molecular Biology 110:387.

17. Studier, F.W. (1973)J. Molecular Biology 79:237.

18. Johnson, P.H.,etal. (1980) Analytical Biochemistry 102:159.

19. Moore, D.H., et al. (1977) NucleicAcids Res.

4:1273.

20. Biggin, M.D., Gibson, T.J. & Hong, G.F.; Proc. Nat.

Acad. Sci. Viol.80, pp.3963-3965 (July 1983).

21. Southern, E.M. Analytical Biochemistry100:34 (1979).

22. Richwood, ibid A. 70.

23. Biggin, ibid.

24. Yang, R. & u, R. J. Viorology 27:700 (1978).

25. Tykva, R. Advances in Physical and Biological Radiation Detectors (Int'l Atomic Energy Agency, Vienna), p. 271 (1971).

26. Tykva, R. & Panek, V. Radiochem. Radioanalytic Letters 14:109(1973).

27. Tykva, R. & Votruba, I Analytical Biochemistry 50:8.

28. Roberts,T.R. Radiochromatography, Elsevier Scientific, Oxford p.20, (1978).

Claims (26)

1. An electrophoresis system for determining the sizes of radiolabelled biological macromolecules, said system comprising: cell means for containing an electrophoresis gel, said cell means defining lanes; voltage source means connected across said cell meansfor establishing a voltage differential along each lane to effect the movement of the macromolecules; detector means located at a portion of the respective lanes of said cell means and fixed relative to the moving macromolecules for detecting signals emitted by the labelled macromolecules passing by said detector means and for converting the signals to electrical pulses;; pulse processing means electrically connected to said detector means for receiving the electrical pulses from said detector means, and for determining the pulse rate from the respective lanes over predetermined successive periods of time; and computational means for receiving said pulse rates and for calculating the function of pulse rate versus time, and computing the correlation function ofsaid pulse ratefunction with a stored function and for comparing the value of said correlation function against predetermined values.

2. The invention according to claim 1 wherein: said cei I cell meanscontainsthegelinanon-horizontal orientation, said cell means having upper and lower portions; said voltage source means establishes the voltage differential along the vertical dimension of the gel to effect movement ofthe fragments from the upper portion to the lower portion of said cell means; and said detector means are located at the lower portions of the respective lanes.

2. The invention according to claim 1 wherein said macromolecules comprise DNA.

3. The invention according to claim 1 wherein said macromolecules comprise RNA.

4. The invention according to claim 1 wherein said macromolecules comprise proteins.

5. The invention according to claim 1 wherein said macromolecules comprise fragments of radio labelled nucleotides.

6. The invention according to claim 5 wherein said cell means comprise a lane for each nucleotide, said detector means detects signals emitted by nuc leotides in the respective lanes, and said com putational means determines the relative sizes ofthe macromolecules in the respective lanes to sequence the nucleotides.

7. The invention according to claim 1 wherein said system further includes means for updating said predetermined value(s) with on-line computed values.

8. The invention according to claim 1 wherein: said cell means contains the gel in a non-horizontal orientation, said cell means having upper and lower portions; said voltage source means establishes the voltage differential along the vertical dimension ofthe gel to effect movement of the macromolecules from the upper portion to the lower portion of said cell means; and said detector means are located atthe lower port ions of the respective lanes.

9. The invention according to claim 8 wherein said cell means comprises a pair of parallel, oppos ing, vertical plates spaced to define between them a cavityforthe electrophoresis gel.

10. The invention according to claim 1 wherein said cell means comprises a tube defining each of said lanes.

11. The invention according to claim 10and further comprising a jacket surrounding said tubes and containing a heat exchange medium for main taining a uniform temperature of gel within said tubes.

12. The invention according to claim 1 wherein said detector means comprises a solid state detector for converting incident quanta of radiation to elec tron flow, and for functioning as a semiconductorto collect the electrons.

13. The invention according to claim 12 wherein said solid state detector is a cadmium telluride crystal detector.

14. The invention according to claim 12wherein said solid state detector is a silicon surface barrier type detector.

15. The invention according to claim 1 wherein said detector means comprises a scintillation detectorforconverting incident radiation to light, and for generating electrical pulses proportional to the value ofthe light.

16. The invention according to claim 15 wherein said scintillation detector is a sodium iodide crystal type detector.

17. The invention according to claim 15 wherein said scintillation detector is a calcium fluoride crystal detector.

18. The invention according to claim 15 wherein said scintillation detector is a plastic detector.

19. The invention according to claim 1 and further comprising collimator means disposed between said cell means and said detector means for collimating signals emitted by radiolabelled macromolecules passing by said collimator means.

20. The invention according to claim 19 wherein said collimator means comprises a radiation imper viols wall having a window for admitting said signals.

21. The invention according to claim 1 wherein said voltage means comprises reservoirs of op positely charged buffer solutions in communication with said cell means.

22. The invention according to claim 21 wherein said voltage means further includes a voltage supply, and oppositely charged electrodes connected to said voltage supply and disposed in the respective reservoirs.

23. The invention according to claim 1 wherein said computational means includes output means for displaying output signals in intelligible form.

24. The invention according to claim 1 wherein said voltage means comprises oppositely charged buffer reservoirs in communication with different portions of the gel, and means for circulating the buffer reservoirs.

25. The invention according to claim 1 wherein said cell means has a narrow transverse dimension.

26. An electrophoresis system substantially as herein described with reference to and as shown in the accompanying drawings.