Gel Inserts Useful in Electrophoresis

The invention is in the field of electrophoresis. It is of particular interest in terms of its application in genetic engineering and molecular biology.

Additional information pertinent to this invention may be found in Schwartz, D. C. and Cantor, C. R., "Separation of Yeast Chromosome-Sized DNAs by Pulsed Field Gradient Gel Electrophoresis," Cell, Volume 37, pg. 67, May 1984; and Van Der Pleog, H. T., Schwartz, D. C., Cantor, C. R. and Borst, P., "Antigenic Variation in Trypanosoma brucei Analyzed by Electrophoretic Separation of Chromosome-Sized DNA Molecules," Cell, Volume 37, pg. 77, May 1984.

The invention which is based upon the discovery of a new kind of electrophoresis makes it possible, inter alia, to carry out important analyses which were not possible or practical with previously known techniques. Potential applications include the separation of chromosomal DNA, chromosomal mapping, the convenient production of genetic libraries, studies on the effects of various drugs on chromosomal DNA, and the convenient characterization of polymers. The invention makes it possible to separate with a high degree of resolution and at high speeds larger particles (molecules) than those capable of resolution with prior art techniques and to concurrently separate particles which differ substantially in mass. In a preferred embodiment, the invention makes it possible to lyse cells for electrophoretic separation of macromolecules e.g. chromosomes contained within the cells with minimal degradation or breakage.

Electrophoresis in which particles such as a mixture of macromolecules are moved, e.g., through a gel matrix, by an electric field, is a widely used technique for qualitative analysis and for separation, recovery and purification. It is particularly important in the study of proteins, nucleic acids and chromosomes. See, e.g., Cantor, C. R. et al., Biophysical Chemistry, Freeman, 1980, Part 2, pp. 676, 683. Indeed, it is probably the principal tool used in most DNA and chromosomal analysis.

The particles to be analyzed and separated by electrophoresis are placed in a support medium such as a gel and are subjected to an electric potential. Difficulties arise when electrophoretic separation of very large particles is attempted. For example, using previously known techniques, the size of the largest DNA molecule routinely handled is that of a bacteriophage (3.2.times.10.sup.7 daltons). Such a limit on size prevents many kinds of desirable analyses from being carried out. For example, intact chromosomal DNAs are larger and are typically reduced in size in order to make it possible to work with them. This, however, destroys important information encoded within the DNA and precludes many important experiments and analyses.

Methods of extending gel electrophoresis to particles of higher mass by reducing the gel concentrations have been proposed. However, this adversely affects resolution, makes experimental conditions difficult to control and has not been successfully applied to DNA molecules having molecular weight greater than about 5.times.10.sup.8 daltons. Fangman, W. L., Nucleic Acids Research, Vol. 5, No. 3, March 1978, pp. 653-655; Serwer, P., et al., Electrophoresis, 1981, Walter, deGreuyter and Coe, pp. 237-243.

It is believed that resolution in previously known electrophoresis techniques is field-dependent since lower electric field intensities generally give higher resolution. As a consequence, electrophoresis runs in which higher resolution is desired often take as long as 100 hours. Moreover, particle mobility, and hence resolution capability, is believed to vary with the logarithm of the mass of the particles to be separated, which of course is not a highly sensitive basis for obtaining separations. Additionally, in known prior art gel electrophoresis, different gel concentrations are typically used for different mass or molecular weight ranges, thereby limiting the range of particles which can be concurrently resolved. Furthermore, previously known electrophoresis techniques are typically used to separate only small amounts of particles, and the process cannot conveniently be extended to larger amounts.

Another problem involved in the electrophoretic separation of large molecules e.g. DNA arises because the molecules (DNAs) must first be isolated since they may not exist as free molecules in the cell. For cells such as yeast and bacterial cells which have a cell wall isolation of DNA generally involves weakening the cell wall by treating it with an enzyme such as lysozyme for bacteria or zymolyase for yeast to form spheroplasts and with a chelating agent e.g. ethylenediaminetetraacetic acid (EDTA). For cells such as mammalian cells which do not have a well-defined cell wall it is of course not necessary to carry out such a treatment step. Cell lysis of spheroplasts or of cells which do not have a well-defined cell wall may be then accomplished by the addition of a detergent such as sodium dodecyl sulfate (SDS) in a buffered saline solution.

Following lysis, the solution is treated with pancreatic ribonuclease to hydrolyze RNA with protease to degrade proteins. Residual proteins and oligopeptides are extracted with an organic solvent, such as phenol or a mixture of equal volumes of phenol and chloroform. Most of the protein will denature and enter the organic phase or precipitate at the interface of the organic and aqueous phases. The clear, viscous aqueous phase containing the DNA may be removed. With the addition of alcohol, the DNA will precipitate out of the aqueous phase as a white fibrous material and may be spooled out on a glass rod. Precipitation from alcohol serves to concentrate the high molecular weight DNA while removing the small oligonucleotides of DNA and RNA, detergent and the organic solvent used in the removal of proteins. Residual detergent and salts may be removed by dialysis of the resuspended DNA solution against the desired buffer. In some instances, it may be desirable to further purify the DNA by centrifugation on isopycnic cesium chloride gradients or hydroxylapatite chromatography.

DNA molecules are extremely susceptible to breakage from shearing forces. As can be seen from the foregoing description of the conventional method for isolating DNA molecules, excessive amounts of shearing forces are applied to the DNA molecules because of the numerous manipulations involved. This results in considerable breakage of the DNA molecules.

Despite the fact that electrophoresis has been used for some time, and despite the fact that important limitations thereof and the need to overcome them have also been long known, no previous proposals are known which have successfully overcome such limitations.

In one embodiment, this invention is a significant departure from the established principles of electrophoresis and is based on the surprising discovery that electrophoresis through deliberately varied electric fields, rather than through the uniform fields sought in previously known electrophoresis methods, unexpectedly yields highly desirable results. More specifically, the invention is based on the discovery that desirable separation results when particles are subjected to respective electrical fields which move them in overall directions generally transverse to the respective general directions of the fields. Particularly desirable results are achieved in at least those cases examined to date when at least one of the electric fields has a deliberate intensity gradient in a direction transverse to its own. As a specific nonlimiting example, two fields can be used which alternate between respective high and low intensities out of phase with each other and are in directions transverse to each other. For example, one of the fields can be on while the other one is off, etc. Particularly good results are obtained when the on and off times of the fields are related to the mass of the particles to be separated, e.g., when the on and off periods are proportional to the mass of the particles raised to a power of about 1.5.

One of the important advantages of this discovery is that it dramatically extends the mass range of particles which can be electrophoretically separated at high resolution. As a nonlimiting example, the new technique can separate at high resolution particles whose mass is about 1.2.times.10.sup.9 daltons, while the upper limit of previously known methods which provide lower resolution, is believed to be about 0.5.times.10.sup.9 daltons. It is believed that the new technique can also resolve particles larger than 1.2.times.10.sup.9 daltons. Another important advantage is that in the new technique, resolution is much less dependent on electric field intensity; consequently, the new kind of electrophoresis can be run at much higher speed, so long as heat produced can be effectively dissipated. As a result, a typical laboratory run can be carried out in 4 to 8 hours, while corresponding runs using prior art techniques require 12 to 100 hours. Another significant advantage of the new technique is that larger amounts of sample, as compared to the known prior art, can be used, thus giving increased resolution and sensitivity. A further advantage is that the new technique can simultaneously resolve, in the same gel, particles from a wider mass range than is believed possible with prior art techniques. As a nonlimiting example, the new technique can resolve simultaneously, in the same gel, particles ranging in mass from about 10.sup.6 to about 10.sup.9 daltons. With previously known techniques several different gel concentrations would have been required to resolve particles in the narrower mass range from about 10.sup.6 to about 10.sup.8 daltons.

As yet another important aspect of the invention, a technique has been found to minimize handling damage to cell derived macromolecules such as DNAs by lysing cells or spheroplasts, in the case of cells having well-defined cell walls, which have been entrapped in a suitable matrix such as a block of gel which is the same as, or compatible with, the electrophoresis gel, and implanting the entire block in the electrophoresis chamber. Another important aspect of this invention is that the blocks of gel may be formed automatically and may be inserted into the electrophoresis chamber automatically with no significant damage to cell-derived macromolecules.

The advantages of releasing macromolecules such as DNA in situ in a solidified gel insert are considerable. For example, the DNA is rendered very stable and may be stored for weeks or even months at room temperature or for days or even a few weeks at temperatures as high as 50.degree. C. This provides major advantages for using electrophoretic methods such as the method of this invention in applications such as in diagnostic applications since DNA entrapped in solid gel inserts may be conveniently shipped from one location to another, thus permitting widespread sample collection and subsequent shipment of samples to a central location for analysis or for storage in a DNA bank where the samples may be maintained indefinitely at low temperatures without risk of damage to the DNA or cross contamination of samples which can easily occur when dealing with liquid samples. Moreover, such solid samples eliminate the need to accurately measure liquid samples and thus reduce error associated with variations in sample size.

The advantages of the use of the gel inserts of the present invention also extend to the ease with which electrophoretic methods may be automated to take advantage of stable and uniformly shaped, modular samples of routine analysis, thus creating a major opportunity for carrying out disease diagnosis at the molecular level by analysis of chromosomal DNAs. Specifically, the DNA in the inserts may be treated in situ with restriction enzymes to produce large intact fragments which can be used for detailed biochemical and molecular analysis, a result not possible with prior techniques. Moreover, the advantages of this approach extend although perhaps to a lesser extent to other macromolecules such as RNAs.

These and other advantages of the invention, as well as additional inventive features, will become apparent from the detailed description which follows.

FIG. 1 is a perspective, partly cut-away view of an electrophoresis chamber useful in explaining certain principles of the invention.

FIG. 2 is a top plan view of the same chamber.

FIG. 3 is a partly schematic and partly block diagram showing an interconnection of exemplary chamber electrodes.

FIGS. 4-7 illustrate exemplary electric fields acting in the electrophoresis chamber.

FIG. 8 illustrates the movement of particles in the new kind of electrophoresis.

FIG. 9 illustrates a hypothesized distortion and movement of a large DNA molecule through agarose gel under the influence of transverse electric fields acting out of phase.

FIG. 10 illustrates the hypothesized effect of a uniform electric field on a large DNA molecule in agarose gel.

FIG. 11 is similar to FIG. 10 but illustrates the hypothesized effect of an electric field which has a substantial intensity gradient in a direction transverse to the field direction.

FIG. 12 illustrates the circulation of cooled buffer through the electrophoresis chamber.

FIG. 13 illustrates the resolution obtained in an experimental example using the new kind of electrophoresis.

FIG. 14 is a perspective view of a mold used for lysing cells or spheroplasts in situ in gel blocks which are later inserted into matching wells in the electrophoresis gel.

An exemplary laboratory device useful in explaining certain principles of the invention is illustrated in FIG. 1 in a perspective, partly cut-away view, and in FIG. 2 in a top plan view. It comprises an open-top, rectangular electrophoresis chamber 10 made of an electrically insulating material, such as 1/4" plexiglass, with dimensions approximately 4".times.4". It supports on its bottom a layer of a medium 12, such as the agarose gel commonly used in electrophoresis, surrounded by electrodes 14. The electrodes are thin (0.032") platinum wires which extend vertically about 3/4" each and are arranged about 1.5 cm apart as seen in the top plan view of FIG. 2.

As one example, the electrode wires can enter the chamber through respective holes arranged in a horizontal row about 3/4" above the interior bottom of the chamber, with each wire extending down, along a respective interior side wall, to the interior bottom of the chamber. In order to generate the desired electrical fields, electrodes 14 are interconnected as shown in FIG. 3. In particular, a d-c power supply 16 (such as Biorad Model 500) supplies d-c power to relay 18 (such as DPDT, 115 volt a-c relay) which is controlled by a programmable timer 20 (such as a Lindberg Enterprises Chrontrol 4-Channel CT Series) to connect a selected one of its two pairs of outputs to the d-c power from supply 16. One output pair of relay 18 (consisting of a negative and a positive output terminal) is connected to the top and bottom rows of electrodes 14 (as seen in FIG. 3), through a respective diode for each electrode. However, it is only when a switch 22 is closed that all the electrodes of the top row are connected to the negative output terminal of relay 18; when switch 22 is open, only the rightmost electode 14 is so connected. The other pair of relay 18 output terminals is similarly connected to the left and right rows of electrodes 14, using a similar switch 24 for the corresponding purpose. Variable resistors R can be used to vary the relevant voltages, as can the controls of power supply 16. The controls of timer 20 determine when a particular pair of relay 18 terminals is energized and when it is de-energized.

When switch 22 is closed and the relay outputs energizing the top and bottom rows of electrodes 14 are on, e.g., at +200 and -200 volts respectively, a substantially uniform electrical field E is established across the bottom of the electrophoresis chamber, as illustrated schematically in FIG. 4. The short arrows in FIG. 4 are uniform in length, to indicate the substantial uniformity of the field, and the longer arrow indicates the general direction of the field (from positive to negative electrodes).

While in reality the field is not perfectly uniform in intensity through out the gell, because of the physical arrangement of individual, spaced-apart electrodes, and for other reasons, and while the general direction may deviate somewhat from the vertical (as seen in FIG. 4), for the purposes of this specification such fields will be called uniform, and are distinguished from fields which are deliberately made nonuniform, e.g., by means of causing an operatively significant intensity gradient in a direction transverse to the overall field direction.

A field E1 which is nonuniform, in that it has an operatively significant intensity gradient in a direction transverse to the general field direction, is illustrated in FIG. 5, and is obtained, in this example, by opening switch 22 such that only the electrode in the upper right-hand corner of FIG. 5 remains at the +200 V potential, while each of the bottom electrodes is at the -200 V potential. The electric field illustrated in FIG. 5 is somewhat fan-shaped, but still has a general direction, illustrated by the longer arrow, which can be viewed as the vector sum of the individual fields that are due to the respective potential differences between the upper right-hand corner electrode and the individual electrodes of the bottom row. The intensity gradient of interest is in a direction transverse to the general field direction, as shown by arrow G, and is due to the fact the distance between the upper right-hand corner electrode and the electrodes of the bottom row increases (and the intensity per unit volume or unit area of the individual fields hence decreases) as one moves to the left along the bottom row, as is indicated by the decreasing lengths of the shorter arrows.

Similarly, when switch 24 is open and the relay outputs connected to the electrodes at the lower left corner and the electrodes along the right-hand row are energized, a similar field E2 is generated, as illustrated in FIG. 6. The only significant difference between the fields in FIG. 5 and FIG. 6 is that the one in FIG. 6 has a different general direction, which is transverse to that of the field E1 in FIG. 5.

One of the unexpected discoveries which this invention utilizes is that if fields such as E1 and E2 alternate out of phase with each other between respective high and low intensities at frequencies selected on the basis of the mass of the particles (e.g., macromolecules) which are to be separated electrophoretically, the particles move from an initial position, such as 26, in an overall direction D which is transverse to both fields E1 and E2, and for any one particle the velocity of movement depends on its mass (or charge). As a result, particles of different masses (charges) travel different distances from the initial position 26, forming bands such as M1, M2, M3 and M4 in FIG. 8, where lighter particles move further distances from the initial position.

It should be noted that the term "transverse" as used in this specification is not limited to an angle of, or close to, 90.degree., but includes other substantial angles of intersection. When used with respect to the angle between electric fields such as E1 and E2, it is meant to exclude only those angles between electric fields in the prior art which resulted from spurious events or from the inability to achieve in practice the design goal of a uniform and unidirectional combination of fields. When used with respect to the angle between the overall direction of particle movement, the term "transverse" is again meant to exclude only angles which resulted from spurious events or from the inability of prior art devices to have the electrophoretic movement coincide with the desired field direction. The term "operationally significant" intensity gradient means here a gradient which is sufficient to enable the relevant fields to move the relevant particles in the direction transverse to the general field directions, for example, as illustrated in FIG. 7.

Satisfactory results can be obtained in some cases with electric fields which alternate and are transverse to each other as discussed above, but are substantially uniform, as is field E in FIG. 4. However, typically better results are obtained when one of the fields has the requisite intensity gradient in the direction transverse to its general direction. Typically, better results are obtained when both fields have such intensity gradients.

While the mechanism by which the new type of electrophoresis works is not entirely understood, it is believed that the application of alternating fields causes a large particle, such as a coiled DNA molecule, to squeeze into the agarose matrix by orienting itself first along the general direction of one of the fields, then along the general direction of the other, etc. Moreover, it is believed that using gradient fields (such as E1 and E2) rather than uniform fields (such as E) produces a shearing effect that helps stretch the molecule in the desired direction. FIG. 9 illustrates this hypothesis by showing a randomly coiled DNA molecule which is pushed into an agarose gel matrix by a uniform electric field E' and is squeezed into the gel by being formed into an elongated cylindrical shape (snake). This snake is then subjected to a uniform electric field E" and is gradually distorted away from its initial snake shape unit it forms a new snake, this time oriented along the general direction of field E", etc., so that its overall direction of movement is along the approximate vector sum of the directions of fields E' and E" . This initial hypothesis has been modified, however, by a later belief that long chain macromolecules such as DNA probably do not snake when their radius of gyration is greater than the effective gel pore radius. Instead, such macromolecules probably condense to a shape more akin to a "beer can" than a snake, as is illustrated in FIG. 10, and therefore do not move easily in a direction transverse to the long axis of the "beer cans." Indeed, it is believed that the use of a gradient rather than a uniform field is one of the critical factors for forcing large molecules, such as DNA molecules, into the desirable elongated cylindrical or snake shape, as is illustrated in FIG. 11. Moreover, it is believed that the proper choice of a frequency at which the change from one field to antoher should occur, is related to the time it takes the particle (molecule) of interest to orient itself into the desired elongated cylindrical or snake shape, and that this time t is related to the mass of the particle (the molecular weight) M, the effective pore radius of the gel r, and the measured velocity of the particle in the gel v, in accordance with the relationship t.varies.M.sup.1.5 /(r.sup.2 v).

It should be emphasized that the hypothesis referred to above, while consistent with experimental results to date, is not to be taken as a factor limiting the scope of the invention, as the invention produces its beneficial results despite the fact that the underlying phenomenon may not be well understood, and despite the possibility that a totally different mechanism may be involved.

The blocks of gel or gel inserts of this invention may be formed by suspending cells in a liquid such as agarose or acrylamide, and then pouring the liquid cell mixture into suitable molds for solidification under appropriate conditions. In the case of cells with cell walls e.g. yeast cells or bacterial cells spheroplasts are prepared by diffusing a reagent or reagents which degrade the cell wall into the inserts e.g. lysozyme in the case of bacterial cells or zymolyase in the case of yeast cells. A high concentration of EDTA, for example, 0.5M, may be maintained to reduce nuclease activity to levels where double-stranded breaks can be detected. The insert containing naked DNA is pressed into slots formed to receive the molded gel inserts in the running gel used for electrophoresis. The advantages of this procedure for producing high molecular weight DNA include the following: Spheroplasting in the case of cells with cell walls does not have to be closely followed; any lysis which occurs during spheroplasting is not harmful to DNA integrity; stationary cells may be used; and centrifugation and precipitation steps are unnecessary. Although these gel inserts are particularly useful for use in DNA separation they are equally well suited for use in separating other particles such as RNA molecules.

An illustrative mold used in this new technique is shown in a perspective view in FIG. 14, and comprises a pair of matching rectangular blocks 14a and 14b which can be secured in the illustrated configuration by means of screws 14c. The top block 14a has a number of molding channels 14d which go through the entire thickness of the block, while the bottom block 14b is solid. When the blocks are assembled in the configuration shown in FIG. 14, suitable agarose gel with suspended cells is poured into the molding channels 14d and allowed to solidify. The blocks 14a and 14b are then taken apart, and the insert blocks such as 14e are carefully extracted, placed in lysing material to lyse the suspended cells are are then carefully inserted snugly into matching wells formed in the electrophoresis gel, e.g., by a comb whose outer shape and dimension matches the molding channels 14d. Alternatively, lysing material may be placed on the inserts while still in the mold to lyse the suspended cells. The molding block for making the gel inserts may be separate from or incorporated as part of the electrophoresis apparatus. It should be appreciated that the gel inserts of this invention can be used with conventional electrophoresis methods and apparatuses as well as with the method and apparatus described in this specification.

Another feature of this invention is the ability to form the gel inserts and to insert the gel inserts into the electrophoresis chamber automatically. The device used for making the gel inserts has a means for metering a desired number of cells into a well of a mold for forming the insert. Liquid agarose or acrylamide is metered into the well either simultaneously with the cells, or subsequent or prior thereto by the same or a separate metering device. After the gel solidifies, the machine, by means of another metering device, meters a desired amount of lysing solution into the wells to form the lysed cells or in the case of cells with cell walls spheroplasts in the gel inserts. This same device or a separate device may be supplied with a means for withdrawing the gel insert containing the lysed cell from the well. A means for inserting this gel insert into the electrophoresis chamber may then be used to automatically load the chamber.

The use of the gel insert of this invention minimizes the handling damage that can be caused by a machine forming lysed cells and loading the lysed cells into an electrophoretic chamber. The following examples demonstrate certain aspects of the invention but, of course, should not be taken as limiting its scope:

Gels about 1 cm thick were cast in 10 cm.sup.2 disposable square Petri dishes. Wells for the sample were formed in a conventional manner using a plastic comb with teeth 0.250".times.0.0787", spaced 0.125" apart. The gels consisted of 1.5% low endoosmosis agarose (Miles Biochemical Company) dissolved in TBE (10.3 g Tris, 5.5 g boric acid and 0.93 g disodium EDTA per liter). Electrophoresis buffer (TBE) was continuously circulated via a magnetically driven polypropylene-housed vane pump (Fischer Scientific) and cooled in a re-circulating refrigerated bath (Haake, type T-52), as illustrated in FIG. 12. The intake and discharge ends of the circulation pipes were close to the gel, and delivered and withdrew liquid buffer at two diametrically opposite corners of the gel square. Samples were loaded into wells using a Gilson Pipetman with the pippette tip ends cut to minimize shear. DNA was visualized after soaking gels in 0.5 micrograms of ethidium bromide per ml of TBE. Photographs were taken using Polaroid 107 film with shortwave U.V. illumination. Exposure times varied from 15 to 180 seconds at f8 depending on samples.

Bacteriophage viruses T7, T2, and G were prepared by lysing a given amount of virus overnight at 50.degree. C. in NDS as described in Laurer et al., Journal of Microbiology, 1975, 95: 309-326. The resulting lysates were then dialysed overnight against the electrophoresis buffer. The bacteriophage DNA masses in daltons are believed to be: T7=2.7.times.10.sup.7 ; T2=1.2.times.10.sup.8 ; and G=5.times.10.sup.8. A 0.02 microgram sample of each DNA was loaded into the wells in 5 microliters of 10% glycerin, TBE and 0.0015% bromphenol blue. Samples were run into gel with a single field for 15 minutes before pulsing. Optimal pulse times, in seconds, for resolution of macromolecules near or at the molecular weight of the following examples were T7=0.25; T2=4; and G=20. The term "pulse time" refers to pulse width, i.e., the time interval over which one of the fields is on (or high) while the other one is off (or low). In this experiment fields of the type and voltage levels illustrated in FIGS. 5 and 6 were used, i.e., both fields had intensity gradients. The relative mobility obtained in this experiment was G=1; T2=2.5; and T7=8.

Various strains of yeast were grown to mid-log phase in 100 to 1000 ml of YPD (YPD:1 g yeast extract, 2 g dextrose and 2 g bactopeptone added to 1 liter of distilled water). Spheroplasts were made as described in Cryer et al., Progress in Cell Biology, Vol. 12, 1975, pp. 30-44. The spheroplasts were then lysed in NDS overnight at 50.degree. C. Yeast lysates were prepared in NDS with concentrations ranging from about 10.sup.9 to 2.times.10.sup.10 cells per ml of lysate. Generally, 90 microliters of lysate were loaded using a blue-tipped (1 ml capacity) Pipetman. Samples were run into 1.5% agarose gel at 100 volts for 45 minutes with a single field. Pulse times of 15-45 seconds at 200 volts (fields E1 and E2 of FIGS. 5 and 6) give the molecular weight resolutions shown in reduced scale in FIG. 13.

The experimental conditions of Example B were used, except that gels were run in the dark and contained 0.5 micrograms per ml ethidium bromide in the gel as well as the circulation buffer, and pulse times were 30 and 45 seconds, using D-273 yeast lysates. Clear resolution of many chromosomes was obtained.

In the examples above, lysing was done in a conventional manner and the lysates were transferred to the electrophoresis gel in a conventional manner. Such handling of lysates results in breakage and other damage to fragile macromolecules. The following example illustrates the method of substantially avoiding such deleterious effects. The example follows the general concept of this invention whereby cells or spheroplasts (cells minus cell walls) are suspended in agarose gel, which is poured into molds to form inserts. The inserts are placed in lysing solution to lyse the suspended cells or spheroplasts, and then the intact inserts are placed snugly into matching wells in the electrophoresis gel. The gel making up the inserts can be the same as, or compatible with, the electrophoresis gel.

Yeast spheroplasts (10.sup.10 to 10.sup.11 cells per ml 1% low gelling agarose in TBE) were suspended in agarose gel and poured into the mold channels to form inserts. The inserts were then placed into NDS at 50.degree. C. overnight, thereby lysing the suspended spheroplasts. Yeast cells, previously treated with mercaptoethanol were also suspended in 1% agarose gel, but in this case 75 microliters of a Zymolyase 5000 mixture (2 mg per ml 0.01M sodium phosphate, 50% glycerine) was added to the insert mixture prior to molding the inserts. 75 microliters of Zymolyase was also added to 0.8 ml of LET (0.5M tetrasodium EDTA, 0.01M Tris, pH=7.5). Molded inserts with the yeast cells were added to the LET, and incubated overnight at 37.degree. C. The resulting suspended spheroplast-derived inserts were placed in matching wells in the electrophoresis gel. Electrophoresis using the conditions discussed above in connection with Examples A-C, provided good chromosomal DNA resolution.

2.5.times.10.sup.7 mouse 3T3-R500 cells were lysed in 0.3 ml of NDS at 50.degree. C. for four (4) days. The lysates were then loaded into 1.5% agarose cells in the TBE and run at 200 Volts with 30 second pulsing. One diffuse band was obtained. It moved as if it had the molecular weight of intact double minute DNA (mol. wt. approx. 600.times.10.sup.6). Marker was G phage (mol. wt. approx. 500.times.10.sup.6).

The new kind of electrophoresis discussed above has numerous applications. As one example, by use of this technique yeast chromosomal DNA has for the first time been successfully separated and characterized by size. Another use of the new technique is exploring the nature of DNA-gyrase complexes in E. coli supercoiled, chromosomal domains to map gyrase locations and thus provide tools for eucaryotic chromosome analysis. The new technique is particularly advantageous when different molecules, such as different DNA molecules, are close to each other in mass. The use of alternating fields each with an intensity gradient, tends to sharpen resolution dramatically and allow unexpected resolution for molecules close to each other in mass. Another use is resolving a great number of bands in the same gel, an important consideration when eucaryotic DNA is being analyzed. Yet another use of the new kind of electrophoresis is to purify molecules such as enzymes, e.g., urokinase, myosins or hyaluronic acids so as to provide a purified sample which can serve as the basis for developing a way to produce the same or an equivalent molecule. As yet another use, the effect of various agents, such as drugs, can be assessed for their effect on chromosomes, nucleic acids and proteins because of the ability to separate such materials provided by the invention. As yet another example, polymers can be accurately and quickly analyzed for molecular weight distribution, branching, and other physical properties by use of the new kind of electrophoresis. As still another example, intact or cut human, animal or plant chromosomes can be analyzed using the new kind of electrophoresis.

It should be clear that the laboratory device discussed in connection with FIGS. 1-8, and the particular kinds of electric fields used thereby, and the insert molding device discussed in connection with FIG. 14, are only specific examples which are convenient for explaining certain principles of the invention. Numerous variations are possible and are within the scope of the invention. For example, a differently shaped electrophoresis chamber, or differently produced, distributed or varied electric fields can be used so long as the particles are acted on by electric fields varying with time so as to move them in overall directions generally transverse to at least two of the relevant, operationally significant fields. For example, the desired fields can be generated by differently shaped electrodes, by suitably excited coils or by other sources or combinations of different (in kind) sources, and the relevant field directions can be controlled by other means, such as without limitation, changing the net direction of the field or changing the electrode characteristics (e.g., potential). Similarly, the desired field gradient can be produced in any number of ways, such as by selecting an appropriate shape for the relevant electrodes, by maintaining different electrode portions at different potentials or by the interaction of two or more fields. Moreover, more than two fields can be used, so long as the net effect is at least to act in the desired manner on a particle first in one direction, then in another direction transverse to the first, etc., so as to move the particle in a third direction transverse to the first two.

It has been found desirable, in the above-described preferred exemplary embodiment of the new electrophoresis device, to have a number of discrete electrodes, and to interconnect them through devices (such as diodes) which allow current flow to each in only a selected direction. Moreover, it has been found desirable to have the wire electrodes extend along the interior sidewalls of the chamber vertically, or nearly so, because such electrodes make it particularly convenient to generate the desired electrical fields, and because with such electrodes when they are long enough in the vertical direction it is possible to have several gel layers on top of each other, each containing samples of particles, and to subject all of them to substantially identical electric fields so as to carry out electrophoresis in all of them concurrently. To generate more complex fields, or to provide more freedom of choice in producing fields of selected characteristics, such as the fields E, E1 and E2 in FIGS. 4-6, each electrode (or at least electrode of a selected plurality of electrodes) can have its own, switchable, power supply connection such that each can be selectively maintained at any positive or negative electrical potential within a selected range (or at ground). In some cases, as few as three electrodes will suffice, and two of them can be connected (intermittently) to the same potential, so long as they cooperate with each other to produce at least two electrical fields which have the desired characteristics (i.e., being transverse to each other).

As one variation, the new kind of electrophoresis arrangement described above can make use of high frequency switching between transverse fields, e.g., at frequencies in the range from about 10.sup.6 to about 10.sup.9 Hz, superimposed on one or more steady, or more slowly switching fields such as the fields E, E1 and E2 discussed above. It is believed that the rapidly switching field or fields can help rotate (or orient) particles such as macromolecules in a desired manner while the steady or slowly switching field or fields can serve to move the particles in the desired overall direction. This arrangement of rapidly switching fields and steady or slowly switching fields can in fact use as few as two transverse fields, at least one of them having a steady or slowly switching intensity component and a rapidly switching intensity component superimposed thereon. For example, mutually transverse fields E1 and E2 as in FIG. 7 can be used, but at least one of the electrodes can have superimposed on the illustrated squarewave voltage waveform, a much higher frequency voltage waveform of a selected amplitude, such as at a frequency from about 10.sup.6 to about 10.sup.9 Hz.