WO1991011709A1

WO1991011709A1 - High-viscosity polymer matrix and methods

Info

Publication number: WO1991011709A1
Application number: PCT/US1991/000545
Authority: WO
Inventors: Robert S. Dubrow
Original assignee: Applied Biosystems, Inc.
Priority date: 1990-01-29
Filing date: 1991-01-25
Publication date: 1991-08-08
Also published as: US5164055A; DE69131124D1; EP0513164B1; DE69131124T2; ATE178990T1; JP2701091B2; EP0513164A1; EP0513164A4; ES2132084T3; JPH06504612A; GR3030209T3; DK0513164T3

Abstract

A supported, substantially uniform matrix for use in electric field-induced separation of molecular components in a sample. The matrix is prepared by forming a viscoelastic polymer matrix and pumping the matrix into a capillary or preparative-scale tube. The polymer type, concentration, and molecular weight are selected to optimize separation of protein or nucleic acid components. The matrix can be expelled from the tube, after fractionation, for analysis and/or recovery of fractionated components.

Description

HIGH-VISCOSITY POLYMER MATRIX AND METHODS

1. Field of the Invention

The present invention relates to separation of molecular components, and in particular, to a flowable, high-viscosity electrophoretic matrix effective as an separation medium for electrophoresis and isoelectric focusing.

2. References

Cohen, A.S., et al. Anal Chem, 59:1021 (1987).

Cohen, A.S., et al, J. Chromatography, 458:323 (1988) . Compton, S.W., et al, BioTechniques, β(5) :432 (1988) .

Herrin, B.J., J Colloid Interface Sci, 115(1) :46 (1987) .

Kaspar, T.J.,- et al, J Chromatography, 458:303 (1988) .

Lauer, H.H., Anal Chem, 58:166 (1985).

McCormic , R.W., Anal Chem, 60(21) :2322 (1988).

3. Background of the Invention: Electrophoresis Electrophoresis is widely used for fractionation of a variety of biomolecules, including DNA species, proteins, peptides, and derivatized amino acids. One electrophore- tic technique which allows rapid, high-resolution separa¬ tion is capillary electrophoresis (CE) (Cohen, 1987, 1988, Compton, Kaspar) . Typically, the CE employs fused silica capillary tubes whose inner diameters are between about 50-200 microns, and which can range in length between about 10-100 cm or more.

In the usual electrophoresis procedure, an electro¬ phoresis tube, such as a capillary tube, is filled with a fluid electrophoresis medium, and the fluid medium is crosslinked or temperature-solidified within the gel to form a non-flowable, stabilized separation medium. A sample volume is drawn into or added to one end of the tube, and an electric field is placed across the tube to draw the sample through the medium. Electrophoretic separation within the matrix may be based on molecular size, in the case of nucleic acid species (which have roughly the same charge density) , or on a combination of size and charge, in the case of peptides proteins.

The polymer concentration and/or degree of cross- linking of the separation medium may be varied to provide separation of species over a wide range of molecular weights and charges. For separating nucleic acid frag¬ ments greater than about 1,000 bases, for example, one preferred temperature-solidified material is agarose, where the concentration of the agarose may vary from about 0.3%, for separating fragments in the 5-60 kilobase size range, up to about 2%, for separating fragments in the 100-3,000 basepair range (Maniatis) . Smaller size fragments, typically less than about 1,000 basepairs, are usually separated in cross-linked polyacrylamide. The concentration of acrylamide polymer can range from about 3.5%, for separating fragments in the 100-1,000 basepair range, up to about 20%, for achieving separation in the size range 10-100 basepairs. For separating proteins, crosslinked polyacrylamide at concentrations between about 3%-20 percent are generally suitable. In general, the smaller the molecular species to be fractionated, the higher the concentration of crosslinked polymer. The resolution obtainable in solidified electropho¬ resis media of the type described above has been limited, in the case of small molecular weight species, by dif¬ ficulties in forming a homogeneous, uniform polymer matrix at high polymer concentration within an electro- phoresis tube, and especially within a capillary tube. In the usual method for forming a high-concentration solidified matrix in a tube, a high-concentration polymer solution, in a non-crosslinked, low-viscosity form, is introduced in fluid form fluid into the tube. The fluid material is then crosslinked, for example, by exposure to light in the presence of persulfate and a cross-linking agent such as bisacrylamide.

At high polymer concentrations, reaction heat gradi¬ ents formed within the tube tend to produce uneven rates of reaction and heat turbulence which can lead to matrix inhomogeneities. Also, entrapped gas bubbles generated during the crosslinking reaction produce voids throughout the matrix. The non-uniformities in the matrix limit the degree of resolution which can be achieved, particularly among closely related, small molecular weight species.

Alteratively, in the case of temperature-solidifying polymers the polymer is introduced into an electrophore¬ sis tube in a fluid form, then allowed to gel to a solid form by cooling within the gel. This approach, however, is generally unsuitable for fractionating low molecular weight species, such as small peptides and oligonucleo- tides, since the polymers, such as agar and agarose, which are known to have the necessary temperature-solidi¬ fying setting properties are not effective for fractiona- ting low molecular weight species, even at high polymer concentrations.

A second limitation associated with crosslinked or temperature solidified matrices is the difficulty in recovering fractionated molecular species within the matrix, after electrophoretic separation. In the case of a preparative-scale electrophoresis tube, the solidified matrix must be carefully separated from the walls of the tube before the matrix can be removed, a procedure which is virtually impossible with capillary tubes. Even after the matrix is removed, and the region of the matrix containing the desired molecular species is identified, the species of interest can be recovered from the matrix region only by a lengthy elution procedure, or by elec- trophoretic elution.

4. Background of the Invention: Isoelectric Focusing

Isoelectric focusing (IEF) is a separation method based on migration of different molecular components to their isoelectric points in a pH gradient. The pH gradi¬ ent is established by subjecting an ampholyte solution containing a large number of different pKi species to an electric field. Molecular components contained in (or added to the equilibrated ampholite solution) will then migrate to their isoelectric points along the pH gradi¬ ent. The components can then be isolated by eluting the gradient and capturing selected eluted fractions.

Although IEF methods are usually carried out in a low-viscosity fluid medium, it is occasionally advantage- ous to perform the IEF separation in a stabilized matrix. One potential advantage of a stabilized matrix is greater resolution, by eliminating the band spreading that occurs when a fluid medium is eluted from a tube, and by elimi¬ nating electroosmotic effects which may be present in capillary IEF. A stabilized separation medium allows use of certain gel analysis techniques, such as gel autoradi- ography which are not possible with a fluid medium. Crosslinked or temperature-stabilized gels of the type described above have been employed in IEF methods, but present some of the same limitations noted above for electrophoretic methods. In particular, the stabilized gels are generally not removable from capillary tubes, and isolating separated molecular species from the matrix may be inconvenient in that exhaustive dialysis or elec- troelution are typically required.

5. Summary of the Invention

It is one general object of the invention to provide a supported matrix for use in electric field-induced separation of molecular components, which substantially overcomes or reduces problems and limitations associated with temperature stabilized or crosslinked polymer matri¬ ces of the type described above. It is yet another object of the invention to provide a method of preparing and using such supported medium.

The invention includes, in one aspect, a method of preparing a supported matrix for use in electric field- induced separation of molecular components in a sample. The method includes first forming an aqueous, viscoelas¬ tic polymer matrix characterized by (i) a water-soluble, substantially non-crosslinked polymer, having a molecular weight of at least about 10,000 daltons, and (ii) a viscosity of at least about 5,000 centipoise. This viscoelastic polymer matrix is then pumped into an elon¬ gate separation chamber, filling the chamber substantial¬ ly uniformly with the matrix.

In one embodiment, the separation chamber may be contained within a capillary tube having a plug at one end to prevent flow of the polymer through that tube end. The polymer forming the matrix is preferably a linear polymer having a molecular weight of at least about 1 million daltons. Preferred polymers for separa- tion of proteins are polyethyleneglycol (polyethylene oxide), polyacrylamide, polymethacrylamide. Preferred polymers for the separation of nucleic acid species are hydroxylated polymers, including hydroxylated alkyl cellulose polymers and polyhydroxyvinyl. In another aspect, the invention includes a sup¬ ported matrix prepared by the above method, for use in electric field-induced separation of molecular components in a sample. The matrix may contain an electrolyte, for electrophoretic separation of sample component, or an ampholyte, for separation of sample components by iso¬ electric focusing (IEF) in a pH gradient.

Another aspect of the invention is an electropho¬ retic method for separating components of a sample. In practicing the method, sample is applied to one end of the supported matrix of the type described above, and an electric field is established across the matrix. The type, molecular weight, and concentration of the polymer may be varied to optimize fractionation of a given sample of peptides or nucleic acids. Another suitable matrix conditions, it is possible to separate phosphorylated oligonucleotides from their non-phosphorylated analogs.

The method may further include, following electro¬ phoretic separation, removing selected regions of the matrix containing separated sample components, and isola- ting separated sample components from removed regions of the gel.

Isolating the sample from the excised region of the matrix can be carried out conveniently by liquefying the matrix, either by dilution or heating. In a preferred embodiment of the electrophoresis method, for fractionating nucleic acid species, the polymer matrix is formed of a mixture of polymers. One of the polymers in the mixture, as exemplified by poly- acrylamide or polyethylene oxide, is effective to frac¬ tionate nucleic acid by a sieving mechanism which is dependent on both the size and concentration of the polymer. The other polymer in the mixture, as exempli¬ fied by a water-soluble, hydroxylated cellulose compound, is apparently acts to separate different molecular weight species on the basis of a hydrophilic inter-reaction which is dependent on polymer concentration, but not polymer molecular weight.

The invention further includes a method of separa- ting molecular components in a sample by IEF in the above matrix. The matrix used in the method contains an am¬ pholyte effective to form a selected pH gradient within the matrix under the influence of an electric field. Sample components separated in the matrix may be iden- tified and/or isolated by pumping the matrix from the chamber.

These and other objects and features of the inven¬ tion will become more fully apparent when the following detailed description of the invention is read in conjunc- tion with the accompanying drawings.

Brief Description of the Drawings Figure 1 is a simplified view of a pumping system used for (a) measuring the viscosity of a viscoelastic polymer solution, and (b) filling a capillary tube with such a matrix, in accordance with the invention;

Figure 2 is a plot of polymer viscosity, as measured by a viscometer (left ordinate, closed circles) or by flow characteristics through a capillary tube (right ordinate, open circles) plotted as a function of polymer concentration;

Figure 3 is an enlarged, fragmentary portion of a capillary-tube supported matrix constructed according to the invention;

Figure 4 is a schematic diagram of a capillary tube system used in practicing electrophoresis or IEF methods, in accordance with the invention;

Figure 5 is a schematic view of steps which may be employed for detecting and isolated fractionated species in a viscoelastic matrix, in accordance with the inven¬ tion;

Figures 6A are electropherograms of proteins frac¬ tionated on a matrix prepared with 4.5 weight percent (6a) , weight percent (6b) , and 10 weight percent poly- acrylamide(6C) ;

Figures 7A and 7B are electropherograms of a 40mer to 60mer oligonucleotide acid ladder fractionated by electrophoresis on linear polyethylene oxide polymers of approximate molecular weight 900 kilodaltons (6A) and 7,000 kilodaltons (6B) , both at 11.2 weight percent polymer;

Figures 8A and 8B are electropherograms of a 40mer to 60mer oligonucleotide acid ladder fractionated by electrophoresis on a linear polyethylene oxide polymer of approximate molecular weight 7,000 kilodaltons at 6 weight percent (8A) and 15 weight percent (8B) ;

Figures 9A-9C are electropherograms of a 40mer to 60mer nucleic acid ladder fractionated by electrophoresis on matrices formed of linear hydroxyethyl cellulose (HEC) polymers having relatively low (9A) , intermediate (9B) , and high (9C) molecular weights, each at a polymer con¬ centration of about 10 weight percent; Figures 10A-10D are electropherograms of a 20mer to

40mer nucleic acid ladder fractionated by electrophoresis on matrices formed of a linear hydroxyethyl cellulose

(HEC) polymer, at polymer concentrations of 3% (10A) , 10% (10B), 15% (IOC), and 25% (lOd) ;

Figure 11 is a plot of migration rate of a 30mer oligonucleotide, as a function of HEC concentration, derived from the plots in Figures 10A-10D; and

Figure 12 is electropherogram of a 40mer - 60mer nucleic acid ladder fractionated by electrophoresis on matrices formed of a linear hydroxyethyl cellulose (HEC) polymer, at polymer concentrations of about 5.5 weight percent, and polyacrylamide, at a concentration of about 11 weight percent; and Figures 13A-13C are electropherogram of sets of polyadenylic acid including both phosphorylated (dominant species) and non-phosphorylated (minor) species (13A) , non-phosphorylated species only (13B) , and a mixture of phosphorylated and non-phosphorylated species (13C) fractionated by capillary electrophoresis in 10 weight percent HEC.

Detailed Description of the Invention Section I describes the supported polymer matrix of the invention, and methods of preparing and characteri¬ zing the polymer solution forming the matrix. Methods of separating biomolecular components, such as proteins and nucleic acids, by electrophoresis in the supported matrix are described in Section II. Isoelectric focusing me- thods employing the supported matrix are given in Section III. I. Supported Polymer Matrix A. Polymers

The polymers used in the preparing the supported matrix of the invention are water-soluble (hydrophilic) polymers which are capable of forming a viscoelastic fluid above a given polymer concentration in an aqueous medium. As defined herein, a "viscoelastic fluid" is one which has a viscosity above about 1,000 centipoise, and may have a viscosity of 400,000 centipoise and greater, and has elastic properties, as evidenced by its tendency, on stretching, to form string-like filament (s) which resist the stretching force. This viscoelastic behavior is similar to the "stringing" which occurs, for example, in rubber cement (which is formed of hydrophobic poly- mers) .

The viscoelastic properties of the polymer are observed generally in linear water-soluble polymers having a molecular weight greater than about 10 kilodal¬ tons (10,000), preferably greater than 250 kilodaltons, and typically between about 1 to 10 million daltons. The concentration of polymer needed to achieve the requisite viscoelastic state is dependent on the molecular weight of the polymer, and can range from about 40 weight per¬ cent for the smallest polymers, e.g., 10-100 kilodaltons, to 1-5 weight percent for polymers greater than 1 million daltons.

This relationship between polymer molecular weight and concentration can be understood from the following simplified picture of polymer interaction. First, the viscoelastic properties of the polymer matrix are due to polymer chain entanglements which allow the material to stretch and exert and counter elastic force, as the intertwined chains are stretched. Secondly, in its ability to form tangled polymer chains, the polymer may be viewed as having opposite end regions which do not contribute to chain entanglements, and a center region which does. Since viscoelastic properties are difficult to achieve below about 10 kilodaltons, the size of the opposite• end regions which does not contribute to en¬ tanglements is assumed to be close to 5 kilodaltons, for a linear polymer. At relatively low molecular weight, the portion of the polymer contributes to chain entangle¬ ments is relatively small, and thus a high polymer con- centration is required, whereas the opposite is true for high molecular weight polymers. Additionally, as the size of the polymer increases, the potential for multiple entanglements increases, reducing the polymer concentra¬ tion at which viscoelastic properties are observed. It will be appreciated from the above that high molecular weight branched polymers would be able to form an entangled-chain matrix, but only if the polymer bran¬ ches are themselves quite long, e.g., corresponding in molecular weight to about 10 kilodaltons. Preferred polymers for use in the present invention are linear polyoxides, polyethers, such as polyethylene oxide (poly- ethylenegl col) , and polypropylene oxide, polyethylene imine, polyacrylic acid, polyacrylamide, polymethacryl- amide, polymethacrylic acid, polyvinylacetate, polyvinyl- pyrrolidone, polyvinyloxazolidone, and a variety of water-soluble hydroxyl polymers, such as natural gums (xanthin, dextran, guar, etc.), water-soluble cellulose compounds, such as methylcellulose and hydroxyethylcel- lulose, and co-polymers and blends of these polymers. Suitable water-soluble polymers having a wide range of molecular weights (often expressed in terms of solu¬ tion viscosity, at a given polymer concentration) are available commercially, or can be prepared under defined polymer formation conditions, as illustrated in the examples below. Example 1A describes matrix formation with polyethylene oxide (PEO) , at molecular weights of 900 and 7,000 kilodaltons. Example IB describes matrix formation with hydroxyethylene cellulose (HEC) having a broad range of viscosities. The matrix described in Example 1C is formed by polymerizing acrylamide monomer in the presence of suitable polymerizing agents to form a polyacrylamide polymer matrix.

As will be discussed below, and in accordance with one aspect of the invention, the polymer matrix with desired viscoelastic properties is formed outside a tube support, and subsequently introduced into the tube under high pumping pressure. An important advantage of this approach, over prior art methods in which a polymer material is solidified within a tube support, is the ability to achieve desired matrix properties by control¬ ling reaction conditions, such as temperature, pressure, mixing conditions, light input (for photopolymerized polymers) , concentration of polymerizing agents (which may require continued addition of the agents to the reaction mixture) , time of reaction termination, and curing conditions. In addition, as indicated above, the reaction can be carried out under conditions which avoid inhomogeneities in the matrix due to local heat gradi- ents, bubble formation, and poor mixing of reaction components. The polymerization reaction is carried out under conditions in which the polymer remains substan¬ tially non-crosslinked.

The polymerization reaction, for use in forming a viscoelastic matrix, is illustrated by the polyacrylamide matrix formed in Example 1C, where polymerization of acrylamide monomer was carried out under controlled reaction conditions until a desired polymer viscosity was achieved. To form a polymer matrix from prepolymerized mate¬ rial, the polymer may be added, for example, in powder form to a given volume of aqueous medium, and the polymer allowed to disperse under selected mixing conditions, for example with vortexing. Typically, the matrix is allowed to sit for several hours after mixing to insure complete dissolution of the material. If desired, additional aqueous medium may be added to the matrix to reduce viscosity, or the matrix may be dehydrated, for example, under reduced pressure, to increase viscosity. This method is illustrated in Example 1C.

Alternatively, the polymer may be pressed into block or slab form, such as by scintering in a press, and the slab then cut into a known-weight sections for dissolu- tion in a known volume of aqueous medium. Typically, the polymer block is allowed to dissolve in the medium at room temperature for several hours until the matrix has a uniform consistency. Example 1A illustrates this ap¬ proach for preparation of different molecular weight PEO matrices.

The polymer matrix prepared for use in electrophore¬ sis (Section II) is formed in an aqueous electrolyte medium, such as a conventional Tris-borate or the like. The electrolyte solution may be formulated to include water-miscible solvents, such as DMSO, ethanol, or the like, if necessary, to increase the solubility of certain molecular species in the matrix. The matrix prepared for use in isoelectric focusing (IEF) is prepared to include standard ampholyte solutions designed to form a selected pH gradient, on equilibration in an electric field.

B. Viscosity Characteristics

The polymer matrix is formed as above to have a viscosity, at room temperature of at least about 5,000 centipoise, and preferably above about 50,000 centipoise.

For polymer solution viscosities less than about 400,000 centipoise, solution viscosity can be measured using a conventional viscometer. In operation, a visco¬ meter rotates a spindle immersed in the test liquid through a spring. The degree to which the spring is wound, detected by a rotational sensor, is proportional to the viscosity of the fluid. Table 1 below shows viscosity values measured for high molecular weight polyacrylamide (PA) at the various polymer concentrations indicated. The polymers were prepared substantially as described in Example IC.

The viscosity of the polymer matrices was measured using a Brookfield Digital Viscometer Model #DV-11 (Stoughton, MA) . A #27 spindle was used in combination with the Brookfield small sample adaptor. The viscosi¬ ties were measured at 25^oC'"^{nd arβ βχPEβ}"^βd *^{» ∞tipoi}".

At high polymer viscosities, where the fluid has a grease-like consistency, the rotating spindle of the viscometer produces a cavity within the fluid, preventing reliable viscosity measurements. At high viscosities in which the fluid has a highly elastic character, the fluid tends to wrap itself, stringlike, about the spindle, also preventing accurate viscosity measurements. With either type of high-viscosity fluid, it is necessary to measure viscosity by alternative means.

One system for performing viscosities measurements of highly viscous, elastic polymer solution is illustra¬ ted at 20 in Figure 1. The system includes a high pres¬ sure pump 22 of the type used for high-pressure liquid chromatography. One suitable pump is a Model 140A high- pressure pump available from Applied Biosystems (Foster City, CA) . The pump is connected to a stainless steel tube 24 which is connected to the pump through a fitting 26, such as connector Part # U437 available from Upchurch Scientific Inc. (Seattle, WA) .

The distal end of the tube is connected to a capil- lary tube 28, also through a high pressure fitting of the type described above. The capillary tube has a preferred lumen diameter of between 50 and 100 microns, and a preferred length of between about 20-40 cm.

To perform a viscosity measurement, the steel tube is filled with the polymer matrix, typically using a syringe to force the matrix material into tube 24.. With activation of the pump, at a selected pumping volume rate, the pump pressure will increase until the preset pumping rate is achieved, up to a maximum pump pressure, e.g., 5,000 psi. The viscosity is then measured as (a) the pressure of the pump when the matrix material has been pumped through the entire length of the capillary tube, or (b) at higher viscosities, the distance the matrix material has traveled when the pump reaches its maximum pressure.

In the viscosity following viscosity measurements, a clean 30 cm length of 75 μm capillary was attached to the high pressure pump for each of viscosity determination. The flow rate was set at 100 microliters per minute on the pump and the polymer matrix was loaded as just de¬ scribed. The pump pressure or travel times are given in the Table 2 below.

As mentioned above, the viscosity of the matrix is measured either as the pressure the pump reaches when the matrix has been pumped through the entire 30 cm length of the tube, or for higher viscosity solutions, the distance traveled in the tube when a maximum pump pressure of 5,000+ psi is reached.

A plot of viscosity as a function of polymer con¬ centration is given in Figure 2, showing the change in viscosity measured by a Brookfield viscometer (solid circles) and as measured by flow pressure (open circles) for various concentrations of polyacrylamide.

As seen from the plot, the viscosity curve has a bend or elbow at about 6% polyacrylamide, corresponding to a polymer viscosity of about 7,000 centipoise and a pumping-pressure viscosity of about 1300 psi, under the above pumping conditions. Experiments aimed at determi¬ ning the lowest polymer concentration effective to re¬ solve proteins by electrophoresis demonstrate that resol- ution of proteins in the 14-30 kilodalton size range is lost significantly below about 5-6% PA, corresponding to a viscosity of about 5,000 centipoise.

The viscosity of the matrix which is used in prac¬ ticing the invention will be dictated by a number of factors whose importance will be described in Sections II and III below. As noted above, a minimum viscosity of about 5,000 centipoise is required for protein fractiona¬ tion by capillary electrophoresis and this value thus defines the lower limit of viscosity in a viscoelastic polymer solution useful in the present invention.

For use in electrophoretic. separation of molecular components in which fractionation is based on a polymer sieving effect (Section II) , lower solution viscosities, such as between 5,000 to 100,000 centipoise are suitable for resolving relatively high molecular weight compo¬ nents, such as proteins in the 20-50 kilodalton and higher range, and higher viscosities. For smaller pep¬ tides, polymer viscosities of about 3,000 psi or greater are suitable for fractionating smaller peptides. For nucleic acid fractionation by electrophoresis, the vis¬ cosity of the medium may vary widely for hydroxylated polymers, which appear to effect fractionation by a non- sieving interaction with nucleic acid species.

The viscosity of the solution may also be dictated by desired physical characteristics of the supported matrix. Thus, where it is desired to process the matrix in an intact form 1 after separation, the viscosity of the matrix must be sufficient to allow expulsion of the matrix under pressure in intact form. C. Supported Matrix

The supported matrix of the invention includes a support, typically a capillary or preparative-scale tube, defining an elongate separation chamber, and a polymer matrix of the type described above which fills the cham¬ ber uniformly and homogeneously, as described below.

Figure 2 shows an enlarged fragmentary portion of a capillary-tube supported matrix 30 formed in accordance with the invention. The capillary tube support 32 is a conventional capillary tube, typically having a length between about 10-200 cm, typically less than about 100 cm, and an inner diameter of preferably between about 25- 200 μ (microns) , typically about 50 μ.

Where the tube is a fused silica capillary, the inner wall has negative silane groups which may produce electroosmotic flow effects, as described, for example, in co-owned U.S. patent application for "Nucleic Acid Fractionation by Counter-Migration Capillary Electropho¬ resis", Serial No. 390,631, filed August 7, 1989. If desired, electroosmotic flow in a charged-wall capillary can be substantially eliminated by one of four approa¬ ches. In the first approach, the viscosity of the poly¬ mer is made sufficiently high, e.g., at a pumping pres¬ sure flow rate of 3,000-5,000 psi, that little or no electroosmotic flow can occur during the period of elec¬ trophoresis.

In a second approach, the capillary tube is provided with a water-permeable plug, such as plug 36 seen in Figure 3, or other obstruction which substantially pre- vents flow of a viscoelastic matrix out of the tube.

Another approach is based on chemical bonding of the matrix, after injection into the tube, with the walls of the tube. Here the tube is first chemically treated with a bifunctional coupling agent capable of reacting with chemical groups on the tube wall and with the polymer in the matrix. Silane coupling agents are suitable, for example, for fused silica tubes, and examples of silane bifunctional reagent chemistry can be found in "Silane Coupling Agents", Plenum Press, NY, 1982.

Finally, the electrophoresis tube may have coated walls, such as Teflon-coated walls, to mask surface wall charges or the tube itself may be formed from Teflon, in which case electroosmotic flow would not occur. Preparative-scale electrophoresis tubes in a variety of tube diameter and length sizes are available for use as a support in the invention. Typically, tube diameters of about 2-10 mm, and tube lengths of between about 15-40 cm are employed for preparative scale fractionation. Flow of matrix material in the tube may be prevented, as above, by the use of a frit or constricted plug or the like which prevents flow of the viscoelastic material out of the tube. Alternatively, or in addition, at high viscosities in the range 3,000 or greater pumping pres- sure, the matrix may be relatively stable against flow under gravity during the period of electrophoresis.

The tube support—either capillary or preparative- scale tube— defines an elongate separation chamber, indicated at 36 in Figure 3, which may include the entire length of the tube lumen or only a portion thereof. For example, it may be desired to fill one end region of the tube with a buffer solution. The matrix support is prepared, in accordance with the invention, by pumping the polymer matrix into the separation chamber, to fill the chamber substantially uniformly with the matrix.

The pumping system shown in Figure 1 is generally suitable for pumping polymer matrix into a capillary or preparative-scale tube. As above, the system is loaded with the polymer solution, and the pump is set to a selected pumping speed typically 50-100 μl/min. The material is pumped into the tube until the separation chamber (which may include only a portion of the tube) is filled. When pumped into the tube, the matrix material fills the chamber substantially uniformly and homogeneously, by which is meant the polymer matrix in the chamber has a substantially uniform density throughout the chamber, with substantially no discontinuities or voids in the matrix. This feature is achieved by virtue of (a) the uniform density and homogeneous bulk properties of the matrix which are achievable by forming the matrix outside of the tube, (b) the ability to pump the matrix into the tube without breaks, cracks, or voids forming in the matrix, and (c) the ability of the matrix to completely fill the chamber space as it is pumped into the tube.

The advantages of the supported matrix, as a medium for electrophoretic separation, can be appreciated from the foregoing. The matrix can be formed to a high vis- cosity, by a combination of high polymer molecular weight and/or concentration, for fractionating closely related molecular species, particularly small peptides and oligo- nucleotides. At the same time, the resolution achievable in the medium is significantly improved over prior art high-concentration gel-electrophoresis methods, by virtue of the greater homogeneity and lack of voids in the medium.

II. Electrophoresis Method A. Capillary Electrophoresis System

Figure 4 is a simplified schematic view of a capil¬ lary electrophoresis system 40 suitable for practicing the method of the invention. The system includes a capillary-tube supported matrix 30, such as described above with reference to Figure 3.

An anodic reservoir 42 in the system contains an electrolytic solution 44. The anodic end of the tube, indicated at 30a, is immersed in the solution, as shown, during electrophoresis. A reservoir 46 in the system may contain a marker solution, or may contain a sample of molecules to be separated, during an electrophoretic separation. Preferably the marker or sample material is dissolved in the electrolytic solution or in water. The two anodic reservoirs may be carried on a carousel or the like, for placement at a position in which the lower anodic end of the tube can be immersed in the reservoir fluid. Although not shown here, the carousel may carry additional reservoirs containing solutions for cleaning and flushing the tube between electrophoretic runs or different solutions, where two or more solutions are employed in a single electrophoretic fractionation me¬ thod. The opposite, cathodic end of the tube, indicated at 30b, is sealed within a cathodic reservoir 48 and is immersed in an cathodic electrolyte solution 50 contained in the reservoir, as shown.

A high voltage supply 52 in the system is connected to the anodic and cathodic reservoirs as shown, for applying a selected electric potential between the two reservoirs. The power supply leads are connected to platinum electrodes 54, 42 in the anodic and cathodic reservoirs, respectively. The power supply may be de- signed for applying a constant voltage (DC) across the electrodes, preferably at a voltage setting of between 5- 50 KV. Alternatively, or in addition, the power supply may be designed to apply a selected-frequency, pulsed voltage between the reservoirs. In general, the shorter the capillary tube, the higher the electric field strength that can be applied, and the more rapid the electrophoretic separation.

When operated in a pulsed voltage mode, the power supply preferably outputs a square wave pulse at an adjustable frequency of about 50 Hz up to a KHz range, and an rms voltage output of about 10-30 KV. Higher pulse frequencies, even into the MHz range may be suit¬ able for some applications. Completing the description of the system shown in Figure 1, a detector 58 in the system is positioned adjacent the cathodic end of the tube, for optically monitoring nucleic acid fragments migrating through an optical detection zone 60 in the tube. The detector may be designed either for UV absorption detection and/ or for fluorescence emission detection. UV absorbance is typically carried out at 240-280 nm, using, for example, a Kratos 783 UV absorbance detector which has been modi¬ fied by Applied Biosystems (Foster City, CA.) , by repla- cing the flow cell with a capillary holder. Fluorescence emission detection is preferably carried out at a .selec¬ ted excitation wavelength which is adjustable between about 240-500 nm, depending on the fluorescent species associated with the nucleic acid fragments, as discussed below. One exemplary fluorescence detector is an HP1046A detector available from Hewlett-Packard (Palo Alto, CA) , and modified as above for capillary tube detection. The detector is connected to an integrator/plotter 45 for recording electrophoretic peaks.

B. Detection and Isolation Methods

Figure 5 illustrates a variety of methods which may be employed to detect and/or isolate fractionated molecu¬ lar species after electrophoretic separation. The system illustrated is preferably a capillary tube system, al¬ though a preparative-scale tube may be required particu¬ larly for isolation of protein and peptide species.

One standard detection system, discussed above, employs a detector for detecting migration of molecular species through a detection zone located within down¬ stream end region of the gel. Alternatively, it may be desired to monitor the positions of fractionated species within the matrix, for example, by autoradiography, or for purposes of isolating fractionated species of inte¬ rest from the matrix. In order to detect or isolated fractionated species in the matrix, the electrophoretic run is stopped before the fastest migrating species of interest reaches the downstream end of the matrix. The tube, here indicated at 62, is then removed from the electrophoretic system and connected to a high-pressure pump 22, through a fitting 26, as described above, and the pump is set at a pumping rate suitable for expulsion of the matrix in intact form, such as 50-100 μl/minute for a capillary tube matrix.

According to an important advantage of the inven¬ tion, the fluid character of the matrix allows it to be expelled from the tube, either capillary or preparative- scale tube— in intact form. However, for this purpose, a polymer matrix with a viscosity of greater than 3,000- 5,000 psi, as defined above, is preferred, in order to provide sufficient cohesiveness and form to the matrix as it is pumped from the tube.

As the matrix is being expelled from the tube, the fractionated species in the matrix can be monitored by a detector 64 whose detection zone is placed either up¬ stream or downstream of the tube, as indicated. As above, the detector can be employed in generating an electropherogram which can be used to identify selected bands in the matrix, after expulsion from the tube, with known distances along the length of the gel. Alterna¬ tively, the matrix can be removed in intact form can "read" in a conventional gel scanner to generate an electropherogram of fractionated molecular species.

To isolate fractionated species of interest, typi¬ cally the electropherogram is divided into short sec¬ tions, as indicated in Figure 5, and those sections corresponding to bands of interest are further processed to obtain the desired isolated species. Two convenient methods are available for isolating separated molecules from the matrix. In the first, shown at the lower left in Figure 5, the matrix is diluted with an aqueous me¬ dium, to produce a low-viscosity protein solution which can then be further processed, for example, by column chromatography or addition of a precipitating agent, to isolated the desired molecular species, or by addition of Polymerase Chain Reaction (PCR) reagents, to amplify nucleic acid species in the dissolved matrix. Alterna- tively, the gel can be heated to a more fluid state from which the molecular species may be isolated, such as by acid precipitation, and filtration.

The lower right portion of Figure 5 illustrates band detection by autoradiography, where the fractionated species are radiolabeled. Here the intact matrix ex¬ pelled from the tube is placed on sensitive film, and after a sufficient exposure time, developed according to standard autoradiography methods to reveal the positions of the fractionated, labeled species. Alternatively, the molecular species can be trans¬ ferred to a suitable filter disc or the like by (a) pla¬ cing the matrix on the disc, (b) applying a vacuum to the disc, while heating the matrix to allow it to be drawn through the filter, and (c) collecting the fractionated material on the disc. The species trapped on the disk may be prelabeled, or identified by reaction with a suitable radiolabeled, species-specific probe, such as a labeled antibody or DNA probe. This technique can be employed, for example, as a modified Western-blot tech¬ nique for transferring fractionated nucleic acid species onto a nitrocellulose filter, and identifying bands of interest by their hydribization with a sequence-specific probe.

C. Protein Fractionation

The electrophoretic method of the invention may be employed for fractionating peptides and proteins over a wide size range, from small peptides to proteins of molecular weight up to 100,000 daltons or greater. Experiments carried out in support of the invention indicate that electrophoretic separation of proteins is based on a sieving effect in which the molecules are retarded, in passing through the matrix, by the polymer mesh forming the matrix, with larger polypeptides migra¬ ting more slowly through the matrix than smaller peptides (with a similar charge density) in an electric field.

Figures 6A-6C show electrophoretic separation of proteins in the 14.2 to 29 kilodalton range, as described in Example 2. The Figure 6A separation was performed at 4.5 weight percent polyacrylamide, prepared as in Example 1C. The viscosity of the medium was about 338 centi¬ poise, and the matrix at this viscosity is clearly unable to effect protein separation. The same proteins frac- tionated in 6 weight percent polyacrylamide, correspond¬ ing to a viscosity of about 35,000 is shown in Figure 6B, showing good separation of the four protein species. Even greater resolution was achieved at a 10 weight percent polyacrylamide concentration, corresponding to a viscosity of more than 200,000 centipoise.

Thus, for proteins in the size range between about 10 and 30 kilodaltons, a matrix viscosity between about 35,000 200,000 centipoise is suitable. For larger mole- cular weight protein, viscosities in the range 5,000 to 100,000 are preferred. Conversely, for low molecular weight peptide, matrix viscosities corresponding to a pumping pressure of 3,000-5,000 psi provide optimal resolution. At the highest viscosities, the method provides enhanced resolution of small peptide species, such as N- or C-terminal fragments and derivatized ana¬ logs thereof, by virtue of the homogeneous nature of the high-density matrix.

The desired viscosities may be achieved by a com- bination of polymer molecular weight and polymer con¬ centration, as described above. Specifically, the vis¬ cosity and degree of resolution of small species are increased with either higher molecular weight polymer or higher concentration of the polymer. Preferred polymers for polypeptide fractionation are polyacrylamide, poly- methacrylamide, and polyalkyl ethers, such as PEO, at polymer molecular weights of at least about 100,000. although a variety of other polymers may be employed.

Electrophoresis is carried out in a conventional electrophoresis system, such as the capillary tube system described with respect to Figure 4. The polypeptide sample is typically introduced by drawing the same into one end of the tube protein electrophoretically, typical¬ ly for a period of several seconds, then typically, then immersing the sample end of the tube in an electrophore¬ sis buffer during electrophoretic separation. The migra¬ tion of the protein bands past a detection zone near the downstream end of the tube can be monitored spectrophoto- metrically, as described above. Alternatively, the fractionated proteins can be detected in and/or isolated from the intact matrix, after removal from the tube support, also as described above.

D. Nucleic Acid Fractionation

Studies conducted in support of the present inven¬ tion indicate a number of unique features of the present invention for electrophoretic separation of nucleic acid species. One feature of the invention is the discovery that nucleic acids may be separated on high-viscosity polymer matrices by either of two distinct separation mechanisms. One separation mechanism involves molecular sieving, similar to the mechanism described above for protein fractionation, and appears to be the predominant in a matrix whose polymer does not include hydroxyl groups. Exemplary polymers which are effective in sieving-type separation are linear polyoxides, polyethers, such as polyethylene oxide (polyethyleneglycol) , and polypro- pylene oxide, polyethylene imine, polyacrylic acid, polyacrylamide, polymethacrylamide, polymethacrylic acid, polyvinylacetate, polyvinylpyrrolidone, and polyvinylox- azolidone, with polyacrylamide, polymethacrylamide and PEO polymers being preferred. As discussed in the section above, a sieving mechan¬ ism produces greater resolution of small molecular weight species as either the molecular weight of the polymer or polymer concentration is increased, since both variables increase the density of the polymer entanglements in the matrix.

Figures 7A and 7B show the resolution of a ladder of single-strand 40mer to 60mer oligonucleotides (40 to 60 bases) in 7.5 weight percent PEO matrices formed with a 900 and 7,000 kilodalton PEO polymers, respectively. Electrophoretic conditions are given in Example 3A. As seen, the oligonucleotides are poorly resolved on the lower molecular polymer matrix, but separated with sub¬ stantially baseline resolution on the higher molecular- weight polymer matrix.

Figures 8A and 8B shows the resolution of the same single-strand ladder on 7,000 molecular weight polymer at 6 and 15 weight percent polymer, respectively, at the electrophoretic conditions given in Example 3B. As seen, the oligonucleotide peaks were much better resolved at the higher polymer concentration.

A second, distinct separation mechanism is observed for nucleic acids separated on a matrix formed of a hydroxylated polymer. This mechanism appears to involve an interaction of the nucleic acid with polymer hydroxyl groups, rather than a sieving effect, and has been ob¬ served in a variety of hydroxylated polymers such as water-soluble hydroxylated cellulose compounds, as exem¬ plified by hydroxyethylcellulose (HEC) and polyvinyl al- cohol. Other suitable hydroxylated polymers include natural gums, such as xanthin, dextran, and guar.

An interaction mechanism would predict that separa¬ tion of nucleic acid species would be dependent on poly¬ mer concentration, since a higher polymer concentration would provide a greater number of nucleic acid/polymer interactions. On the other, separation of nucleic acid species should be independent of polymer molecular weight, since this variable would effect the density of polymer entanglement, but not the density of polymer hydroxyl groups.

Figures 9A through 9C are electropherograms of a ladder of single-strand 40mer to 60mer oligonucleotides (40 to 60 bases) in 10 weight percent HEC prepared from low-, intermediate-, and high-viscosity polymers, respec- tively. As seen, there is no appreciable effect of polymer molecular weight on oligonucleotide resolution. Electrophoretic conditions are described in Example 4A.

By contrast, when the intermediate-viscosity HEC polymer is formulated at increasing concentrations, as detailed in Example in Example 4B, a dramatic increase in oligonucleotide resolution is obtained. This is seen in Figures 10A through 10B, which show electropherograms of a 20mer to 40mer single strand nucleotide ladder at HEC weight percent concentrations of 3 (10A) , 10 (10B) , 15 (IOC) , and 25 (10D) . The spreading of the peaks with increasing polymer concentration is evident.

When time of travel through the matrix of the inter¬ mediate 30 mer peak is plotted as a function of polymer concentration in the above studies, the plot shown in Figure 11 is obtained, demonstrating the greater polymer concentration is associated with slower migration rate through the matrix.

According to another feature of the method, it has been discovered that electrophoretic separation of low molecular weight nucleic acids is enhanced by a combina¬ tion of sieving and interaction effects in the matrix. The polymer used in forming the matrix is a mixture containing at least non-hydroxyl polymer, such as poly- acrylamide or PEO, which can produce a sieving effect, and a hydroxylated polymer, such as a water-soluble hydroxylated cellulose compound, which interacts with nucleic acids hydroxyl groups.

The mixed-polymer matrix method is illustrated in Example 6, for fractionation of a 0mer to 60 mer single strand oligonucleotide ladder. The matrix is prepared from a polymer mixture containing about 11 weight percent polyacrylamide, as the "sieving" polymer, and about 5.5 weight percent HEC as the "interacting" polymer. The electropherogram of the fractionated oligonucleotides in seen in Figure 12. As seen, the peaks are resolved to an extent seen only in an HEC matrix at 25 weight percent (Figure 10D) . A comparison of the resolution achievable in a 10 weight percent sieving polymer (Figure 8A) and 5 weight percent "interaction" polymer (Figures 10A, 10B) indicate that the combination of the two polymer types provides significantly better resolution than what would be expected from either polymer effect alone. A third feature of the invention, as it applies to electrophoresis of nucleic acids, is the ability to resolve oligonucleotide analogs, and in particular, phosphorylated and non-phosphorylated oligonucleotide analogs. The method is illustrated in Example 6, which describes electrophoretic separations of a ladder of single-strand 12mer to Iδmer oligonucleotides, resolved on a matrix support formed of 10 weight percent HEC, as detailed in Example 5.

Figure 13A is an electropherogram of the phosphory- lated oligomer ladder, showing the baseline resolution of the seven major oligomer peaks. Also seen in the elec¬ tropherogram are seven minor peaks which are "offset" from the major peaks by migration distance corresponding to about 1.5 nucleotide units. An electropherogram of the non-phosphorylated oligomers is seen in Figure 13B, showing seven clearly resolved peaks. When a mixture of phosphorylated and non-phosphorylated ladders are mixed, the electropherogram seen in Figure 13C is obtained. A comparison of the three figures indicates that the two fastest moving peaks (at the left) are the phosphorylated 12mer and 13mer. Thereafter the non-phosphorylated and phosphorylated peaks alternate, with the two slowest peaks (at the right) being the non-phosphorylated 17mer and Iδmer. The results demonstrate the ability of the electro¬ phoresis method to separate phosphorylated oligomers from their non-phosphorylated analogs.

III. Isoelectric Focusing Method

The polymer matrix of the invention is also useful for capillary or preparative-tube IEF, by providing a stable matrix which can be expelled in intact form from the tube after separation of sample components on the pH gradient.

The supported matrix used in carrying out the IEF method is prepared according to the general guidelines above, substituting a standard IEF ampholyte solution for the electrolyte solution used in an electrophoresis matrix. Ampholyte solutions for producing a range of pH gradients, on equilibration in an electric field, are well known.

The type, molecular weight, and concentration of polymer are less critical than in electrophoretic separa- tion, since the polymer matrix does not function as a separation medium. Rather, the polymer composition is selected to (a) minimize band spreading after the elec¬ tric field is removed, and (b) to facilitate removal of the matrix from the tube after equilibrium is reached. Typically these objective are met in a relatively high matrix viscosities. However, for separation of high molecular weight proteins and nucleic acids, the vis¬ cosity must be kept low enough to allow free migration of the molecular species in the gel. The electrophoresis system like the one illustrated in Figure 4 is suitable for use in IEF on a capillary- tube supported matrix. Sample loading, voltage settings, and run times are carried out conventionally. Typically the sample contains proteins or peptides which are readi- ly separable, on the basis of different isoelectric points, on the selected pH gradient. After equilibrium is reached, the tube may be removed from the system and scanned. Alternatively, and according to an important advantage of the method, the matrix can be expelled in intact form from the tube, for analysis of separated bands in the matrix by one of the methods described with reference to Figure 5.

The method enhances the resolution of bands which can be achieved by IEF, since band spreading and wall distortion effects which smear separated bands when a low-viscosity medium is drawn from the tube are eliminat¬ ed. At the same time, the matrix provides the advantages of a low-viscosity medium in that the pH gradient can be expelled from the tube readily for analysis of the sepa¬ rated molecular components. The expelled matrix also allows for analysis by autoradiography or blotting tech¬ niques, as described with reference to Figure 5.

The following examples illustrate methods of prepar¬ ing supported matrices, and methods of fractionating proteins and nucleic acids, in accordance with the inven¬ tion. The examples are intended to illustrate, but in no way limit, the scope of the invention.

Example 1 Preparation of Capillary-Tube Matrices A. Polyethylene Oxide Matrix

High molecular weight linear polyethylene oxide (PEO) was dissolved into viscoelastic gels by the follow¬ ing technique. Two grades of PEO powder were obtained from Union Carbide Corporation (Danbury, CT) , WSR-1105 (MW=900,000) and WSR 303 (MW=7,000,000) . These powders were sintered into 2.0 x 0.1 cm slabs in a press. The slabs were then weighed and placed into a volume of Tris- Borate buffer containing 10 mM Tris-borate, pH 8.3, 5 mM NaCl, and 0.1 mM EDTA, (TB buffer), in an amount suffi¬ cient to produce a final polymer concentration of 11.2 weight percent. The buffer solution was obtained from Sigma Chemical (St. Louis, MO) .

The polymer solution was allowed to come to equili¬ brium for 24 hours. The lower molecular weight polymer (WΞR1150) formed a polymer matrix with a greaselike consistency, and a the higher molecular weight polymer formed a matrix with a soft elastic gel.

B. Hydroxyethyl Cellulose Matrix

High molecular weight linear hydroxyethyl cellulose was obtained from Union Carbide Corporation (Danbury, CT) . Five grades were used in studies on the effect of polymer molecular weight. The molecular weights are quantified by solution viscosity by Union Carbide Cor¬ poration and are listed below

The viscosities were determined on a Brookfield viscome¬ ter at- 25°C. and are reported in centipoise. The various grades were dissolved in TB buffer. The polymer was dissolved in the buffer, at selected weight percentages, by vortexing the buffer and adding the pow¬ dered polymer rapidly. Within five minutes the solution viscosity increased to a high-viscosity state. The mixture was allowed to sit 24 hours at room temperature to insure complete dissolution.

C. Polyacrylamide Matrix

Acrylamide monomer (Bio-Rad Inc., Richmond, CA) was prepared as a 30% solution in deionized water. This was then combined with a lOx buffer concentrate (Tris- Glycine-SDS, #832-7603, Kodak Inc., Rochester, NY) and water to obtain various monomer concentrations. The acrylamide was polymerized by adding ammonium persulphate and tetraethylene diamine. Once the catalysts were added the mixture was placed under nitrogen and allowed to cure for two hours at 25° C.

D. Matrix Pumping Procedure Each of the above polymer matrix material was pumped into a fused silica capillary tube, typically a #TSP75/350 capillary tube obtained from Poly icro Tech¬ nologies, Inc. (Phoenix, AZ) . This tube has a 75 micron ID and was cut to 42 cm. Injection of the polymer matrix solutions into the into capillary tubes was done as follows. One gram of the finished matrix was placed with a spatula into a 5cc syringe and then injected into a 5 cc stainless steel tube with a 0.125" OD and a 0.040" ID. The stainless steel tube was coupled to Model 140A high-pressure pump

(from Applied Biosystems) at one tube end, and to the above capillary tube at the other end. Connections were with low dead volume fittings, such as connectors Part

#U437 from Upchurch Scientific Inc. (Seattle, WA) . The pump was filled with water and used to apply a pressure of between about 2,500- 7500 psi to the in order to pump the matrix into the tube. The pressure was typically set to achieve a pumping rate of about 250 μL per minute. Example 2 Fractionation of Proteins on a Polyacrylamide Matrix A mixture of proteins containing lactalbumin, molecular weight 14,200 daltons (O.lmg/ml), lactalglobulin, molecular weight 17,500 daltons (O.lmg/ml), trypsin inhibitor, molecular weight 20,000 daltons (O.lmg/ml), and carbonic anhydrase, molecular weight 29,000 daltons (O.lmg/ml) was prepared in the above TB buffer.

Capillary electrophoresis was carried out using an ABI Model 270 capillary electrophoresis system. The system includes a built-in high-voltage DC power supply capable of voltage settings up to 30 KV. The polyacrylamide matrix used was the 42 cm capillary tube prepared as in Example 1C. The two reservoirs ware filled with TB buffer. The protein sample was electrophoretically drawn into the tube for 1.5 seconds at 5 kV. The electrophoretic system was run at a voltage setting of about 9 kV (about 200 V/cm) through the run. UV detection was with a Kratos 757 UV detector designed for capillary tube detection. The detector output signal was integrated and plotted on an HP Model 3396A integrator/plotter.

The electropherogram obtained is shown in Figures 6A-6C for the 4.5, 7.5 and 10 weight percent polymers, respe tively. The numbers above the major peaks (A-D) are electrophoresis times, in minutes. Total run time was about 30 minutes. As decribed above, the method effectively separated proteins in the range of 14 to 30 kilodaltons at 7.5 and 10 weight percent polymer. Example 3 Fractionation of Nucleic Acid Fragments on a PEO Matrix

A. Effect of Polymer Molecular Weight

Capillary tubes with supported PEO matrices, at two different PEO molecular weights, were prepared as de¬ scribed in Example 1A. An oligonucleotide ladder ( 0mer to 60mer) obtained from Pharmacia (Bromma, Sweden) was prepared in TB buffer and introduced into each of the tubes, at the cathodic end, in the capillary electro- phoresis system described in Example 2. The applied voltage was nine kilovolts and the detection wavelength was 260 nm. The electropherograms from the 900 kilodal¬ ton and 7,000 kilodaltons PEO polymers are shown in Figures 7A and 7B, respectively. As seen, the nucleic acid fragments are significantly better resolved in the higher molecular weight polymer matrix.

B. Effect of Polymer Concentration

Capillary tubes with supported PEO matrices, were prepared using the above WSR3030 PEO polymer (7 million average molecular weight) , at polymer concentrations of 6% and 15%, substantially as described in Example 1A. A 40mer-60mer oligonucleotide ladder was fractionated on the supported matrices, substantially as described in Part A above. The applied voltage was nine kilovolts and the detection wavelength was 260 nm. The electrophero¬ grams for the 6% and 15% polymer solutions are shown in Figures 8A and 8B, respectively. As seen, the nucleic acid fragments are significantly better resolved in the more concentrated polymer matrix. Example 4 Fractionation of Nucleic Acid Fragments on an HEC Matrix A. Effect of Polymer Molecular Weight

Capillary tubes with supported HEC matrices prepared from the QP-40, QP-300, and QP-4400 polymers described in Example IB, to a final polymer concentration of 10 weight percent in TB buffer, as described in Example 1. The oligonucleotide ladder (40mer to 60mer) from Example 3 was introduced into each of the tubes, at the cathodic end, in the capillary electrophoresis system described in Example 2. The applied voltage was nine kilovolts and the detection wavelength was 260 nm. The electrophero¬ grams for the QP-40, QP-300, and QP-4400 HEC polymers are shown in Figures 9A, 9B, and 9C, respectively. It can be seen in the electropherograms that all three polymer matrices gave similar resolution at the same polymer concentration, despite the different polymer molecular weights.

B Effect of Polymer Concentration Capillary tubes with supported HEC matrices prepared from the QP-300 polymer described in Example IB, at polymer concentrations in TB buffer of 3, 10, 15 and 25 weight percent QP-300. An oligonucleotide ladder (20 mer to 40mer) obtained from Pharmacia was introduced into each of the tubes, at the cathodic end, in the capillary electrophoresis system described in Example 2. The applied voltage was nine kilovolts and the detection wavelength was 260 nm. The electropherograms for the 3%, 10%, 15%, and 25% polymer matrices are seen in Figures 10A, 10B, 10C, and 10D, respectively.

As seen, resolution can be enhanced to obtain a well- defined separation between nucleic acid peaks at increas¬ ing polymer concentrations. When the migration rate of the 30mer (the central peak) is plotted as a function of polymer concentration, the curve shown in Figure 11 is obtained. The plot shows that the migration of small nucleic acid species is strongly dependent on polymer concentratio .

Example 5 Separation of Phosphorylated and Non-Phosphorylated Nucleic Acid Fragments An HEC gel was formulated as described in Example IB using QP4400H at a concentration of ten percent in TB buffer, and injected into a capillary tube. A set of polyadenylic acids (12-18 bases) was fractionated on the matrix under the electrophoretic conditions detailed in Example 3, with the results shown in Figure 11A. As seen from the figure, all of the phosphorylated nucleotides are baseline resolved, but with shadow peaks.

When the same set of nucleotides in non phosphorylated form were fractionated under the same conditions, the electropherogram shown in Figure 12B was obtained. The electropherogram is distinguished by the lack of shadow peaks, but otherwise resembles the Figure 12A electro¬ pherogram.

When a mixture of roughly equal amounts of phosphory¬ lated and non-phosphorylated nucleotides were fractiona- ted under the same conditions, the electropherogram seen in Figure 12C was obtained. It is clear from a compari¬ son of the three Figure 12 electropherograms that (a) the shadow peaks seen in Figure 12A are non-phosphorylated contaminants and (b) the electrophoretic system is effec- tive to resolve phosphorylated and non-phosphorylated nucleotide analogs. Example 6 Mixed Polymer Matrix A mixed polymer solution containing both HEC and poly¬ acrylamide was prepared from the following components: 0.5g QP3L

0.5ml TBE (5x cone.)

4.5ml Deionized Water

3.5ml 30% Acrylamide in Deionized Water

The ingredients were mixed well and then catalyst was added as described in Example 1C to polymerize the acryl¬ amide. The final percentage of HEC was about 5.5 by weight; of polyacrylamide, about 11.5 by weight. After curing the polymer mixture was injected into a capillary tube ar.d a separation of pd(A)40 to 60 was performed. The electropherogram obtained is shown in Figure 13. The separation between peaks was comparable to that achieved at an HEC polymer concentration of about 25 weight per¬ cent (Figure 8D) , and much greater than either of the two components would be expected to give, indicating an additive separation mechanism.

Although the invention has been described with respect to exemplary supported matrices and methods of use, it will be appreciated that various modifications and chan¬ ges may be made without departing from the invention.

Claims

IT IS CLAIMED:

1. A method of preparing a supported, substantially homogeneous matrix for use in electric field-induced separation of molecular components in a sample compri¬ sing: forming a viscoelastic, flowable aqueous polymer matrix characterized by:

(i) a water-soluble, substantially non-crosslinked polymer, having a molecular weight of at least about 10,000 daltons,

(ii) a viscosity of at least about 5,000 centipoise, and pumping the polymer matrix, in its high-viscosity state, into an elongate separation chamber, to fill the chamber uniformly with the matrix.

2. The method of claim 1, wherein the separation chamber is contained within a capillary tube which has a plug at one end to prevent flow of matrix material from that tube end.

3. The method of claim 1, wherein the separation chamber is contained within a capillary tube having a substantially uncharged wall surface.

4. The method of claim 1, wherein the polymer is a linear polymer having a average molecular weight of at least about 200,000 daltons, and viscosity of at least about 100,000 centipoise. 5. The method of claim 4, wherein the viscosity of the matrix, as measured by as measured by the pressure needed to pump the matrix through a 30 cm length of 75 μ capil¬ lary tube, at a flow rate of 100 μl/ in, is at least about 3000 psi.

6. The method of claim 4, wherein the concentration of polymer in the matrix is at least about 10 percent by weight.

7. The method of claim 1, wherein the polymer is selected from the group consisting of polyoxides, poly- ethers, polyethylene imine, polyacrylic acid, polyacryl¬ amide, polymethacrylamide, polymethacrylic acid, poly- vinylacetate, polyvinylpyrrolidone, polyvinyloxazolidone, naturals gums, polyvinylalcohol, and water-soluble cel¬ lulose compounds, including hydroxymethylcellulose and hydroxyethylcellulose, and co-polymers and blends of the polymers.

8. The method of claim 1, for preparing' a matrix for use in separating molecular components by electrophore¬ sis, wherein the polymer matrix contains an electrolyte.

9. The method of claim 8, for use in separating poly- peptides wherein the polymer is selected from the group consisting of polyethylene oxide, polyacrylamide, and polymethacrylamide.

10. The method of claim 8, for use in separating nucleic acids, wherein the polymer is a hydroxylated polymer selected from the group consisting of water- soluble hydroxylated cellulose compounds, pectin, and polyvinyl alcohol. 11. The method of claim 8, for use in separating nucleic acid components wherein the polymer includes a mixture of a polymer selected from the group consisting of polyethylene oxide, polyacrylamide, and polymethacryl- amide, and a polymer selected from the group consisting of a water-soluble, hydroxylated cellulose compounds, amylose, pectin, and polyvinyl alcohol.

12. The method of claim 1, ^' for preparing a matrix for use in isoelectric focusing, wherein the polymer matrix contains ampholyte species effective to establish a substantially continuous pH gradient over a selected pH range.

13. A supported matrix for use in electric field- induced separation of molecular components in a sample comprising: a support containing an elongate separation chamber, and contained within said chamber, and filling the chamber substantially uniformly, an aqueous viscoelastic matrix characterized by:

(i) a water-soluble, substantially non-crosslinked polymer having a molecular weight of at least about 10,000 daltons, and

(ii) a viscosity of at least about 5,000 centipoise.

14. The matrix of claim 13, wherein the separation chamber is contained within a capillary tube which has a plug at one end to prevent flow of matrix material from that tube end. 15. The matrix of claim 14, wherein the separation chamber is contained within a capillary tube having a substantially uncharged wall surface.

16. The matrix of claim 13, wherein the polymer is a linear polymer having a average molecular weight of at least about 200,000 daltons, and a viscosity of at least about 100,000 centipoise.

17. The matrix of claim 16, wherein the viscosity of the matrix, as measured by as measured by the pressure needed to pump the matrix through a 30 cm length of 75 μ capillary tube, at a flow rate of 100 μl/min, at room temperature, is at least about 3000 psi.

18. The method of claim 16, wherein the concentration of polymer in the matrix is at least about 10 percent by weight.

19. The matrix of claim 18, wherein the polymer is selected from the group consisting of polyoxides,' poly- ethers, polyethylene imine, polyacrylic acid, polyacryl¬ amide, polymethacrylamide, polymethacrylic acid, poly- vinylacetate, polyvinylpyrrolidone, polyvinyloxazolidone, naturals gums, polyvinylalcohol, and water-soluble cel¬ lulose compounds, including hydroxymethylcellulose and hydroxyethylcellulose, and co-polymers and blends of these polymers.

20. The matrix of claim 19, for use in separating polypeptides, wherein the polymer is selected from the group consisting of polyethylene oxide, polyacrylamide, and polymethacrylamide. 21. The matrix of claim 19 for use in separating nucleic acids, wherein the polymer is a hydroxylated polymer selected from the group consisting of water- soluble, hydroxylated cellulose derivatives, amylose, pectin, and polyvinyl alcohol.

22. The method of claim 21 for use in separating nucleic acids wherein the polymer includes a mixture of a polymer selected from the group consisting of polyethy- lene oxide, polyacrylamide, and polymethacrylamide, and a polymer selected from the group consisting of water- soluble, hydroxylated cellulose compounds, amylose, pectin, and polyvinyl alcohol.

23. The matrix of claim 19, for use in electrophoretic separation of related biopolymer components which differ from one another by a single charge or by a single number of subunits, wherein the polymer matrix has a molecular weight of at least about 1 million daltons, and a vis- cosity, as measured by as measured by the pressure needed to pump the matrix through a 30 cm length of 75 μ capil¬ lary tube, at a flow rate of 100 μl/min, at room tempera¬ ture, is at least about 3000 psi.

24. The matrix of claim 11, for use in isoelectric focusing, wherein the polymer matrix contains ampholyte species effective to establish a substantially continuous pH gradient over a selected pH range.

25. A method of electrophoretic separation of molecu¬ lar components in a sample comprising: adding said sample to one end region of a support defining an elongate separation chamber which is filled, substantially uniformly, with an aqueous electrolyte- containing polymer matrix characterized by:

(i) a water-soluble, substantially non-crosslinked polymer having a molecular weight of at least about 100,000 daltons, and

(ii) a viscosity of at least about 5,000 centipoise, and separating said components by applying an electric field across opposite end regions of the chamber, until a desired degree of electrophoretic separation of the components is achieved.

26. The method of claim 25, wherein the separation chamber is contained within a capillary tube having a charged wall surface, and tube has a plug at one end which prevents flow of the matrix through that tube end.

27. The method of claim 25, wherein the elongated support is a capillary tube having an uncharged wall surface.

28. The method of claim 25, wherein the elongate support is an electrophoresis tube, and said method further comprises, following said separating, forcing the solution out of the tube and monitoring the solution for the presence of discrete sample component bands as the sample is forced from the tube.

29. The method of claim 25, which further comprises, following said separating, removing selected regions of the solution containing separated sample components, and isolating separated sample components from removed re¬ gions of the gel. 30. The method of claim 29, wherein said isolating includes liquefying the solution to form a low-viscosity liquid, and isolating the sample component from the polymer in such liquid.

31. The method of claim 30, wherein said liquefying is carried out by diluting the solution in an aqueous medi¬ um.

32. The method of claim 30, wherein said liquefying is carried out by heating.

33. The method of claim 25, wherein the sample com¬ ponents are proteins, and said polymer is selected from the group consisting of polyethylene oxide, polymeth¬ acrylamide, and polyacrylamide.

34. The method of claim 33, wherein the polymer has a molecular weight of at least about 1 million daltons, and a viscosity, as measured by as measured by the pressure needed to pump the matrix through a 30 cm length of 75 μ capillary tube, at a flow rate of 100 μl/min, at room temperature, is at least about 3000 psi.

35. The method of claim 34, wherein the polymer is polyethylene oxide having a molecular weight of about 3 million and a concentration of about 10 weight percent.

37. The method of claim 25, wherein the sample com- ponents are oligonucleotides, and the polymer is a water- soluble hydroxylated cellulose compound. 38. The method of claim 37, wherein the polymer is selected from the polymer includes a polymer selected from the group consisting of polyethylene oxide, poly¬ acrylamide, and polymethacrylamide, and a polymer selec- ted from the group consisting of a water-soluble hydroxy¬ lated cellulose compound.

39. The method of claim 37, for use in separating a phosphorylated low molecular weight nucleic acid com- ponent from its non-phosphorylated analog.

40. The method of claim 25, for use in separating a phosphorylated low molecular weight nucleic acid com¬ ponent from its non-phosphorylated analog, wherein the polymer has a molecular weight of at least about 1 mil¬ lion daltons, and the concentration of polymer is at least about 10 weight percent.

41. A method of fractionating a sample of nucleic acid components in the size range less than about 100 bases, comprising adding said sample to one end region of a support defining an elongate separation chamber which is filled, substantially uniformly, with an aqueous electrolyte- containing polymer matrix characterized by:

(i) a mixture of polymers, one of which is selected from selected from the group consisting of polyethylene oxide, polyacrylamide, and polymethacrylamide, and ano¬ ther of which is selected from the group consisting of a water-soluble hydroxylated cellulose compound,

(ii) polymer molecular weights of molecular weight of at least about 100,000 daltons, and

(iii) a viscosity of at least about 100,000 centi¬ poise, and separating nucleic acid components by applying an elec¬ tric field across opposite end regions of the chamber, until a desired degree of electrophoretic separation of the components is achieved.

42. The method of claim 41, wherein the polymer mix¬ ture contains at least about 5 weight percent of each of the polymers in the mixture.

43. The method of claim 42, for separation of oligo¬ nucleotides having sizes less than about 50 bases, where¬ in the polymer mixture has a viscosity, as measured by as measured by the pressure needed to pump the matrix through a 30 cm length of 75 μ capillary tube, at a flow rate of 100 μl/min, at room temperature, is at least about 3000 psi.

44. A method of separating molecular components in a sample by isoelectric focusing comprising: adding the sample to a matrix supported in an elongate separation chamber which is filled, substantially uni¬ formly, with an aqueous ampholyte-containing polymer matrix characterized by:

(i) a water-soluble, substantially non-crossliήked polymer having a molecular weight of at least about 10,000 daltons, and

(ii) a viscosity of at least about 5,000 centipoise and separating said components by applying an electric field across opposite end regions of the chamber, until the ampholyte has equilibrated to produce a selected pH gradient across the end regions of the chamber, and the molecular components have migrated to their isoelectric points within the chamber. 45. The method of claim 44, wherein the elongate support is a tube, and said method further comprises, following said separating, forcing the solution out of the tube and monitoring the solution for the presence of discrete sample component bands as the sample is forced from the tube.

46. The method of claim 44, which further comprises, following said separating, removing selected regions of the solution containing separated sample components, and isolating separated sample components from removed re¬ gions of the gel.

47. The method of claim 46, wherein said isolating includes liquefying the solution to form a low-viscosity liquid, and isolating the sample component from the polymer in such liquid.

48. The method of claim 47, wherein said liquefying is carried out by diluting the solution in an aqueous medi¬ um.

49. The method of claim 47, wherein said liquefying is carried out by heating.