WO1997026531A1

WO1997026531A1 - Fractionation process

Info

Publication number: WO1997026531A1
Application number: PCT/SE1997/000074
Authority: WO
Inventors: Johan Roeraade; Mårten STJERNSTRÖM
Original assignee: Johan Roeraade; Stjernstroem Maarten
Priority date: 1996-01-19
Filing date: 1997-01-17
Publication date: 1997-07-24
Also published as: SE9600213D0; SE9600213L; SE506461C2

Abstract

A process for the size-fractionation of high-molecular-weight molecules in a mixture thereof using an electroosmotic flow of said molecules contained in a low ionic strength buffer, said process being performed in a capillary tube. The invention is based on the concept of using a capillary tube having an inner diameter not exceeding the cross-sectional size of the largest molecules by a factor greater than about 10 (ten).

Description

FRACTIONATION PROCESS

The present invention relates to a process for the size-fractionation of high-molecular-weight molecules using an electroosmotic flow of said molecules contained in a low ionic strength buffer. Such process is performed in a capillary tube to the effect of obtaining separation of the molecules according to size.

Molecules with a high molecular weight can be size- fractionated by a number of instrumental separation tech¬ niques such as size-exclusion chromatography (SEC) (Potschka, M., Macromolecules 24 (91) 5023), capillary gel electrophoresis (Cohen, A.S.,; Karger, B.L., J. Chro- ma togr. 397 (87)- 409) or field flow fractionation (Giddings, J.C, Anal . Chem . 67 (95) 592A) . For example molecular weight determinations of polymers are commonly made by the SEC methods. There is a considerable need for separation of deoxyribonucleic acid (DNA) molecules in a size range above 30 kbp. This has not been possible by using free solution electrophoresis or gel electrophore¬ sis, because the sieving action is lost for large olecu- les. However, this can be accomplished by a procedure called pulse-field electrophoresis (Schwartz, D.C; Can¬ tor, C.R., Cell . 37 (84) 67) in which the resolving power of ordinary gel electrophoresis is extended. The mecha¬ nism is based on a repeated change of direction of the electric field. The larger the DNA molecules, the longer time is needed by the molelcules to adjust their shape to the applied field and hence their change of movement is reduced. However, the procedure is associated with a num¬ ber of problems, such as time consuming analyses, compli- cated instrumentation (complex power supplies), difficul¬ ties in quantification, low reproducibility, wasteful use of sample etc.

Hydrodynamic chromatography (HDC) is another tech¬ nique that is used for size fractionation. It can be car- ried out in a packed column or in an open tubular mic- rocapillary (Tijssen, R. ; Bos, J. ; van Kreveld, M.E., Anal . Chem . 58 (86) 3036) using a single solvent for mass transport. The separation is due to the presence of a pressure-induced Poisseuille flow or Taylor flow velocity profile within the tubing in which large particles based on their size are excluded from the slowest moving streamlines closest to the wall. This causes them to be eluted before the smaller analytes. Electrostatic ef- fects, especially pronounced in low ionic strength buf¬ fers, can also influence this separation process (DosRamos, G.J.; Silebi, C.A., J. Colloid In terface Sci . 133 (89) 302) by an electrical double layer repulsion between charged surfaces and analytes. It is well known that the electrokinetic bulk flow behavior in a capillary is influenced by characteristics of the electrical double-layer that resides in close proximity to the capillary inner surface. In capillary electrophoresis separations are usually performed in tu- bes having a radius of 25-50 μm and in buffers with salt concentrations in the mM range. When a potential is applied over such a capillary, the observed electroosmo¬ tic flow profile is essentially flat (Jorgensson, J.W.; Lukacs, K.D., J. Chroma togr. 218 (81) 209) and is super- imposed on the electrophoretic separation. However, theo¬ retical calculations have shown that it is possible to obtain an electroosmotic flow velocity profile similar to that of pressure induced Poisseuille flow in tubes with small inner diameters and liquid media of low ionic strengths (Rice, CL.; Whitehead, R. , J. Phys . Chem . 69 (65) 4017) . According to these calculations, a near- parabolic flow profile is generated, when the radially opposite electrical double-layers begin to overlap.

One object of the present invention is to provide new techniques, whereby the problems associated with the known art as discussed above are eliminated or at least essentially reduced. Another object of the present invention is to provi¬ de a process for the size fractionation of molecules of high molecular weight, said process being based on utili¬ zation of an electroosmotic flow of said molecules con- tained in a buffer, preferably of low ionic strength, such as less than about lOmM. The lower limit is determi¬ ned by the requirement that the electrical double-layer is just established.

Yet another object of the invention is to provide a process for such size-fractionation of a spectrum of mo¬ lecules using very narrow capillary tubes with inner dia¬ meters of less than about 1 μm.

Still another object of the invention is to provide for separation of high molecular weight molecules, the larger ones of which have cross-sectional dimensions ap¬ proaching the inner dimensions of the capillary tubes used.

Still another object of the invention is to provide a process, whereby an acceptable flow in a capillary tube can be generated which could not be obtained by resorting to hydrodynamic separation using an external inlet pres¬ sure.

For these and other objects that will be understood by the following disclosure the invention provides a pro- cess for the size-fractionation of molecules of high mo¬ lecular weight using an electroosmotic flow of said mo¬ lecules contained in a low ionic strength buffer, said process being performed in a capillary tube. The inven¬ tion is based on the concept of using a capillary tube that has an inner diameter which does not exceed the cross-sectional size of the largest molecules by a factor greater than about 10.

In the instant disclosure the expression "cross- sectional size" means the gross dimension in a plane nor- mal to the longitudinal direction of the capillary tube under operating conditions.

A preferred operative range for such factor is from about 2 to about 10. Even if this range is only a prefer¬ red one a factor less than 2 may reduce the free movabi¬ lity of larger molecules, whereas at factors exceeding 10 the fractionation efficiency tends to decrease. Expressed in absolute numbers it is preferred that the inner diameter of the capillary tubes used is prefe¬ rably less than about 1 μm and may especially be about 500 nm or less.

The lower limit of the inner diameter of the capil- lary tubes used in the fractionation process is not par¬ ticularly critical and can be substantially less than 500 nm. In practice the lower limit is perhaps restricted by the possibility of providing very small diameters, yet obtaining useful capillary tubes. The tube length can vary within very broad limits and a practical range may be from about 10 cm to about 10 m, a particularly preferred range being from about 20 cm to about 2 m.

In regard to the fractionation of molecules carrying an electric charge it is, according to one embodiment of the invention using charged molecules, necessary that the molecules have a net charge sign which is the same as that of the capillary tube wall. By such arrangement the¬ re is obtained a fractionation whereby small molecules travel faster than large ones.

On the other hand, when the molecules to be separa¬ ted are essentially uncharged, the process results in a situation where large molecules travel faster than small ones. The molecules to be separated by using the process of the present invention may have a molecular size vary¬ ing within very broad limits. Thus, the molecular weight may vary from the order of 10⁶, and may reach very high values, such as several tenths of millions and even up to several billions. Any high molelcular molecules can be separated resorting to the techniques of the present in¬ vention and as examples there may be mentioned proteins, DNA, RNA, polysaccharides, organic polymers, etc. The present invention is particularly suited for fractiona^¬ tion of molecules belonging to the biotechnological area, such as proteins, DNA:s and RNA:s. Accordingly, the present invention is based on the new concept of resorting to the use of an electroosmotic flow, where the flow profile is utilized for size frac¬ tionation. This is accomplished by the use of very narrow capillary tubes within a diameter dimension in the nano- meter range. The target analytes are molecules with mo¬ lecular sizes almost approaching the inner diameter of the capillary tube. It is preferred to use a low ionic strength buffer, and an electric field is applied across the capillary tube in order to propel the liquid separa- tion medium through the tube. It is again emphasized that it is not practicable to use an external inlet pressure to generate the required flow in a capillary of such dia- mension. In other words, pressure-driven hydrodynamic se¬ paration is not practically useful for providing the de- sired fractionation.

The present invention will now be further illustra¬ ted more in detail with reference to the appended draw¬ ings, wherein:

Fig. 1 is a diagrammatic illustration of the flow conditions in a capillary tube, where the molecules to be fractionated have a charge of the same sign as that of the surrounding capillary inner walls;

Fig. 2 is a corresponding illustration, where neu¬ tral or uncharged molecules are subjected to fractiona- tion;

Fig. 3A is a diagram showing separation in accor¬ dance with the invention at 4 different ionic strengths in the buffer solutions used;

Fig. 3B shows the sizes of the molecular fractions of the mixture used given in thousands of basepairs (kbp);

Fig. 4A shows a diagram on the fractionation of a mixture of molecules using the invention; and

Fig. 4B shows the sizes of the molecules of said mixture given in kbp.

Two main cases for the new separation system can be distinguished, which rely on different mechanisms.

Figure 1 shows the first case, where the analytes have a net charge with the same charge sign as the capil¬ lary wall. A size separation between the smaller molecu¬ les (1) and the larger molecules (2) of the analyte is due to the velocity difference which the two components experience by occupying a different fraction of the cross sectional area of the capillary tube (4) . During the transport of the analytes by electroosmotic flow (3), they are at the same time forced towards the center of the capillary by an electrostatic repellation exerted by the charged capillary walls (fat arrows, Fig. 1) . The ve¬ locity of a molecule can by a first approximation be cal¬ culated by integrating its velocity components in the different streamlines over its cross section and dividing the obtained value by its cross-sectional area. The lar¬ ger components (2) will be more affected by the slower moving streamlines closer to the capillary walls and will therefore be retarded compared to the smaller components. The electrostatic forces, which repel negatively charged molecules from the capillary wall and focus them in the potential well in the center of the capillary also redu¬ ces the possibility for detrimental sample adsorption.

In the second case uncharged molecules can be sepa¬ rated by the same mechanism as in hydrodynamic chromato- graphy as illustrated in Figure 2. Here, the elution or¬ der is reversed compared to the first case. Since the mo¬ lecules do not experience an electrostatic repellation from the capillary wall, they will distribute statisti¬ cally over the entire cross section of the capillary, the driving force being molecular diffusion. However, molecu¬ les with a radius in the size range of the capillary in¬ ner diameter will be fractionated according to size, sin- ce the entire cross section of the smaller molecules can also reside in the slower moving regions near the wall, which is not possible for the larger molecules. Thus, in contrast to the situation in the first case, they will be transported at a slower speed than the larger molecules. This separation principle is basically similar to classi¬ cal hydrodynamic chromatography. The important differen¬ ce, however, which is embodied by the present invention, is the way the flow is generated to accomplish the hydro- dynamic separation.

The first case illustrated in Figure 1, where char¬ ged molecules are separated, is particularly directed towards free flow size separations of large biomolecules, such as DNA, in a pulseless electric field. However, the invention is not limited to this particular application. The invention will now be further illustrated by ex¬ amples with reference to Figures 3A, 3B, 4A and 4B, but it is to be noted that these examples are not construed to be restricting the scope of the invention.

EXAMPLE 1

In this example a mixture of lamda-DNA fragments di¬ gested using Hind III in a concentration of 0.7 ng/μl is used. The composition of this mixture is given in Figure 3B, the respective numbers referring to kilo-basepairs. The size fractionation in accordance with the pre¬ sent invention is carried out using capillary tubes ha¬ ving an outer diameter of 40 μm and an inner diameter of 0.5 μm. The active length of the capillary tube is 60 cm. Initially, the capillary tubes are filled with a borate buffer, the ionic strength of which is varied from 50 mM down to 50 μM as shown in Figure 3A. Each capillary con¬ taining initally the borate buffer is then filled at one end thereof with the DNA-mixture by electrokinetic injec- tion from one end thereof for 8 sec at an applied voltage of 4700 V.

The capillary tubes thus loaded are then subjected to size fractionation of contained analytes in accordance with the present invention using a separation voltage across the tube of +10 000 V, a field strength of about 170 V/cm, and the result of fractionation is determined by the use of laser-induced fluorescence detection. For the detection the DNA fragments are prior to loading first stained with a bis-intercalator YOYO (Molecular Probes, Eugene OR, USA) . An Ar+ laser induced confocal fluorescence detector (488 nm exitation) is employed as a detection device.

Figure 3 shows the results of the fractionation per¬ formed in accordance with the present invention, and it can be seen that lowering the ionic strength of the buf¬ fer used results in improved separation. Thus, at an io- nic strength of 50 μM efficient separation of the DNA mo¬ lecules can be seen, a separation which could not be achieved by resorting to hydrodynamic separation of a conventional nature since a capillary tube of an inner diameter of about 0.5 μm could not by exerting external pressure provide for a useful flow through the capillary to give a practically useful separation.

EXAMPLE 2

An experiment similar to that described in Example 1 above is performed using as a DNA fragment mixture a gui¬ deline ΘX-174 ladder, having a composition according to Figure 4B, again given as kbp. In this example an ionic strength of the buffer of 50 μM is used and the injection of the sample is performed for 25 sec at an applied vol- tage of 4000 V. The separation proper is performed at an applied voltage of 2000 V.

Figure 4A shows the result of the separation made in accordance with the present invention, and as can be seen from the diagram a very efficient separation or fractio- nation of the 25 components of the mixture is obtained. In fact, the individual peaks seen in Figure 4A corre¬ spond to single DNA molecules. Again, this separation could not have been obtained resorting to hydrodynamic separation or other conventional separation techniques.

The size, orientation and conformation of large DNA molecules in aqueous solutions are dependent on a number of factors. It is well known that the DNA coil swells and increases in size when the salt concentration decreases. Under such conditions the thickness of the electrical double layer near the capillary walls will also increase. The spherical coil of DNA can be stretched by the appli- cation of strong electric field and/or by being forced through a narrow passage. Furthermore, the magnitude of the shape change in the electric field is also dependent on the salt concentration (Diekmann, S; Porschke, D., Biophys . Chem. 16 (82) 261) . Such factors can be taken in- to consideration, when optimizing separations according to the principles in the described invention.

The optimal choice of inner diameter of the submic¬ ron separation capillaries, to obtain optimal separation performance is related to the geometrical radius of the molecules to be separated, as shown in both Fig. 1 and

Fig. 2. However, it is not always straightforward to pre¬ dict the molecular radius, by only knowing the molecular weight such as in dalton. Most often, the shape, size, conformation and radius of gyration of polymeric molecu- les are dependent on a number of factors, such as the surrounding medium (ionic concentration, solvation power, dielectric properties, etc.) . Thus, polymers of a parti¬ cular molecular weight can swell or stretch, depending on the medium. Also, the applied electrical field across the separation tube can change the shape of the analytes (this is particularly true for charged molecules like DNA) . For the charged analytes, the ion cloud (electrical double layer) , present around the analytes is an impor¬ tant factor, which influences the separation. Furthermo- re, the differential hydrodynamic force, exerted on the analytes during the separation can change the conforma¬ tion of the analytes (the molecules simply allongate (stretch) or deform into another conformation) . This can occur also for non-charged species.

When optimizing the separation for a particular type and size of analyte, it is therefore not only the capil- lary inner diameter, but also the above-mentioned fac¬ tors, which need to be taken into consideration.

An interesting possibility optional to this inven¬ tion, is to employ a pulsed electric field (cf. Schwarz et al., ibid.), instead of a continuous electric potenti- al across the separation column. This is extensively uti¬ lized e.g. for separation of large DNA fragments by gel electrophoresis. Basically, the DNA molecules stretch during application of the electric field. Under the sub¬ sequent period, where no electric field is present, a re- laxation of the DNA molecules will occur, where the mo¬ lecules will attempt to attain their original coiled con¬ formation. The relaxation procedure occurs in time domain and is dependent on molecular size. By repeating the pro¬ cedure and adjusting the frequency of the intermittent voltage application, the separation can be optimized for a given size range of molecules. The same procedure can be utilized in the present invention for improving sepa¬ ration of charged analytes. By applying a pulsed field across the submicron sized capillary separation column, a size selective separation will take place. The periodic conformation changes of the analytes enhance the action of the hydrodynamic effects as described in the second case.

The Taylor velocity profile will be influenced by a number of operational parameters. Of first importance is the inner diameter of the capillary, which has to be suf¬ ficiently small. Of great importance is also the ionic strength of the separation medium. The lower the ionic strength, the thicker the electrical double layer and the better the fractionation. Therefore the generation of the electroosmotic flow profile will be enhanced by reducing the ion concentration of the medium. For comparatively "large" capillaries, such as up to 1 μm, it is favourable to employ media with low ionic strength.

The thickness of the electric double layer (and the¬ reby the tendency to form a Taylor-type flow profile) can further be influenced by applying an external potential across the inner and outer wall of the capillary. The non-conductive capillary material (glass, fused silica) will act as a dielectricum, and the charge at the inner wall will be affected (compare the action of an electric condensor) . To accomplish this mode of operation, it is necessary to coat the outside of the capillary wall with a conductive material, e.g. a metal. A voltage is then applied between the electrolyte inside the capillary tube and the surrounding conductive material. Said voltage may be of the same order magnitude as the voltage applied for the separation and may also be pulsed (cf. Tung-Liang Huang et al., "Mechanistic Studies of Electroosmotic Con¬ trol at the Capillary-Solution Interface", Anal . Chem . 1993, 65, 2887-2893) . Although the present invention has been illustrated mainly with reference to the separation of DNA fragments or molecules in a mixture thereof it is important to note that the invention is equally applicable to other high molecular weight molecules. Therefore, the invention is to be construed to be restricted only by the wording of the appended claims.

Claims

1. In a process for the size-fractionation of high- molecular-weight molecules in a mixture thereof using an electroosmotic flow of said molecules contained in a low ionic strength buffer, said process being performed in a capillary tube, the improvement of using a capillary tube having an inner diameter not exceeding the cross-sectio¬ nal size of the largest molecules by a factor greater than about 10 (ten) .

2. A process according to claim 1, wherein said fac¬ tor is from about 2 (two) to about 10 (ten) .

3. A process according to claim 1 or 2, wherein said inner diameter is less than about 1 μm.

4. A process according to claim 3, wherein said dia¬ meter is less than about 500 nm.

5. A process according to any one of the preceding claims, wherein the tube length is from about 10 cm to about 10 m.

6. A process according to any one of the preceding claims, wherein the molecules to be separated have a net charge sign which is the same as that of the capillary tube wall, whereby small molecules travel faster than large ones.

7. A process according to any one of claims 1 to 5, wherein the molecules to be separated are essentially un¬ charged, whereby large molecules travel faster than small ones.

8. A process according to any one of the preceding claims, wherein the molecules to be separated are selec¬ ted from proteins, polysaccharides, organic polymers, DNA, RNA.

9. A process according to claim 8, wherein the mo¬ lecules to be separated are DNA molecules comprising up to about ten million kbp.

10. A process according to any one of the preceding claims, wherein a constant separation voltage is applied across the capillary tube.

11. A process according to any one of claims 1 to 9, wherein a pulsating electric separating field is applied across the capillary tube.

12. A process according to any one of the preceding claims, wherein a constant or pulsating voltage is app¬ lied between the electrolyte inside the capillary tube and a conductive coating on the outside of the capillary tube.