Application of Continuous Zone Electrophoresis to Preparative Separation of Proteins

Sunil Nath* Biochemical Engineering Research Centre, Indian Institute of Technology, Hauz Khas, New Delhi 110 016, India

Horst Schutte, Helmut Hustedt, and Wolf-Dieter Deckwer National Research Institute of Biotechnolog y, Mascheroder Weg 1, 3300 Braunschweig, Germany

Received October 13, 1992/Accepted April 27, 1993

A comprehensive study of the application of continu- ous zone electrophoresis to preparative separation of proteins in free solution is presented. First, the influ- ence of electric field strength, buffer residence time in the chamber, sample flow rate, and sample concen- tration on separation resolution and throughput were studied. Using multiple injections of sample into the electrophoresis chamber, a throughput of 500 mg protein/h was achieved for partially purified model proteins. Experi- ments on Escherichia coli crude extracts yielded a fivefold purification of p-galactosidase along with a simultaneous separation of proteins from cell debris in a single step. Experiments correlating the electrophoretic mobility in continuous electrophoresis with the elution behavior in ion-exchange chromatography were performed on more than a dozen proteins which conclusively showed that separation of proteins in continuous zone electrophoresis is governed by net surface charge. Based on these results, the fraction numbers in which the proteins eluted could be correctly predicted. Proteins and enzymes with differences >0.05 M elution molarities in ion-exchange chromatogra- phy were separated by continuous zone electrophoresis on a preparative scale (mg/h or g/h) with >90% recovery. This corresponds to a preparative scale separation of proteins and enzymes which differ in apparent electrophoretic mobility by only 0.70 X cm2/V . s. 0 1993 John Wiley & Sons, Inc. Key words: preparative separation proteins continu- ous zone electrophoresis ion-exchange chromatography surface charge electrophoretic mobility

INTRODUCTION

Continuous zone electroph~resis~~~-~,~'-~~~~~ is a continu- ous technique for the separation of biological molecules and particles. Using this technique, separations involving cells,' cell membranes," plant ~ h l o r o p l a s t s , ' ~ ~ ~ ~ malaria parasites," and viruses2' have been reported. Most of the work has been directed toward separation of cells and cell membranes-more than 70% of the papers in the above categories deal with cell separation. Further, most of the research has been confined to analytical applications and

* To whom all correspondence should be addressed. Present address: Department of Chemical Engineering, Massachusetts Institute of Technol- ogy, Cambridge, MA 02139-4307.

is of little relevance to separations on a preparative scale (mg/h or g/h).24 Because of the continuous and mild nature of the technique and due to the fact that recovery of biologi- cal activity is high in free solution, it is eminently suitable for preparative separations. However, compared to cell separation, continuous separation of proteins on a prepara- tive scale has received attention only r e ~ e n t l y . ~ . ~ , ~ ~ In the last few years, preparative separations involving recombi- nant tissue plasminogen a~ t iva to r ,~ porin,I2 m-amylase from Aspergillus oryzae,I6 alcohol dehydrogenase from baker's yeast26 and formate dehydrogenase, formaldehyde dehydro- genase, and methanol oxidase from Candida boidinii" have been reported in the literature. Evidence for the importance of net surface charge in continuous zone electrophoresis has been presented by the authors, based on the electrophoretic titration curve.18

In this contribution, the various parameters affecting resolution and throughput in continuous preparative zone electrophoresis have been studied. Examples of single- step preparative separations of proteins and simultaneous separation of proteins from cell debris are shown. Fi- nally, comparison with the elution behavior in ion-exchange chromatography is undertaken to elucidate the separation mechanism, which characterized the separation technique and enabled prediction of the fraction numbers in which the proteins elute from the electrophoresis chamber.

MATERIALS AND METHODS

Materials

Catalase from bovine liver and bovine thyroglobulin were part of the gel filtration calibration kit purchased from Pharmacia-LKB (Freiburg, Germany) while ovalbumin from chicken egg, bovine serum albumin, and ferritin from horse spleen were purchased from Serva (Heidelberg, Germany). Myoglobin from horse skeletal muscle and dithioerythritol were from Sigma (Deisenhofen, Germany), and DNase from bovine pancreas was from Boehringer (Mannheim, Germany). Tris(hydroxymethyl)aminomethane, acetic acid,

Biotechnology and Bioengineering, Vol. 42, Pp. 829-835 (1993) 0 1993 John Wiley & Sons, Inc. CCC 0006-3592/93/070829-07

and hydrochloric acid of analytical grade were bought from Merck (Darmstadt, Germany).

Crude Extract

Escherichia coli cells were stored frozen at -25°C. For use, 200 g wet weight cells were thawed and suspended in 10 mM Tris-acetate buffer, pH 7.5, and 1 mM diihio- erythritol to give a final concentration of 40%(w/v). The pH was checked and readjusted to 7.5 using Tris. The cells were disrupted by wet grinding in a 0.31 Dyno-Mill type KDL at 3000 rpm with 0.25450-mm glass beads. Three passes of the suspension were made (flow rate 2.5 L/h). After disruption, the pH was brought back to 7.5 with Tris. DNase was added to the crude extract to break the nucleic acids and reduce the viscosity of the suspension. A technical centrifugation was carried out at 15000g for 20 min. This operation removed whole cells and larger particles of cell debris that could clog the outlet fractions of the electrophoresis chamber, but finer cell debris was still present in the crude extract. Next, the crude extract was dialyzed against the chamber buffer. After enzyme activity measurement, the dialyzed crude extract was ready for injection into the continuous electrophoresis apparatus.

Experimental Apparatus

Experiments were carried out using the VaP-22 continuous electrophoresis apparatus (Bender and Hobein, Munich, Germany; Table I) in which the separation chamber is formed by two parallel plates-a Plexiglass plate at the front and a mirrored copper plate at the back. A rigid polypropylene spacer separates the two plates from each other and maintains a constant gap. Platinum wire elec- trodes are placed deep in a slit in the Plexiglass plate. Cellulose acetate membranes isolate this slit from the separation chamber. The separation chamber itself measures 50 X 10 X 0.05 cm with a volume of 25 mL. Both buffer and sample are continuously admitted in laminar flow at the top of the chamber and flow perpendicular to an applied electric field (Fig. 1). The separated products are collected in an array of 90 outlet tubes (Table I) at the bottom. A cooling system continuously circulates high-conductivity buffer past the platinum wire electrodes and carries off the heat dissipated by the electrophoretic process.

Principle and Operation

In continuous zone electrophoresis, sample components are separated based on differences in electrophoretic mobilities

Table I. Continuous electrophoresis apparatus.

Number of separation chambers 1 Dimensions of separation chamber Chamber volume 25 cm3 Number of outlet fractions 90

50 x 10 X 0.05 cm

Cooling Power (w) 375

and are deflected by the electric field at different angles. Zones broaden as the sample travels through the chamber and the sample components are diluted with buffer. Thus, sample and buffer flow continuously in the x direction and the electrophoretic migration takes place in the z direction (Fig. 1). The degree of dilution of samples was 3-6 for model proteins and 10-12 for enzymes from microorganism crude extracts. No surface coatings were used in this work. The direction of zone broadening due to electro-osmosis (toward the anode for a closed chamber with negatively charged walls) was opposite to that due to the parabolic velocity profile (toward the cathode for a negatively charged sample). Taking advantage of this fact, balancing conditions (Table 11) of buffer residence time and electric field strength were selected such that electro-osmosis did not lead to non-Gaussian peaks, tailing, or other deleterious effects in the separation profiles. All experiments were performed at 15°C using electric field strengths between 100 and 150 V/cm, a chamber buffer residence time between 2 and 5 min (corresponding to a chamber buffer flow rate between 12 and 5 mL/min, respectively), a sample flow rate from 1.5 to 6.2 mL/h, a sample concentration from 2.5 to 40 mg/mL, and a chamber buffer conductivity between 1.0 and 1.5 mS/cm. All protein samples were dissolved in the Tris-acetate chamber buffer. The electrode buffer was the same as the chamber buffer, except for its conductivity, which was 10 times higher. All electrophoretic migrations were measured from the position where the sample eluted at zero electric field.

Ion-Exchange Chromatography

Ion-exchange chromatography was performed on the Pharmacia fast protein liquid chromatography (FPLC) system22 using the anion-exchanger Mono Q (column diameter 5 mm, column length 5 cm, buffer flow rate 2 mL/h) and a linear NaCl gradient (0-0.5 M) (Table 111) in 25 mM Tris-acetate or potassium phosphate buffer (buffer A). Buffer B was made by adding 1 M NaCl to buffer A.

Sample I Buffer

Electrodes

Outlet fractions Figure 1. Principle of continuous zone electrophoresis.

830 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 42, NO. 7, SEPTEMBER 20, 1993

Table 11. electrophoresis.

Influence of operating parameters in continuous zone

Parameter HighiLow Resolution Throughput Bandwidth

Buffer residence time high high low high Sample flow rate high high Sample concentration high low high Electric field strength high high high

RESULTS AND DISCUSSION

First, the influence of various operating parameters in continuous zone electrophoresis [i.e., buffer residence time in the chamber (2-5 min), sample flow rate (1.5-6.2 mL/h), sample concentration (2.540 mg/mL), and electric field strength (100-150 V/cm)] on separation resolution, band- width (peak-width at half peak-height), and throughput was explored. These experiments were performed on a mixture of partially purified D- and L-hydroxyisocapro- ate dehydrogenase (D- and L-Hic dehydrogenase) model proteins.

Buffer Residence Time

High buffer residence times (5 min on more) lead to higher resolution but smaller throughputs in continuous zone electrophoresis (Table 11). Since the sample is exposed to the electric field for a longer duration, the bandwidths are larger; moreover, the parabolic buffer velocity profile in the separation chamber is responsible for zone broadening effects (particles closer to the chamber walls stay longer in the electric field and migrate farther). However, employing too low a residence time (< 2 min) may not lead to complete separation of sample components. As a result, a balance has to be made.

Table 111.

Volume

Program for ion-exchange chromatography experiments.

(mL) Setting Value

0.0 0.0 0.0 0.0 0.0 2.0 5.0

35.0 35.0 37.0 37.0 42.0

Conc %B mL/min cm/mL Port set Valve position Valve position Conc %B Conc %B Conc %B Conc %B Conc %B Conc %B

0.0 2.00 0.25 6.1 1.2 1 .1 0.0

50.0 100 100

0.0 0.0

Sample Flow Rate

The residence time of buffer in the chamber controlled the maximum sample flow rate that could be injected into the electrophoresis chamber without overloading it and thereby determined the maximum throughput (Table 11). In experiments on model proteins, low (< 2 min) buffer residence times [i.e., high ( > l o mL/min) buffer flow rates in the electrophoresis chamber] were used to enable high sample flow rates (> 6 mL/h) without overloading the chamber or adversely affecting the separation resolution (Figs. 2 and 3). The ratio of sample flow rate to buffer flow rate in the separation chamber varied between 0.005 : 1 and 0.01 : 1.

Sample Concentration

Increasing the sample concentration 10-fold from 2.5 to 25 mg/mL and from 25 to 40 mg/mL did not adversely affect the separation resolution in experiments on D- and L-Hic dehydrogenase mixtures, as long as the electric field or the buffer residence time was appropriately increased.

Electric Field Strength

Increasing the electric field strength in the experiments on model proteins led to larger bandwidths and larger current loads. For Tris-acetate buffer of pH 6.9 and conductivity 1.5 mS/cm, the electric current increased from 200 to 300 mA for an increase in electric field strength from 100 to 150 V/cm. Electric field strengths of 100 V/cm or more are required to achieve separation with good resolution when high sample concentrations and low buffer residence times are used.

In all cases, flat pH and conductivity profiles were found across the width of the separation chamber. The effect of the operating parameters on separation resolution, throughput, and bandwidth is summarized in Table 11.

Throughput in Continuous Zone Electrophoresis

In continuous zone electrophoresis experiments on mixtures of D- and L-Hic dehydrogenase model proteins, multiple injections of sample into the chamber were employed to increase the throughput. The optimized run is shown in Figures 2 and 3. Using a buffer residence time of 2 min, a field strength of 140 V/cm, a protein concentration of 40 mg/mL, and a sample flow rate of 6.2 mL/h at each injection point, a throughput of 500 mg/h protein was obtained, which is five times higher than the value reported earlier for continuous zone electrophoresis separation^.',^^

Experiments on E, coli Crude Extract

Experiments were carried out on E. coli crude extracts using continuous zone electrophoresis with a Tris-acetate buffer (pH 7.3, conductivity 1.5 mS/cm). The buffer residence time measured 5 min, the sample flow rate 1.5 mL/h,

NATH ET AL.: PREPARATIVE SEPARATION OF PROTEINS 83 1

- 12 120 7- I r Tris acetate butter !

Fraction number

Figure 2. Throughput optimization for a model system of a mixture of partially purified D- and L-hydroxyisocaproate dehydrogenase by continuous zone electrophoresis. Two injections of sample at frac- tion numbers 10 and 50. The throughput was 500 mg protein per hour. Operating conditions: buffer, Tris-acetate, pH 6.9, 1.5 mS/cm; electric field strength, 140 V/cm; electric current, 280 mA; buffer residence time, 2 min; sample flow rate, 6.2 mL/h per injection; sample concentration, 40 mg/mL.

Fraction number

Figure 3. Absorbance at 280 nm for the throughput experiments on partially purified D- and L-Hic dehydrogenase (Fig. 2). Two injections of sample at fractions 10 and 50.

the electric field strength 100 V/cm, and the electric current 250 mA. In a single step, a purification factor of greater than 5 was achieved for P-galactosidase, the enzyme assayed, along with a simultaneous separation of enzyme from cell debris (Fig. 4). Similar results have been obtained for the separation of enzymes from C. boidinii crude extracts.” To verify the stability of the separation, the experiment was operated continuously for a period of 12 h. The proteins and the cell debris eluted in the same fractions throughout the duration of the long-time run and the recovery of P-galactosidase activity was greater then 90%. More than 90% of the enzyme activity was found in only five fractions (Fig. 4). The pH and conductivity profiles of the electrode buffer as a function of time are shown in Figures 5 and 6. As a thumb rule, it is advisable to replace the spent electrode buffer if the pH change (due

60 Fraction number

Figure 4. Separation of P-galactosidase from E. coli crude ex- tract by continuous zone electrophoresis. Tris-acetate buffer, pH 7.3, conductivity 1.5 mS/cm; electric field strength, 100 V/cm; electric current, 250 mA; buffer residence time, 5 min; sample flow rate, 1.5 mL/h.

to electrolysis products) is greater than 0.5 units (Fig. 5). This prevents the chamber buffer pH and composition from being altered substantially.

When the electrophoresis experiment on E. coli crude extract was repeated with Tris-HC1 buffer, pH 8.0 (all other conditions identical to those given in Fig. 4), the separation failed; 90% of the P-galactosidase activity was spread over more than 40 fractions while the absorbance at 280 nm was distributed in all 90 fractions. This shows the importance of the choice of buffer in continuous zone electrophoresis.

Mechanism of Separation in Continuous Zone Electrophoresis

Continuous preparative zone electrophoresis separations can be characterized by the apparent electrophoretic mo- bility, ii:

- dAk u = - t l

where d is the measured electrophoretic migration in a con- tinuous electrophoresis experiment, A the chamber cross- sectional area, k the conductivity of the chamber buffer,

L

6.9- 11

0 2 4 6 8 1 0 1 2 ‘ 6. 5

Time (h)

Figure 5. time in a continuous long-time preparative zone electrophoresis run.

pH of the Tris-acetate electrode buffer as a function of

832 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 42, NO. 7, SEPTEMBER 20, 1993

Figure 6. Conductivity of the Tris-acetate electrode buffer as a function of time in a continuous long-time preparative zone elec- trophoresis run.

r the buffer residence time, and I the electric current. Experimental evidence indicated the importance of net surface charge in continuous zone electrophoresis; thus, in experiments with E. coli crude extracts, the separation failed when the acetate anion of the chamber buffer was replaced by the chloride, keeping all other conditions the same (text, Fig. 4). Further, it was found that the elec- trophoretic titration curves (the net surface charge vs. pH curves) of proteins could be used to choose buffer pH for optimal separation of a mixture of proteins." Therefore it was decided to compare continuous zone electrophore- sis with ion-exchange chromatography, which is a more

sensitive technique to elucidate the separation mechanism than the electrophoretic titration curve. Figure 7 depicts, for more than a dozen proteins, the correlation of the electrophoretic mobility [Eq. (l)] measured in a continuous zone electrophoresis experiment with the elution molarity measured in ion-exchange chromatography. The deviations from the linear correlation occur due to the effect of the Tris-acetate buffer pH, which varied from 6.8 to 7.9 in the continuous zone electrophoresis experiments but was kept constant at 7.5 in the ion-exchange chromatography runs. The linear correlation between the apparent elec- trophoretic mobility and the elution molarity confirms the primary role of net surface charge in continuous zone electrophoresis separations. Further, based on Eq. (1) and the results shown in Figure 7, i.e. for a known value of the elution molarity, it is possible to correctly predict the fraction numbers in which the proteins elute in a continuous preparative electrophoresis experiment. However, this does not mean that the separation mechanisms in continuous zone electrophoresis and ion-exchange chromatography are identical. In fact, at low salt, the electrophoretic mobility is very sensitive to variation in salt concentration while the elution molarity varies only slightly (the high-salt limit should show the opposite trend). Thus, continuous zone electrophoresis should prove to be a superior, more sensitive, higher resolution technique over ion-exchange chromatography in the low-salt range. However, mobil- ity-molarity plots of the kind shown in Figure 7 can be used as guidelines for predictions. Proteins and enzymes with differences greater than 0.05 M elution molarities

0.6r 1

C 0

Electrophoretic mobility (cmL/Vs)x lo5 Figure 7. Correlation between the electrophoretic mobility measured in continuous preparative zone electrophoresis and the elution molarity in ion-exchange chromatography for various proteins. Tris-acetate buffer; the pH varied from 6.8 to 7.9. (1) Myoglobin from horse heart. (2) Formate dehydrogenase from C. boidinii. (3) Catalase from bovine liver (Serva). (4) o-Hic dehydrogenase. (5) Catalase (Pharmacia). (6) Formaldehyde dehydrogenase from C. boidinii. (7) Bovine serum albumin. (8) Ovalbumin from chicken egg. (9) Ferritin from horse spleen. (10) P-Galactosidase from E. coli. (11) Bovine thyroglobulin. (12) Methanol oxidase from C. boidinii. (13) L-Hic dehydrogenase.

NATH ET AL.: PREPARATIVE SEPARATION OF PROTEINS a33

in ion-exchange chromatography could be separated by continuous zone electrophoresis on a preparative scale. This corresponds to a preparative separation of proteins which differ in apparent electrophoretic mobility by only 0.70 X cm2/V s.

The advantages of continuous electrophoresis over chro- matographic and other techniques are its continuous char- acter and the fact that it can readily handle particles and cell debris. There are no problems with column clogging, as in ion-exchange chromatography. Since protein samples are diluted with buffer during continuous zone electrophore- sis, problems due to protein-protein interactions and pro- tein precipitation [especially near the protein’s isoelec- tric point (PI) value] are avoided. Further, the continuous electrophoresis equipment offers considerable flexibility in that focusing and concentration techniques such as continuous isoelectric focusing and continuous field-step electrophoresis can also be implemented. A future article shall address the important problems of separation and simultaneous concentration and simultaneous reaction and separation of biomolecules by continuous electrophoresis.

In all continuous zone electrophoresis experiments, the reproducibility of the results was excellent and the recovery of enzyme activities greater than 90%. Equation (1) proves to be invaluable in reducing the number of variables that need to be optimized; optimum pH can be chosen using analytical electrophoresis. However, the type of buffer and buffer additives (e.g., detergents) still need to be experimentally optimized. A fundamental study of the interaction of detergents on proteins is sorely needed.

CONCLUSIONS

The present work is a comprehensive study in the application of continuous zone electrophoresis to protein separation in free solution on a preparative scale (mg/h or g/h). Resolution and throughput studies along with several practical separations are shown. The advantages of the technique over other techniques are its continuous character and its ability to handle particle-containing crude extracts. Preparative scale separations are shown to be feasible for proteins which differ in apparent electrophoretic mobility by only 0.70 X lop5 cm2/V s. Experimental evidence is provided to conclude that separations in continuous zone electrophoresis are based on net surface charge of the molecules; this also enables accurate prediction of the separation. Future research should address the important problems of separation and simultaneous concentration as well as reaction and simultaneous separation of biomolecules. A fundamental study of the role of buffer additives (e.g., detergents) in protein separations shall also represent a significant advance.

The first author (S. N.) sincerely acknowledges the financial sup- port and the experimental facilities par excellence extended to him during his stay at the National Research Institute of Biotechnology. The expert technical assistance of R. Kraume-Fliigel and I. Schmidt

is also acknowledged. S. N. thanks N. Papamichael and S. Das for help with literature and P. Ghosh and V. Wray for encouragement that uplifts. Thanks are also expressed to R. Kapoor and R.K. Khurana for drafting the figures and to R.N. Shukla and R. Rajagopalan for typing the manuscript.

References

1.

2.

3.

4.

5 .

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Baier, T. G., Weber, G., Hartmann, K., Heinrich, U., Schonberg, D. 1988. Preparative separation of human B and T lymphocytes by free-flow electrophoresis. Anal. Biochem. 171: 91 -95. Barrollier, J., Watzke, E., Gibian, H. 1958. Einfache Apparatur fur die tragerfreie praparative Durchlauf Elektrophorese. Z. Naturforschung 13b: 754-755. Barth, F., Gruetter, M.G., Kessler, R., Manz, H.J. 1986. The use of free-flow electrophoresis in the purification of recombinant human tissue plasminogen activator expressed in yeast. Electrophoresis 7: 372-375. Basset, P., Froissart, C., Vincendon, G., Massarelli, R. 1980. A continuous flow system for protein fractionation; description and application with partial purification of cholin acetyl transpherase. Electrophoresis 1: 168-172. Bradford, M. M. 1976. A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of protein- dye binding. Anal. Biochem. 72: 248-254. Glad, M., Larsson, P.-0. 1991. New methods for separation and recovery of biomolecules. Curr. Opinion Biotechnol. 2: 413-418. Hamel, J . F. P., Hunter, J. B. 1990. Modeling and applications of downstream processing: A survey of innovative strategies. pp. 1-35 In: J. F. P. Hamel, J. B. Hunter, and S. K. Sikdar (eds.), Downstream processing and bioseparation, ACS Symposium Series 419. American Chemical Society, Washington, DC. Hannig, K. 1961. Die tragerfreie kontinuierliche Elektrophorese und ihre Anwendung. Z. Anal. Chem. 181: 244-248. Hannig, K. 1982. New aspects in preparative and analytical continu- ous free-flow cell electrophoresis. Electrophoresis 3 235 -243. Hannig, K., Heidrich, H. G. 1974. The use of continuous preparative free-flow electrophoresis for dissociating cell fractions and isolation of membranous components. pp. 746-761 In: S. Fleischer and L. Packer (eds.), Methods in enzymology, vol. 31, Biomembranes, Part A. Academic, New York. Heidrich, H.G., Danfnrth, H.D., Beaudoin, R. L. 1980. Free-flow electrophoresis separation of whole body plasmodium berghei sporo- zoites. Eur. J. Cell Biol. 22: 516-522. Holzenburg, A., Engel, A., Kessler, R., Manz, H. J., Lustig, A,, Aebi, U. 1989. Rapid isolation of Ompf porin-LPS complexes suitable for structure-function studies. Biochemistry 28: 4189-4193. Ivory, C. N. 1988. The prospects for large-scale electrophoresis. Sep. Sci. 23: 875-912. Janson, J.-C., Ryden, L. 1989. Protein purification: Principles, high resolution methods and applications. VCH, New York. Kappler, R., Kristen, U., M o d , D. J. 1986. Membrane flow in plants: Fractionation of growing pollen tubes of tobacco by preparative free- flow electrophoresis and kinetics of labelling of endoplastic reticulum and Golgi apparatus with (3H) leucine. Protoplasma 132: 38-50. Kuhn, R., Wagner, H. 1989. Free-flow electrophoresis as a method for the purification of enzymes from E. coli cell extract. Electrophoresis 10: 165-172. Nath, S., Schutte, H., Hustedt H., Deckwer, W.-D. 1989. Contin- uous separation of proteins by free-flnw zone electrophoresis. Ber. Bunsenges. Phys. Chem. 93: 1042- 1046. Nath, S., Schiitte, H., Hustedt, H., Deckwer, W.-D. 1990. Corre- lation of migration behavior in free-flow zone electrophoresis and electrophoretic titration curve. Electrophoresis 11: 612-616. Nath, S., Schiitte, H., Weber, G., Hustedt, H., Deckwer, W.-D. 1990. Separation of enzymes from Candidu boidinii crude extract by continuous flow zone electrophoresis. Electrophoresis 11: 937-941.

a34 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 42, NO. 7, SEPTEMBER 20, 1993

20. Ramsey, W.S., Nowlan, E.D., Simpson, L.B. 1980. Resolution of microbial mixtures by free-flow electrophoresis. Eur. J. Appl. Microbiol. Biotechnol. 9: 217-226.

21. Sandelius, A. S., Penel, C., Anderset, G., Brightman, A,, Millard, M., M o d , D.J. 1986. Isolation of highly purified fractions of plasma membrane and tonoplast from the same homogenate of soybean hypocotyls by free-flow electrophoresis. Plant Physiol. 81: 177- 1x5.

22. Schiitte. H., Hummel, W., Kula, M.-R. 1985. Improved enzyme screening by automated fast protein liquid chromatography Anal. Biochem. 151: 547-553.

23. Scopes, R. K. 1987. Protein purification: Principles and practice. Springer, New York.

24. Thormann, W., Mosher, R. A,, Bier, M. 1986. Free flow electrophore- sis. pp. 153-168 In: C.J. King and J.D. Navratil (eds.), Chemical separations, vol. 1, Principles. Litarvan Literature, Denver, CO.

25. Wagner, H., Mang, V. 1985. Voruntersuchungen und Empfehlungen zum Einsatz and zu wirtschaftlichen Nutzung der Elektrophorese unter Weltraumbedingungen. Universitat des Saarlandes, Saarbriicken, Germany.

26. Wagner, H., Kuhn, R., Hoffstetter, S. 1989. Praxis der elektrophoretis- chen Trennmethoden. Springer, Heidelberg.

NATH ET AL.: PREPARATIVE SEPARATION OF PROTEINS a35