Continuous Separation of Model Proteins by Free-Flow Field Step Electrophoresis and its Electrokinetic Basis

Continuous separation of cells, cell membranes, and proteins by electrophoretic techniques such as free-flow zone electrophoresis (FFZE) has a rich history since its introduction sixty years ago by Hannig. However, the results of FFZE, capillary zone electrophoresis, and other similar electrophoretic techniques are limited to analytical scale separations that are not readily extendable to the preparative scale. Moreover, a five- to ten-fold dilution of samples by buffer is common in separations by these techniques. Free-flow field step electrophoresis (FFFSE) is an electrophoretic technique that appears to be suitable for continuous simultaneous preparative separation and concentration and therefore has the power to overcome the above limitations. Here we apply FFFSE to a simple system of model proteins to show proof-of-concept of the technique. A continuous, preparative-scale separation of myoglobin from BSA with a throughput of 20 mg/h and a yield of >98% is shown to be successfully obtained using single-step FFFSE. Most important, it is shown that the preparative FFFSE experiment can be rationally designed, and the results predicted theoretically by use of electrokinetic data recorded in a simple analytical-scale FFZE experiment. This is the first paper to present a theory of separation in FFFSE. The separation is continuous , single-step , and environmental-friendly as no adjuvants are used, with no need for regeneration of components. The electrokinetic basis of the separation appears to be of a general nature. In further research we are testing the limits of the technique, by exploring its extension to more complex systems, and to higher preparative-scale throughputs.


Introduction
Electrophoresis in free buffer offers an attractive approach for carrying out bioseparation in a continuous mode [1]. Among an array of free-flow electrophoretic techniques [2,3], free-flow zone electrophoresis (FFZE) is the best known, and its field of application is very broad. It has been successfully used for the separation of cells, cell debris, populations of organelles as well as protein mixtures [1−7]. A merit of the FFZE process, which ought to be stressed, is its ability to maintain biological activity of samples given that the separation is achieved in free fluid in the absence of any solid support or gel. The technique has also been shown to be very useful for complex mixtures such as fermentation crude extracts without the necessity of any additional purification step [6,8,9].
However, like all techniques, the FFZE technique presents certain limitations.
Most of the documented results of FFZE are applicable to analytical scale separations [10] and cannot be easily extended to the preparative scale.
Moreover, at least a five-fold dilution factor of samples by buffer is commonly observed in FFZE, and hence the technique is not ideally suitable for continuous simultaneous electrophoretic separation and concentration [6,9]. Free-flow field step electrophoresis (FFFSE) appeared to be a promising alternative technique to achieve this aim, although free-flow isoelectric focusing, moving reaction boundary electrophoresis and other predictive techniques have been pioneered in the literature [11−19], notably by Ivory and colleagues [3, 15−18].
A literature search revealed that other than the continuous separation of the dye rhodamine B from fluorescein [2], no applications to preparative protein separation had been documented by the FFFSE technique. Turgeon and Bowser [11] also reported that no further data using FFFSE was available. The situation has remained virtually unchanged since then, though many interesting approaches to the general problem of protein complexation and separation have evolved, and new media and approaches have been tried out [20−25]. It is the purpose of this work to apply FFFSE to a simple system of model proteins, develop a theory of separation, and highlight its electrokinetic basis. However, it is not our aim to provide a catalog of successful applications. Above all, it is shown that a simultaneous separation and concentration process in FFFSE can be successfully designed and predicted using electrophoretic mobility data obtained from a simple FFZE experiment. The underlying principle of the FFFSE process is outlined in Section 2.

Principle of FFFSE
As mentioned in Section 1, analytical techniques such as FFZE are beset by a large number of difficulties. In these techniques, the sample is diluted by buffer, and a number of factors lead to band/zone broadening and negatively influence the quality of the separation. Primary factors of band/zone broadening and loss of separation efficiency arise from disturbances arising from the parabolic flow profile and from electroosmosis [2]. Although some innovative ways to mitigate the effects of these disturbances and optimizing FFZE separations have been described in the literature [6], the fact remains that most FFZE applications are still limited by resolution.
The practical implementation of FFZE leads to high dispersion of samples and therefore to poor separation performance, as discussed above. A major cause of these deleterious effects in FFZE is the use of buffer with constant physical properties throughout the width of the separation channel. These difficulties are circumvented in FFFSE by changing the assumption of constant physical properties of the buffer in the separation chamber (Fig. 1). In particular, differences in electrical conductivity, of media within the separation chamber ( Fig. 1) permits rapid migration and separation of analytes in the low regions in FFFSE (Fig. 2). It also serves to focus and concentrate the charged species as they approach the boundaries of the high regions, since the species now experience a small value of the electrical field strength, given that ∝ −1 (Figs. 1, 2). Thus at high sample flow rates, band broadening of analytes is minimized in FFFSE, and simultaneous separation and concentration can be achieved continuously on a preparative scale (Fig. 2).
In current practice, FFFSE uses high conductivity buffers only along the edges of the separation chamber. These create boundary walls at the two edges along which the electrophoretic mobility of the samples decrease, causing focusing of the bands (Fig. 2). This design requires some modifications for preparative separation of proteins by FFFSE, as discussed in subsequent Sections.

Electrophoretic apparatus
The rectangular plexiglass electrophoretic apparatus used for the FFZE and

Analytical methods
Separation results were evaluated from concentration measurements at the exit fraction ports of the chamber. Myoglobin was estimated from optical density T h e f i r s t w as from 0.5 to 1.5 mS/cm with the second from 1.5 to 9.0 mS/cm (Fig. 3).

Design of separation process by FFFSE
Several preliminary experiments conducted on model proteins revealed that it  Next an FFFSE experiment using two anodic conductivity steps from 0.5 -1.5 mS/cm and 1.5 -9. 0 m S / c m i n a T r i s -acetate buffer of pH 7.7 was designed and BSA along with their simultaneous separation was obtained, with a throughput of 20 mg/h was obtained using FFFSE with a yield >98% (Fig. 4).
In order to rationalize the FFFSE separation shown in Fig. 4, an analytical FFZE experiment on more than a dozen proteins was conducted and their electrophoretic mobilities determined, as shown in Fig. 5. The electrophoretic mobility of the subset of myoglobin and BSA proteins obtained from this analytical experiment is tabulated in Table 1. From this data, the migration distance of myoglobin and BSA was calculated based on Eq. (1) [6], and this was used to determined the position of the conductivity steps in preparative FFFSE. The equation and model calculations are given by: where measures the distance of migration in the electrical field, is the electrophoretic mobility of the protein, the residence time in the electrophoresis chamber, the current, the cross-sectional area of the chamber, and is the conductivity of the medium. The factor 2 3 accounts for the correction due to the parabolic profile of the Poiseuille flow of buffer and sample in the electrophoresis chamber [6,9].
Applying Eq. (1) with the parameters in Table 1, we find for BSA The number of fractions migrated, by BSA is given by which works out to = 24 fractions. Hence if the BSA is injected at fraction number 27 in a buffer of = 1.5 ⁄ , then its peak will elute at fraction number 51. Similar calculations based on the data of table 1 show that the peak of a sample of myoglobin injected at fraction 27 will elute at fraction 30 in buffer of the same conductivity.
The above discussion has shown that application of the governing electrokinetic equation of FFZE [6,9] together with the values of electrophoretic mobilities of the components in the separation process shown in Table 1 (Figs. 3, 4). For the conditions of the FFFSE experiment of Fig. 4, the zone electrophoresis equation [6] applicable to predict the elution fractions in FFFSE therefore adopts the form, = where , represent the distance migrated in the electrophoretic experiment in any two zones and of the FFFSE chamber with varying conductivity properties, , .  Table 2.
A sample calculation for BSA will help to elucidate the calculation procedure for the spread of the distribution of the BSA concentration. Consider for example the right front of the BSA. The right edge of the injection at fraction 30 will move a number of fractions equal to 6 +  Table 2). The spread of the distributions calculated using Eq. (5) and tabulated in Table 2 are in good agreement with the observed distributions of the individual species (Fig. 4).
The results of Table 2 suggest that the reason for the high yield obtained in the FFFSE separation of Fig. 4 arises from the fact that the BSA migrates towards the anode and reaches the edge of the highest conductivity, κ = 9.0 mS/cm at fractions 36 -40, and thereby focuses at the second anodic conductivity step (Fig. 4). On the other hand, the lower electrophoretic mobility of myoglobin (Table 1) only lets it migrate to fraction 33, as results of similar calculations (Table 2) show that it distributes between fractions 29 and 33, and that it does not reach fraction 36 where κ = 9.0 mS/cm. Thus the myoglobin fraction focuses at the first anodic conductivity step, as found in Fig. 4.
The above analysis leads to the thumb-rule that experimental conditions in

Conclusion
The present work describes a continuous, single-step procedure for preparative separation by free-flow field step electrophoresis by a new design approach of that employs multiple anodic electrical conductivity steps. The experiment is designed such that each component of the system focuses at a separate conductivity step. A continuous, preparative scale separation of model proteins myoglobin from BSA with a throughput of 20 mg/h could be successfully obtained with a yield of >98% using the above methodology by FFFSE.
Variables that need to be experimentally optimized include the electrical conductivity, κ of the intermediate stage, the concentration of the protein mixture, the magnitude of the applied electrical field, and the operating pH. The above results appear to be of a general nature and have the potential to be extended to other binary systems using the electrokinetic properties determined

Consent for publication
Not applicable

Availability of data and materials
All are included in the article

Competing interests
The author declares no competing interests

Funding
No funding was received for this particular work

Author contributions
S.N. conceived the study, arranged materials, conducted the study, analyzed the data, and wrote the paper Legends to the Figures   Fig. 1. Principle of free-flow field step electrophoresis (FFFSE). Profile of electrical conductivity, κ and electrical field strength, E in the standard FFFSE configuration.     Tables   Table 1. Electrophoretic mobility of model proteins myoglobin and BSA determined by analytical scale FFZE (Fig. 5).     Focuses at first κ step

-40
Focuses at second κ step Figure 1 Principle of free-ow eld step electrophoresis (FFFSE). Pro le of electrical conductivity, κ and electrical eld strength, E in the standard FFFSE con guration.

Figure 2
Operation of FFFSE. In the standard con guration, the two components in the mixture migrate in opposite directions and focus at the edges of the single electrical conductivity step.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.