Method for the Fractionation of Dextran by Centrifugal Precipitation

Recent advances in biotechnology and biochemistry have been facilitated by efficient separation methods for biopolymers, such as proteins and nucleic ...
0 downloads 0 Views 178KB Size
Download PDF
Anal. Chem. 2002, 74, 440-445

Method for the Fractionation of Dextran by Centrifugal Precipitation Chromatography Fuquan Yang and Yoichiro Ito*

Laboratory of Biophysical Chemistry, National Heart, Lung, and Blood Institute, National Institutes of Health, Room 3334, Building 50, 50 South Drive MSC 8014, Bethesda, Maryland 20892-8014

Recent advances in biotechnology and biochemistry have been facilitated by efficient separation methods for biopolymers, such as proteins and nucleic acids. On the other hand, research on polysaccharides is hindered by problems in their fractionation. For many decades, polysaccharides have been fractionated by stepwise ethanol precipitation, and even at the present time, almost all of their purification protocols include at least one such step, although it is tedious and inefficient. This paper describes a novel approach for chromatographic fractionation of polysaccharides by ethanol gradient precipitation using an open column under a centrifugal force field. Using a unique column design, chromatographic separation of polymers is achieved by subjecting the sample to a repetitive process of precipitation and dissolution along a long, spiral channel. In this article, we fully describe the principle, design of the prototype, and basic studies on various parameters for optimization of chromatographic conditions, using dextran as an example. Centrifugal precipitation chromatography (CPC) has been developed for separation and purification of materials ranging from small-molecular-weight compounds to macromolecules.1 The principle and basic studies of protein separation by ammonium sulfate precipitation were described earlier,2 and more recently, preliminary applications of the method to the separation of polycatechin3 and chondroitin sulfate4 have been reported. PRINCIPLE AND DESIGN OF THE APPARATUS Consider a pair of channels partitioned with a dialysis membrane, as shown in Figure 1. Absolute ethanol is eluted through one channel, and water, through the other in the opposite direction at a lower flow rate, as indicated. This countercurrent process allows migration of ethanol through the membrane from the ethanol channel toward the water channel and water, to the ethanol channel. Assuming that the rate of ethanol migration is proportional to the difference in ethanol concentration between the two channels and that the ethanol concentration is nearly * Corresponding author. Phone: 301-496-1210. Fax: 301-402-3404. E-mail: [email protected]. (1) Ito, Y. J. Liq. Chromatogr. Relat. Technol. 1999, 22 (18), 2825-2836. (2) Ito, Y. Anal. Biochem. 2000, 277 (1), 143-153. (3) Degenhardt, A.; Engelhardt, U. H.; Winterhalter, P.; Ito, Y. J. Agric. Food Chem. 2001, 49, 1730-1736. (4) Shinomiya, K.; Kabasawa, Y.; Toida, T.; Imanari, T.; Ito, Y. J. Chromatogr. A 2001, 922, 365-369.

440 Analytical Chemistry, Vol. 74, No. 2, January 15, 2002

Figure 1. Principle of centrifugal precipitation chromatography.

constant throughout the ethanol channel (since V . v), an increment in ethanol concentration in the water channel is expressed by the following equation,

dc ) k(C - c) dt

(1)

c ) C(1 - e-kt)

(2)

which becomes

Eq 2 indicates that an exponential gradient of ethanol concentration is formed through the water channel. This gradient is stable as long as the flow rate through each channel remains unaltered. Under a centrifugal force field, polysaccharides introduced into the water channel are exposed to a gradually increasing ethanol concentration and, finally, are precipitated at various locations along the channel according to their solubility in the ethanol solution. Chromatographic elution of polysaccharide is then initiated by gradually decreasing the ethanol concentration (C) in the ethanol channel, which results in a proportional reduction of the ethanol concentration in the gradient, as indicated by eq 2. This, in turn, causes once-deposited polysaccharide to dissolve and reprecipitate at an advanced location in the channel. Because the ethanol gradient concentration in the channel falls at a rate much lower than the flow rate of the mobile phase, the polysaccharides undergo a repetitive process of precipitation and dissolution along the channel and elute in a decreasing order of their solubility in the ethanol solution. The centrifugal precipitation chromatograph (Figure 2) used in the present study consists of a compact seal-free centrifuge (Pharma-Tech Research Corporation, Baltimore, MD). The separation column was newly designed by one of us (Y.I.) and 10.1021/ac010948r Not subject to U.S. Copyright. Publ. 2002 Am. Chem. Soc.

Published on Web 12/15/2001

Figure 2. Centrifugal precipitation chromatograph: (a) column channel; (b) cross section of the centrifugal precipitation chromatograph prototype through the central axis.

fabricated at the NIH machine shop, as shown in Figure 2A. The column consists of a pair of disks (high-density polyethylene, 13.2 cm diameter and 1.5 cm thick), each with a spiral groove measuring 1.5 mm wide, 0.5 mm (upper) and 2 mm (lower) deep, and ca. 2 m in length. With proper alignment, these grooves form a single channel. A dialysis membrane is sandwiched between these two disks to form two identical channels partitioned by the membrane, as shown in Figure 1. One channel has a 2-mL

capacity, and the other channel, a 7.3-mL capacity. The disk assembly is tightly bolted and mounted on the sealless continuousflow centrifuge, which allows elution through multiple channels without the use of a conventional rotary seal device. The crosssection view of the apparatus is schematically shown in Figure 2B. The motor drives the rotary frame through a pair of toothed pulleys and a toothed belt. The frame holds a miter gear assembly consisting of a stationary lower gear, a pair of horizontal idlers, Analytical Chemistry, Vol. 74, No. 2, January 15, 2002

441

Figure 3. Ethanol transfer and osmosis rates through the dialysis membrane: (A) ethanol transfer through the membrane measured at various water input rates; (B) osmosis rate through the membrane measured at various water input rates. Experimental conditions: ethanol channel, ethanol at 1.0 mL/min; water channel, water at various flow rates of 0.1 to 1.0 mL/min. Sample: not charged. Revolution speeds: 0 rpm (2), 1000 rpm (9), and 2000 rpm (b).

and an upper gear that is connected to the separation disk assembly through the central shaft. Because the lower miter gear is rigidly mounted at the bottom of the centrifuge, rotation of the frame produces synchronous rotation of the horizontal idlers around their own axes in the rotating frame. This motion is further conveyed to the upper miter gear and the separation disk assembly, which rotates at double speed around the central axis of the centrifuge. This 2:1 rotation ratio between the disk assembly and the frame prevents flow tubes from twisting during the revolution. As shown in the diagram, two pairs of flow tubes (0.5-mm i.d. PTFE, Zeus Industrial Products, Raritan, NJ) from the disk assembly are successively led through the hollow central shaft downward, the hollow idler gear shaft horizontally, and then the tube support upward, finally exiting at the top center of the centrifuge, where they are tightly secured with a pair of clamps. These tubes are bundled, lubricated, and protected with a sheath of Tygon tubing at each flexing portion to prevent direct contact with the hard surface. If this precaution is taken, then these tubes can maintain their integrity for many runs. The revolution speed of the centrifuge is regulated up to 1500 rpm using a speed controller, which rotates the top disk assembly at a doubled rate of 3000 rpm. The present studies were mostly performed at 2000 rpm (ca. 260g at the peripheral terminal of the spiral channel). EXPERIMENTAL SECTION Reagents. Ethyl alcohol (anhydrous) was purchased from the Warner-Graham Company (Cockeysville, MD); water (HPLC grade), from Fisher Scientific (Fair Lawn, NJ); and dibasic and monobasic potassium phosphates, both of reagent grade, from Mallinckodt Baker, Inc. (Paris, KY). Dextran standards, including 25, 50, 80, 150, 270, 410, 670, and 750 kDa and crude dextran (average molecular weight, 77 kDa), were purchased from SigmaAldrich (St. Louis, MO). Basic Studies on Ethanol Transfer and Osmosis Rates. The ethanol transfer rate through the membrane was measured according to the following procedure. The anhydrous ethanol was eluted through one channel at a fixed flow rate of 1 mL/min while 442

Analytical Chemistry, Vol. 74, No. 2, January 15, 2002

water was eluted through the other channel at various flow rates ranging from 0.1 to 1 mL/min. No dextran sample was used in this study. After hydrodynamic equilibrium was reached, the output flow rate from the water channel and the relative ethanol concentration from the water channel were determined using the on-line flow injection system. The experiments were performed under three different column conditions: (1) the stationary column, by eluting ethanol through the lower channel; (2) the column rotated at 1000 rpm; and (3) the column rotated at 2000 rpm. Preparation of Sample Solutions. Individual molecular mass standard solutions were prepared by dissolving 0.5 mg each of dextran standards with a MW range from 23.8 to 771 kDa in 1.0 mL of 50 mM phosphate mobile-phase buffer. The crude dextran solutions were prepared by dissolving 40 mg each of the crude sample in 3 mL of 50 mM phosphate mobilephase buffer. CPC Fractionation of Dextran. In each experiment, both the ethanol channel (lower channel) and the water channel (upper channel) were completely filled with 100% ethanol. After the sample solution was introduced through a sample loop, the column was rotated at the desired rate, usually 2000 rpm. The ethanol channel was then eluted with the ethanol gradient using a gradient pump (Shimadzu SCL-10A and LC-10AD, Shimadzu Scientific Co., Columbia, MD) at a total flow rate of 1 mL/min, and the water channel was eluted with HPLC grade water using a Harvard syringe pump (model 980532, Harvard Apparatus, South Natick, MA) at a flow rate of 0.12 mL/min. The effluent from the water channel was continuously monitored through a UV monitor (Uvicord S, LKB Instruments, Bromma, Sweden) at 275 nm and collected in test tubes at 4-min intervals using a fraction collector (Ultrorac, LKB Instruments, Bromma, Sweden). Analysis of CPC Fractions. Total polysaccharide in each collected fraction was quantitatively determined by phenolsulfuric assay according to Dubois et al.5,6 (5) Dubois, M.; Gilles, K. A.; Hamilton, J. K.; Roberts, P. A.; Smith, F. Anal. Chem. 1956, 28 (3), 350-356. (6) Pazur, J. H. In Carbohydrate AnalysissA Practical Approach; Chaplin, M. F., Kennedy, J. F., Eds.; IRL Press: Oxford, 1987; pp 55-142.

Analytical Chemistry, Vol. 74, No. 2, January 15, 2002

443

Figure 4. Dextran fractionation by centrifugal precipitation chromatography (CPC). CPC experimental conditions: revolution speed, 2000 rpm; flow rates, 1.0 mL/min (ethanol channel) and 0.12 mL/min (water channel); sample size, 10 mg dextran crude sample dissolved in 1 mL of water; ethanol gradient in ethanol channel, ethanol concentration 100-0% in 100 (A), 200 (B), 300 (C), and 400 (D) min; detection, 460 nm.

Figure 5. Dextran fractionation by centrifugal precipitation chromatography (A) and HPSEC analysis (B). CPC experimental conditions: revolution speed, 2000 rpm; flow rates, 1.0 mL/min (ethanol channel) and 0.12 mL/min (water channel); sample size, 10 mg dextran crude sample dissolved in 1 mL of water; ethanol gradient in ethanol channel, ethanol concentration 100 (0.01min)-0% (400 min); detection, 460 nm. HPSEC conditions: column, TSK-GEL G5000PWXL (30 cm × 7.8 mm i.d., 10 µm); mobile phase, 50 mM potassium phosphate (pH 6.9); flow rate, 0.7 mL/min.

Analysis of the crude dextran sample solutions and the collected dextran fractions was performed using a high-performance size-exclusion chromatography (HPSEC) system consisting of a Waters 515 HPLC pump, a model 7725 injection valve with a 20-µL loop, a Waters 2410 refractive index detector set at 40 °C (Waters, Milford, MA), and a CR501 Chromatopac integrator (Shimadzu Corporation, Kyoto, Japan). A TSK-GEL G5000PWXL column (30 cm × 7.8 mm i.d., 10 µm, TosoHaas, Montgomeryville, PA) was used for analysis. The mobile phase (pH 6.9), composed of aqueous potassium phosphate monobasic and dibasic (50 mM), was pumped at a flow rate of 0.7 mL/min, resulting in typical operating pressures of ∼500 psi.7 The samples were prepared by dissolving each CPC fraction in 0.2 mL of 50 mM phosphate mobile phase buffer. Calibration Curve of Molecular Mass. Using the HPSEC system, an eight-level molecular weight calibration curve was drawn with a set of dextran standards (0.5 mg/mL). It showed an excellent linearity (coefficient of correlation, r ) -0.999) in a range of 23.8-771 kDa when the log of the molecular weight was plotted against the HPSEC retention time. RESULT AND DISCUSSION Ethanol Transfer and Osmosis Rates through the Membrane. A series of basic studies on ethanol transfer and osmosis rates through the membrane was performed without a sample by varying the water input rate from 1 to 0.1 mL/min while the (7) White, G.; Katona, T.; Zodda, J. P. J. Pharm. Biomed. Anal. 1999, 20, 905912.

444

Analytical Chemistry, Vol. 74, No. 2, January 15, 2002

anhydrous ethanol was eluted through the ethanol channel at a fixed flow rate of 1 mL/min. As shown in Figure 3A, the ethanol concentration in the water output (ethanol transfer rate) increases as the water input flow rate is decreased and very slightly as the revolution speed is increased. The ethanol concentration in the water output reaches 82 and 88% at 0.1 mL/min of the water input rate at revolution speeds of 1000 and 2000 rpm, respectively. The osmosis rate through the membrane is similarly illustrated in Figure 3B, in which the relative water output rate (%) is plotted against the water input rate. The osmosis curves form an inverted shape of the ethanol transfer curves, that is, the osmosis rate becomes enhanced by the centrifugal force, and at 2000 rpm, the water output reduces to 23% at a 0.1 mL/min water input rate. When the water input is further reduced to 0.05 mL/min, no liquid is collected from the outlet of the channel. CPC Fractionation of Crude Dextran. Figure 4 shows the results of the CPC fractionation of a crude dextran mixture using linear gradients from 100 to 0% ethanol at elution times of 100, 200, 300, and 400 min. Quantitation was performed using the phenol-sulfuric assay, as described earlier. It appears that the dextran standards are split into three fairly distinct peaks as the applied gradient time is increased to 400 min. Figure 5 shows the CPC chromatogram of the same crude mixture using linear gradient from 100 to 0% at an elution time of 400 min in Figure 5A (identical to the chromatogram shown in Figure 4D) and a set of HPSEC analysis data of collected fractions in Figure 5B. It demonstrates that the CPC method enables fractionation of dextran according to its molecular mass, as in size exclusion

chromatography, suggesting that the method may be useful for the preparation of dextran standards. In the past, chromatographic separations of polymers according to their solubility using the liquid chromatographic column filled with a solid support that is eluted with a gradient of suitable solvent were reported by several workers8-13. The use of solid support, however, may cause some problems, such as clogging of the column, carryover of fine precipitates, and irreversible adsorption and denaturation of samples at the solid-liquid interface. As described above, the present method uses an open channel where a desired ethanol/water gradient is established by osmosis through a dialysis membrane. Under a centrifugal force field, the sample molecules are subjected to the repetitive process of precipitation and dissolution and fractionated according to their solubility. The centrifugal force, which is unique to our chromato(8) Porath, J. Nature 1962, 196, 47-48. (9) King, T. P. Biochem. 1972, 11, 367-371. (10) Petro, M.; Svec, F.; Fre´chet, M. J. J. Chromatogr. A 1996, 752, 59-66. (11) Petro, M.; Svec, F.; Gitsov, I.; Fre´chet, M. J. Anal. Chem. 1996, 68, 315321. (12) Staal, W. J. Thesis, Elndhoven University of Technology, 1996. (13) Zhang, L. N.; Zhou, J. P.; Yang, Z.; Chen, J. H. J. Chromatogr. A 1998, 816, 131-136.

graphic system, enables the following three major functions: stabilization of the density gradient in the sample channel; acceleration of the osmosis process through the membrane; and most importantly, the retention of precipitated polymers at the critical spot in the channel. Basic studies on these factors are described in detail for the protein separation by ammonium sulfate precipitation.2 In addition to high sample recovery, the method offers various advantages, as compared to regular size-exclusion chromatography, including the ability to concentrate the sample solution within a channel, more flexible manipulation of the gradient through the channel, thorough elimination of low-molecular-weight impurities from solid support, etc. Fractionation of crude polysaccharides from plant extracts using the system is currently underway in our laboratory. ACKNOWLEDGMENT The authors are deeply indebted to Dr. Henry M. Fales for editing the manuscript and for valuable suggestions. Received for review August 24, 2001. Accepted November 1, 2001. AC010948R

Analytical Chemistry, Vol. 74, No. 2, January 15, 2002

445