Separation of prostaglandins using the new horizontal flow-through coil planet centrifuge

Journal o f Biochemical and Biophysical Methods, 3 (1980) 77--87

77

© Elsevier/North-Holland Biomedical Press

SEPARATION OF P R O S T A G L A N D I N S USING THE NEW HORIZONTAL FLOW-THROUGH COIL PLANET CENTRIFUGE

ELISE A.B. BROWN 1 and YOICHIRO ITO 2

1 Hypertension-Endodrine Branch, and 2 Laboratory o f Technical Development, National Heart, Lung and Blood Institute, National Institutes o f Health, Bethesda, MD 20205, U.S.A. (Received 13 November 1979; accepted 18 February 1980)

Prostaglandins and their metabolites have been separated using the new horizontal flow-through coil planet centrifuge to perform countercurrent chromatography. The system is a continuous-flow system which provides separations similar to column and thin-layer chromatography without the use of solid supports. Either phase of a twophase solvent system can be used to elute the compounds with the low pressure produced by a metering pump. Separations were produced in PTFE tubing, 0.55 mm i.d. (wound in 0.68 cm helical diameter) with a total capacity of 24 ml. A solvent system of chloroform/acetic acid/water (2 : 2 : 1, v/v) was used at flow rates of 2.4 and 6 ml/h and at a rotation of 500 rev./min. Microgram quantities of prostaglandins E2, F2~, A2, and B2 and thromboxane B2 can be separated. This method can be scaled up to 260-ml columns with comparable results. Key words: prostaglandins; flow-through coil planet centrifuge; countercurrent chromatography.

INTRODUCTION

Separation of prostaglandins has been accomplished by a number of methods including column chromatography and high-pressure liquid chromatography. Partitioning chloroform extracts of biological samples on silicic acid columns led to highly variable recoveries in the range of 25--75% of the amount applied [1] when using the solvent elution sequence by Jaffe et al. [2]. Similarly, problems of resolution and of sample preparation have been reported for high-pressure liquid chromatography [3]. The theoretical calculations used for separations achieved by high-pressure liquid chromatography and thin-layer chromatography are based on the phenomenon of partition between two liquid phases, as is countercurrent chromatography. This fact allowed us to extrapolate the known characteristics of partition solvent systems used in these methods to our new method. The method of countercurrent chromatography using the new horizontal flow-through coil planet centrifuge has been applied to the separation of a spectrum of prostaglandins and their metabolites.

78

The horizontal flow-through coil planet centrifuge [4] used for these experiments carries a pair of coiled separation columns with different modes of planetary motion, each providing a specific advantage for performing countercurrent chromatography. One column (the pulley-side) permits efficient mixing of the two solvent phases to yield highly efficient separations in an analytical scale, while the other column (the gear-side) enables stable retention of the stationary phase in large-bore columns for preparative-scale separations. Both of these columns also allow continuous elution without the use of rotating seals. Consequently, the present apparatus is capable of both small-scale and large-scale separations. In this paper, we describe the separations of prostaglandins and their metabolites achieved by a single two-phase solvent system composed of chloroform, acetic acid and water (2 : 2 : 1, v/v). Partition efficiency of up to a few thousand theoretical plates was achieved, yielding peak resolutions comparable to those attainable with the high-pressure liquid chromatography. Recovery of compounds used in the present method ranged from 80 to 95% in the expected peak and total recovery of all radioactive material when impurities are measured. MATERIALS

All the solvents used for these studies were of the highest purity available from Burdick and Jackson Laboratories. The tritiated prostaglandins PGF2~, PGB2, 13,14-dihydro-15-keto-PGE2, E2, 13,14-dihydro-15-keto-PGF2~ were obtained from the Radiochemical Center, Amersham, as well as tritiated eicosatrienoic acid. Tritiated PGA1, PGA2, PGE2, 6-keto-PGF~, PGD2, carbon-labelled PGF2~ and PGE2, as well as tritiated thromboxane B2 and arachidonic acid, were obtained from the New England Nuclear Corporation. The highest specific activity available was used. All non-labelled compounds which were used as carriers were obtained through the courtesy of Dr. John E. Pike (Upjohn Co.) except for arachidonic acid (Nuchek) and eicosatrienic acid (Sigma). A standard amount of 80 ~g of unlabelled prostaglandin and 0.2 pCi of labelled compound were blown to dryness with nitrogen and taken up in 20--50 ~1 of the mobile phase for application to the column. A two-phase solvent system composed of chloroform/acetic acid/water at a 2 : 2 : 1 (v/v) ratio was thoroughly equilibrated in a separatory funnel at room temperature and separated before use. METHODS

Motion of two immiscible liquids in a rotating helical column under an acceleration field has been described [5--7]. A two-phase solvent system in the column tends to distribute itself in such a way that each phase occupies

79 nearly equal amounts in every turn o f the helical column and any excess of either phase is accumulated at one end o f the column. Once this hydrodynamic equilibrium is reached, further rotation results in mixing of the t w o phases w i t h o u t changing the overall phase distribution throughout the column. Consequently, elution of either phase through the column in the proper direction results in an efficient partition of s o l u t e s between the mobile and the stationary phases within each turn of the column.

Apparatus Fig. 1 shows a photograph of the machine used in this study. The rotary frame of the centrifuge consists of a pair of aluminum plates rigidly linked and driven b y a m o t o r around the hollow stationary shaft fixed at the central axis of the centrifuge. The frame holds a pair of rotary shafts symmetrically spaced at a distance of 15 cm from the center of the apparatus. One rotary shaft is coupled through a pair of idler t o o t h e d pulleys and a belt to a planetary gear which engages an identical stationary gear fixed on the stationary shaft. The second rotary shaft is equipped with a t o o t h e d pulley which is coupled to an identical pulley m o u n t e d around the central stationary shaft. Because of the different geometry of the flow tubes together with the symmetrical orientation of the holder systems, t w o holders are conveniently

Fig. 1. A photograph of the new horizontal flow-through coil planet centrifuge.

80 paired in one apparatus without interfering with the pathway of the flow tubes, as shown in Fig. 2, where one pair of identical gears and one pair of identical pulleys provide a different planetary motion to each column holder. Consequently, t w o types of separations can be carried out simultaneously in both columns, all w i t h o u t the use of rotating seals.

Column preparation The separation columns were prepared by winding PTFE tubing {Zeus Industrial Products, Raritan, N.J.) onto a metal pipe to make a short column or column unit. The long column for microscale separations was prepared from one piece of tubing without interconnection between the column units. Each segment o f the column consisted o f 340 turns of a 0.55 mm i.d. PTFE tubing coiled around a 0.68 cm o.d. core with a total capacity of 2.4 ml. The long column is made b y connecting the desired n u m b e r of column units, usually 10 units, in series for a total volume of 24 ml. The larger bore column consisted of 100 helical turns of a 2.6 mm i.d. PTFE tubing coiled onto each 1.25 cm o.d. metal core with a total capacity of a b o u t 26 ml and with 10 units connected to give a total volume of 260 ml. Revolutional speed of the apparatus is continuously adjustable for 0--600 rev./min. A Chromatronix Cheminert p u m p was e m p l o y e d for elution.

Sample application and collection Samples dissolved in either phase are introduced with a syringe through a septum in the side arm of a T-shaped coupling placed in the input line (PTFE tubing, 0.55 mm i.d.) from the pump. The column is filled with the stationarty phase. After the sample is charged into the line, elution with the mobile phase is started. The displacement of the stationary phase and the initial collection of the mobile phase results in a 'front' which is easily determined. The a m o u n t displayed varies with the flow rate; higher flow rates give greater

Gears

Pulleys Fig. 2. Schematic drawing which illustrates the direction of motion of the two types of columns. The upper column rotates in the same direction as the revolution around the central axis of the apparatus; the lower column rotates counter to the rotation around the central axis which gives more efficient mixing of the two phases. × indicatesthe place where the inlet and outlet tubes are tightly supported on the stationary shaft.

81

displacement o f the stationary phase. Samples were collected at time intervals into counting vials with an Instrumentation Specialties Co. Model 328 fraction collector. 5 ml o f Aquasol (New England Nuclear) scintillation fluid were added to each vial and the a m o u n t of radioactive c o m p o u n d eluted was measured with an ambient temperature scintillation counter. Fractions from the column where the chloroform phase was eluted were blown to dryness under a stream of nitrogen and counted in Aquasol to which 1 ml of glacial acetic acid per 100 ml of solution was added to minimize spurious counts.

Calculation of partition coefficients Partition coefficient (PC) is defined as solute concentration in the mobile phase divided by that in the stationary phase. Using a purified c o m p o u n d , the PC can simply be determined b y a one step extraction process in a separatory funnel. When the sample is a mixture of multiple compounds, the PC of each c o m p o n e n t can be c o m p u t e d from the chromatogram o b t a i n e d by the present m e t h o d according to the following equation: PC = ( V c - Vs)/(VR-- Vs)

(1)

where V¢ denotes the total column capacity, Vs the retention volume of the solvent front, and VR the retention volume of the peak maximum, all expressed in milliliters. Table 1 contains the calculations made from the results of the same column (small-bore) run at t w o different flow rates. Arachidonic acid and eicosatrienoic acid were n o t included in this table, since they would have been retained in the column for extremely long per.iods of time under this set of operative conditions. The partition coefficient, TABLE 1 Partition coefficients Partition coefficients calculated using Eqn. 1 from the experimental results obtained with a small-bore column at two different flow rates and with the upper (acetic acid) phase mobile. Prostaglandin

6-Keto-PGFla Thromboxane B 2 PGF2a PGD2 PGE2 13,!4-Dihydro-15-keto-PGS2a 13,14-Dihydro-15-keto-PGE2 PGB 2 PGA 2

Flow rate 2.4 ml/h

6.0 ml/h

1.0 0.71 0.64 0.53 0.47 0,25 0,22 0.20 --

1.0 0.71 0.67 -0.49 0.34 0.24 0.15 0.15

82

once determined, can be used to predict the location of the solute peak as discussed earlier [7]. The partition coefficient is also useful for the choice of the stationary phase in the present method. In separation of similar compounds, high partition coefficients tend to produce poor peak resolution while extremely low partition coefficients require long elution times without sufficient separation. Ideal partition coefficients m a y range between 1.0 and 0.2 which can usually be obtained by adjusting the phase composition and/or choosing the proper phase as the stationary phase. RESULTS AND DISCUSSION

Fig. 3. shows the structure of the c o m p o u n d s employed in this study, except for arachidonic acid and eicosatrienoic acid. This figure illustrates the slight variations in structure which m a y be separated by this method. Many of these c o m p o u n d s are sensitive to degradation by oxygen so that collection under nitrogen atmosphere would be desirable if the m e t h o d is used for preparative separation of compounds. We purified commercially obtained tritium-labelled arachidonic acid in this manner and used two passes through the column in order to assure the removal of extraneous breakdown products. It has been n o t e d before [8] that thin-layer chromatography is a convenient way of checking on the purity of radiolabelled c o m p o u n d s before use in radioimmunoassay b u t that the minute amounts of c o m p o u n d used decompose rapidly under the conditions of a thin-layer plate. The result is a 'purified' c o m p o u n d containing perhaps an even higher percentage of decomposition products than did the original substance. We would suggest that our m e t h o d can be utilized to purify c o m p o u n d s with the advantage of having no support matrix and having the c o m p o u n d in solvent at all times until it is decided to remove the solvent. The c o m p o u n d s which we obtained from commercial sources varied from a radiopurity of 75--95% even though the stated purity on all samples was greater than 90%. The separation of c o m p o u n d s by this m e t h o d using the smaller column (0.55 mm i.d., 24 ml volume) and a 6 ml/h flow rate is shown in Fig. 4. Smaller fractions, such as were used in Figs. 5 and 6, would have minimized the overlap of PGF2~ and PGE2. As can be seen, satisfactory resolution of 6-keto-PGFl~, TBX2, PGE2, 13,14-dihydro-15-keto-PGF2~ and 13,14-dihydro15-keto-PGE2 and PGB2 is obtained. This m e t h o d would be useful for separating these c o m p o u n d s for subsequent quantitative measurement b y radioimmunoassay. Sample preparation for application to this column should be the same as the preparation for methods using silicic acid columns with mixtures of benzene, methanol and ethyl acetate. Recoveries in this m e t h o d are greater and more predictable than have been reported for some siticic acid column methods [2,3]. Many laboratories separate PGA2 and PGB2, PGE2 and PGF2~ and then convert the PGE2 to PGB2 and run another silicic acid column. With our method, direct conversion of PGE2 to PGB2 before the

83

6-Keto-PGF~o

ThromboxaneB2

~

H

~,,~..~.Av,~"*.~ OOH O

PGFz=

~

~

.

~ N,

f H~ ~ H

~

OOH I 6 Keto PGFzo

II t,

PGDz

OH ~ 50 g

JJTXBz Chloroform:Acetic Acid:

~

u~I

Ez

Water (2:2:1)

UpperPhaseMobile 6 ml perhour AnalyticalSeparation

°

Dihydro-keto-PGFz=

OH ~.~: 30 E z

Dihydro-keto-PGEz

OH

PGBz

OOH

20

~

~

H

OOH

n II

PA

~-

0

PGA=

rlll I,I

i ,

0

20 40 60 80 100 TUBENUMBER=MILLILITERSOFEFFLUENT HOURS2 4 6 8 10 12 14 16 18

Fig. 3. S c h e m a t i c structures o f the c o m p o u n d s used in this study. These c o m p o u n d s are arranged in order o f their increasing lipophilic character in the c h l o r o f o r m / a c e t i c acid/ water (2 : 2 : 1, v/v). Fig. 4. The separation o f prostaglandins at a flow rate o f 6 m l / h with the upper phase (acetic acid) mobile. 1 ml fractions were collected into c o u n t i n g vials. This figure is a c o m p o s i t e o f several runs. When possible, different radioactive labels were used for adjac e n t c o m p o u n d s and t h e y were run on t h e same c o l u m n so that overlap o f the different species could be observed.

initial s e p a r a t i o n s h o u l d b e feasible. Fig. 5 is a c o m p o s i t e o f s e p a r a t i o n s using o n e or m o r e c o m p o u n d s at a t i m e w i t h t h e s a m e c o l u m n as in Fig. 4 b u t w i t h a s l o w e r f l o w r a t e a n d

,

84

0

Chloroform: Acetic Acid: Water (2"9:1)

40

Upper Phase Mobile

I-.-

2.4 ml per hour

z !3o ~

Anelytical Separation PGF2a Q

2o

TXB,!, ~O,

6-Keto-PGFIo 0

~

II

10

Z

i 2O

, I';

/:l

.PGE2

i i!i!l

~___/,I_~.~\,\ 40

~h~,o-~o-~,.

/, Oih,ro-.o- E.

,

,./ ,,.<..'y ' , , ~

60 80 I00 120 140 160 180 200 TUBE NUMBER (0.4 ml samples) TEN MINUTES PER TUBE

--~._ 220

Fig. 5. The separation of prostaglandin using a flow rate of 2.4 m l / h with the upper (acetic acid) phase mobile. Samples o f 0.4 ml were collected into c o u n t i n g vials. The colu m n size (0.55 m m i.d.; 24 ml v o l u m e ) was the same as in Fig. 2, o n l y the flow rate was changed.

smaller (40% of those used before) fractions. Clearly, 6-keto-PGF~, TXB2 and PGD2 would be separated on one run. Also PGF2 and PGE2 would be separated on one run. The degree of overlap of c o m p o u n d s was assessed with c o m p o u n d s labelled with different isotopes (~4C or 3H). The results in Fig. 5 are of the same order and degree of separation that was obtained by Green et al. [9] using HPLC and a reversed-phase column. The advantages of our m e t h o d include less expense, the column is infinitely reusable, and all material applied can be recovered. Fig. 6 shows one such run. The PGF2~ was carbon-labelled, the TXB2 and PGD2 were tritium-labelled. Similar runs were made with carbon-labelled PGE2 and other c o m p o u n d s labelled with tritium. Some alterations of elution conditions will be necessary before TXB2 and PGF2~ can be separated well. In Fig. 7, we have illustrated t w o features of this m e t h o d of separation; the column capacity is 260 ml, 10 times that of the previous figures, and the mobile phase is the lower (chloroform) phase. As would be expected, the order of appearance of the prostaglandins is reversed as compared :with the columns which are run with the upper phase mobile. The PGB2 appears first, PGE2 second and the PGF2a third. Smaller cuts at the point of confluence of the F and E prostaglandins might separate these fractions more clearly. All these c o m p o u n d s were tritium-labelled so that no assessment o f overlap could be made. If it was desirable to convert PGE2 to PGB2 and then to separate the PGB2 for radioimmunoassay with an antibody to B2, the~e conditions would be suitable. Under these conditions b o t h arachidonic acid and eicosatrienoic acid appear at the solvent front and thus are completely separated from these prostaglandins, since the solvent front occurred at fraction

85 Chloroform: Acetic Acid: Water (2:2:1) Upper Phase Mobile 2.4 ml per hour Analytical Separation

5O

¢M 40

30 PGFza

20

PGD=

10

TxB,I I ,,',

0 20 40 60 80 100 TUBE NUMBER (0.4 ml samples) 10 MINUTES PER TUBE Fig. 6. T h e c o n d i t i o n s o f this s e p a r a t i o n w e r e t h e same as f o r Fig. 3. These t h r e e c o m p o u n d s were run simultaneously with the thromboxane and PGD2 (tritium labels) and the

PGF2a (carbon label).

No. 51. For the isolation of metabolites of arachidonic acid or eicosatrienoic acid after an in vitro incubation, it should be possible to elute the column initially with the upper aqueous phase until all the metabolites are eluted and then to switch phases and recover the remaining precursor by elution with the lower nonaqueous phase. Since this system is a closed system, no alterations in structure should take place during the run. The same a m o u n t of c o m p o u n d was used on the large column as for the smaller column. Obviously, more than ten times the a m o u n t could be purified under the conditions of this figure, since the smaller column was n o t overloaded. Sample sizes as large as 10 ml have been used with the preparative column as has been described in the separation of peptides and DNP-amino acids [10]. The partition efficiency of chromatograms obtained by the present method can be expressed in terms of the n u m b e r of theoretical plates using the conventional equation [ 11 ] n = (4R/w)

2

(2)

where R denotes the elution time, and w the peak width. When this equation is applied to the data seen in Fig. 5, the theoretical plates calculated for 6-keto-PGFl~ are 2500 which is attained in 10 h, an estimated 5700 transfers

86 Chloroform: Acetic Acid: Water (2:2:1) Lower Phase Mobile Column: 2.6 mm i.d., 250 ml Capacity

Sample Volume: 150 tJI Revolution: 300 rpm Flow Rate: 24 ml/hr Fraction: 3 ml/bottle Preparative Sel~aration

PGB 6

-

5

-

PGF==

PGE 4

-

3

-

2

-

o

x E ¢L

o

1 -

0J 50

60

70 80 90 FRACTION NUMBER

100

Fig. 7. This illustrates a p r e p a r a t i v e - t y p e s e p a r a t i o n w i t h a c o l u m n o f t e n t i m e s t h e capacity o f t h e p r e v i o u s samples, w i t h t h e l o w e r ( c h l o r o f o r m ) p h a s e m o b i l e , a n d w i t h a f l o w r a t e f o u r t i m e s as great. T h e o r d e r o f a p p e a r a n c e o f p r o s t a g l a n d i n s is c o m p l e t e l y reversed. All t h e c o m p o u n d s were labelled w i t h t r i t i u m so t h a t n o a s s e s s m e n t o f o v e r l a p p i n g c o u l d •b e m a d e .

with a Craig t y p e machine. Similarly the theoretical plates for the c o m p o u n d PGD2 are 1370 within 16 h, and an estimated 5000 transfers in the Craig machine. Clearly, if a countercurrent separation is needed to resolve a mixture, the coil planet centrifuge saves both time and the a m o u n t of labor necessary to collect fractions and to load the Craig machine. SIMPLIFIED DESCRIPTION

OF THE METHOD

AND

ITS A P P L I C A T I O N S

The horizontal flow-through coil planet centrifuge provides a method for the separation of compounds based on their partition coefficients in a two-phase solvent system. This method can be compared to a Craig countercurrent distribution of 1000 to 3000 plates. A m o u n t s from micrograms to several hundred milligrams can be used for either analytical or preparative separation in coils of P T F E tubing with an appropriate twophase solvent system. The advantages of this system over other partition methods are high partition efficiency without the use of solid supports, ease of sample recovery, excellent reproducibility from one experiment to the next, and the peak location can be determined by a simple test tube experiment.

87 Several prostaglandins can be isolated from one sample extraction as has been done with thin-layer chromatography; however, unlike the latter method, oxygen can be omitted easily to protect sensitive compounds and the compounds are easily recovered. It is well suited for the purification of highly radiolabelled prostaglandins either from in-house synthesis or from commercial sources, since the column can be relied on to elute in a reproducible fashion. ACKNOWLEDGEMENT

The authors are deeply indebted to Dr. John E. Pike, The Upjohn Co., for gifts of the unlabelled prostaglandins, to Mr. Howard Chapman for fabrication of the apparatus, and to Miss Sue Engle for the drawings. REFERENCES 1 Yamamoto, M., Herman, E.A. and Rapoport, B. (1979) J. Biol. Chem. 254, 4046-4051 2 Jaffe, B.M., Behrman, H.R. and Parker, C.W. (1973) J. Clin. Invest. 52, 398--405 3 Hubbard, W.C., Watson, J.T. and Sweetmen, B.J. (1979) in Biological/Biomedical Applications of Liquid Chromatography (Hawk, G.L., ed.), pp. 31--55. Marcel Dekker, New York 4 Ito, Y. (1979) Anal. Biochem., in press 5 Ito, Y., Hurst, R.E., Bowman, R.L. and Achter, E.K. (1974) Sep. Purif. Methods 3, 133--165 6 Ito, Y. and Bowman, R.L. (1971) Science 1 7 3 , 4 2 0 - - 4 2 2 7 Ito, Y. and Bowman, R.L. (1973) J. Chromatogr. Sci. 11,284--291 8 Granstrom, E. and Kindahl, H. (1978) Adv. Prostaglandin Thromboxane Res. 5, 119-210 9 Green, K., Hamberg, M., Samuelsson, B. and Frohlich, J.C. (1978) Adv. Prostaglandin Thromboxane Res. 5, 15--38 10 Ito, Y. (1979) J. Chromatogr., in press 11 Keulemans, A.I.M. (1957) Gas Chromatography, p. 113. Reinhold, New York