2271
583 SEPARATION OF ESTROGEN CONJUGATES BY HIGH PRESSURE LIQUID CHROMATOGRAPHY Musey, P. 1.1, D. C. Collins, and J. R. K. Preedy, Departments of Medicine and Biochemistry, Emory University School of Medicine, Atlanta, Georgia 30303.
Received:
l-19-78 ABSTRACT
High pressure liquid chromatography using a prepacked commercial strong anion exchanger column (u Partisil 10 SAX, 25 cm x 4.6 mm) was used to separate a mixture of eight estrogen conjugates. Chromatographic conditions using aO.OlM potassium phosphate or 0.1 M NaCl as solvent in the isocratic mode are described for the separation of estrone glucosiduronate, 17B-estradiol-3_glucosiduronate, 17&estradiol17_glucosiduronate, estriol-3_glucosiduronate, estriol-16a-glucosiduronate, estriol-17_glucosiduronate, estrone sulfate and 17B-estradiol3-sulfate. This system gives high resolution of the estrogen conjugates in small eluent volumes in less than 30 min. The advantages of this high pressure liquid chromatographic system over other methods of separation are discussed. INTRODUCTION There are three main classes of estrogen conjugates found in biological fluids: glucosiduronates,
sulfates and sulfoglucosiduronates.
Analysis of complex mixtures of estrogen conjugates requires elaborate preparation and
multiple chromatographies
(2-4).
In 1969, we improved
the original (5) technique of DEAE-Sephadex chromatography of estrogen conjugates using a linear gradient of NaCl (6).
However, dextran or
cellulose based ion exchangers have the disadvantage of being soft easily collapsable gels in which the bed volume changes continuously with changes in pH, ionic strength and pressure.
In this situation,
there is no fixed relationship between elution volume and retention of the estrogen conjugate.
This problem was partly solved by development
of an isocratic mode of separation by Musey -et al (7) which allows calculation of the elution volume of a conjugate from the molarity of the eluent. Though high pressure ion exchange chromatography has been used extensively
Voibne
31,
in the analyses of nucleotides
Nwnber
4
S
(8). very little work has
wnnoltmm
April,
2978
been done with steroid conjugates (g-11). have lately been used with some success
Cellulose ion exchangers
in high speed systems to sepa-
rate organic acids (12) and estrogen conjugates
(13).
With the recent
introduction of rigid permanently bonded supports and the speed advantage provided by high pressure liquid chromatography
(HPLC), we have
explored the possibility of separating complex mixtures of estrogen conjugates normally found in physiological
fluids using a high pressure
system. EXPERIMENTAL Materials Estrogen conjugates used in this study included estrone-3-glucosiduronate (ElG) (14); 17B-estradiol-3-glucosiduronate (E23G); 17Bestradiol-17-glucosiduronate (E217G); estriol-3-glucosiduronate (E33G); estriol-16-glucosiduronate (E316G); estriol-17-glucosiduronate (E317G); estrone sulfate (ElS) and 17B-estradiol-3-sulfate (E2S). All the steroid conjugates were obtained from Sigma Chemical Co., St. Louis, MO. The purity of each steroid conjugate was checked by DEAE-Septadex chromatography (7) followed by sulfuric -acid fluorescence (15).. The only impurity found in some of the samples was free steroid which was removed by ether extraction. Estrogen sulfoglucosiduronate was not available to us at the time of this study. Sodium chloride and potassium dihydrogen phosphate solutions were prepared from analytical grade reagents and deionized distilled.water. The pH of the solutions was adjusted with 1 M acetic or phosphoric acid. The eluents were all filtered throu h a 5 um Millipore Solvinert filter (Millipore Corp., Bedford, Mass. 3 before use. Apparatus A relatively simple liquid chromatograph was assembled from comIt consisted of a mercial modules and regular laboratory equipment. Milton Roy (Philadelphia, Pa.) Simplex high pressure pump with a pulse dampener(Lab Data Control, Riviera Beach, Fl.), an in-line 2 micron stainless steel filter and a Valco six-port rotary injection valve (Glenco Scientific, Houston, TX.). The detection system was an inexpensive dual wave'length (280/254 nm) UV Spectrophotometer with a 18 vl flow cell (Glenco Scientific) connected to a 10 inch flat bed potentiometric recorder (Beckman Instrument, Inc.). When radioactive samples were involved in the chromatography, the effluent was collected directly into scintillation vials on a fraction collector. The columns used in this study were prepacked commercial strong anion exchanger, u Partisil 10 SAX, 25 cm x 4.6 mn (Whatman, Inc., Clifton, N. J.). Two of these columns were often operated in series
S
585
T'mmOIDI
to improveresolution. METHODS AND RESULTS All experiments were carried out in the isocratic mode using potassjum phosphate buffer or sodium chloride solutions.. Phosphate buffer concentration of 0.02 M was found adequate for elution of the estrogen conjugates. Standard solutions of authentic estrogen conjugates were prepared in 20% methanol in distilled water in approximate concentrations of 1 ug/pl. Aliquots of up to 10 pl per sample were injected Znto the column. The elution of some estrogen conjugates with phosphate buffer are shown in Figure 1.
Potassium phosphate concentratims aslow as 0.008
or 0.01 M were enough to elute even EJS.
However, the elution of the
Ez3G COLUMN: y Partisil SAX t 4mm x 25cml MODE: ISOCRATlC ELUENT.: 0.01 M KH,PO, pH4.2
FLOW RATE : 0.4mi /min SAMPLE:
0 Figure I.
5 TIME ~~N*~
STANDARDCONJUGATES
15
High pressure liquid ch~atogr~p~ of 52176, E23G, Elf;and ElS on anion exchange micropartisil column using a phosphate buffer as eluent; inlet pressure = 500 psi.
S
586
TPBEOSDl
conjugates in phosphate buffers tion of some of the conjugates.
too fast to allow adequate resolu-
was
Preliminary runs with sodium chloride
solutions indicated that better separation could be achieved with this solvent system. this eluent.
All subsequent chromatographies
Sodium chloride concentration
were conducted with
of 0.1 M was found to give
maximum resolution of the conjugates. Figure 2 shnws the separation of a mixture of four estrogen conjugates using 0.1 M NaC.1 as the eluent.
This column was operated
at 500 psi with a solvent delivery of 0.4 ml/min. system showed a separation of ElG, ElS, E23G and
This chromatographic
E217G that
was
COLUMN : y Par tisil SAX (4mm x 25crn)
m
:,G
MODE : ISOCRATIC ELUENT FLOW
W
RATE
SAMPLE
: O.IM
N&l
pH 5
: Q4ml/min ; STANDARD
CONJUGATES
.
o-5 Figure 2.
IO TIME Separation of E217G, E23G, ElG and E S by HPLC on anion exchange micropartisil column using aaC1 as eluent. Inlet pressure = 500 psi.
S
TDEOXDI
587
superior to the potassium phosphate system shown in Figure 1. Resolution of the estrogen conjugates could be improved by chromatography on two SAX ion exchange columns in series (Figure 3).
In this
system the back pressure was raised to 1200 psi. This gave a flow rate of approximately 0.8 ml/min. Comparison of Figures 2 and 3 indicates that the retention time of each of the estrogen conjugates was increased significantly in the dual column system leading to a much better separation of the individual compounds. The retention times of the conjugates were found to be consistent and reproducible with standard conjugates. In order to test the effect of urinary residue on the separation of estrogen conjugates in the HPLC system, 50 ml of urine was processed on Amberlite XAD-2 resin as described elsewhere (16). Methanolic solutions of standard estroaen
COLUMN :
Ez17G EIS
Ea E316C E,l7G
MODE:
i
r
Partisil SAX 4mm x 25cm)x2
IS OCRATIC
ELUENT : O.IM Ndl pH 4.8 08 ml/min FLOW RATE: SAMPLE : STANDARD CONJUGATES
s % ol
a 0
‘I‘
0
20
25
TIME t MN -1 Figure 3.
Separation of a test mixture of estrogen conjugates on two micropartisil anion exchange columns connected in series.
ST
588
DEOXDI
conjugates were added to the XAD extract and evaporated to dryness. The residue was redissolved in 2 ml of 20% aqueous methanol and aliquots were injected into the chromatograph.
Table I shows the
retention times of five repetitive injections of the urinary extract. There was insignificant variation in the retention times of the conjugates over several applications.
However, the elution volumes of the
estrogen conjugates were significantly
increased after several months
of repeated use or on applications of heavy extracts from large volumes of urine or plasma. (Whatman
Inc.,
This was obviated by introduction of a pre-column
Clifton, N. J.).
A pre-column
is a short guard column
which protects an analytical column when dirty samples are to be run routinely or when the contaminating
components of the sample may include potentially
compounds.
It is usually packed with pellicular equiva-
lent of the analytical column. We have used HPLC extensively to analyze radioactive metabolites Table I.
Retention times (minutes) of estrogen conjugates added to urine extract. Chromatographic conditions were as described in Figure 3. 1
2
3
4
5
Average
E316G/E317G
8.6
8.6
8.9
9.4
9.6
9.0
E33G
10.1
10.2
10.5
10.6
11.0
10.5
E217G
14.1
14.0
14.4
14.8
15.3
14.5
E23G
14.6
14.6
14.9
15.3
15.1
15.1
ElG
17.5
17.5
17.9
18.1
18.8
18.0
E25
20.0
20.2
20.7
21.0
21.6
20.7
El5
22.4
22.6
23.1
23.8
24.4
23.3
Run
Conjugates
S of estrone in human plasma.
TmPEOfDII
Figure 4 shows a representative chromato-
gram of such analyses using chromatographic described in Figure 3. intervals.
589
conditions similar to those
Radioactive fractions were collected at 20 sec.
The retention times of the radioacitve estrogen conjugates
were very similar to those obtained for the standard compounds in Figure 3.
EIG
15
E217G
12
9
6
3
0 Figure 4.
5
IO
15 20 25 30 TIME (MIN.)
35
Separation of radioactive metabolites in plasma following iv administrationof 3H-El into human subject. Chromatographic conditions were similar to Figure 3 but with a solvent flow rate of 0.6 ml/min at 1000 psi
DISCUSSION The high pressure chromatographic
system that we have described
for the separation of estrogen conjugates is more convenient and faster than conventional methods currently in use.
The separation of
is complete in about 30 min usingHPLC,whereas
these conjugates
tional ion exchange cellulose or Sephadex chromatography
conven-
takes over 18
hrs and requires about 1 liter of eluent (6,7). The sequence of elution of the conjugates on conventional anion exchange Sephadex or high pressure anion exchange cellulose is generally in order of their relative polarities.
However, on the bonded
silica gel columns used for HPLC investigation, reverse order.
they tend to elute in
This reverse sequence was maintained even when we ran
the samples with NaCl solutions without adjusting the pH. therefore that reverse phase chromatography
It seems
on bonded silica gel
columns would be ideal for weak acids like estrogen conjugates. The selectivity and resolution of the estrogen conjugates in this study is superior to ion exchange Sephadex or cellulose chromatographic methods. E317G.
There is, however, insufficient separation of E316G and Other authors (13) have found this to be the case for chroma-
tography on ion exchange cellulose. little consequence
Since E317G is thought to be of
in normal extracts of estrogen conjugates (17)
inability to separate it from E316G is not a major problem. significant to note that none of the chromatographic
It is
procedures
currently in use allow for simultaneous separation of estriol and estradiol-estrone
mixed conjugates, be it either celite partition,
paper, Sephadex LH-20, TLC or ion exchange chromatography. We have already demonstrated that estrogen conjugates can be
,ed rather successfully in isocratic mode on DEAE-Sephadex and ;ellulose systems (7). The excellent resolution we have achieved with isocratic elution in this HPLC is very significant and will reduce the substantial initial investment in HPLC system at least for estrogen analyses. A gradient elution in HPLC requires a complicated solvent delivery system consisting of two separate high pressure pumps interfaced with a sophisticated solvent prograannerand a pressurized mixing chamber.
It also requires a detector with a flow through reference
cell to correct for baseline shifts. The use of NaCl solutions does however have a disadvantage because of the long term corrosive effect of NaCl on steel columns and tubing. van Del Wal -et al (13) have reported at length on the factors and mechanics of separation of estrogen glucosiduronates on a high pressure cellulose system. Our experiences with cellulose ion exchangers using a similar system suggest that sulfates and glucosiduronates are poorly resolved. Therefore, it seems that bonded silica gels are the preferred supports for HPLC separation of steroid conjugates based on selectivity, speed, resolution and resistance to high pressure. The advantages of the present high pressure liquid chromatographic system over other methods currently in use for separating estrogen conjugates are the convenience, speed of analyses and the small volumes of eluent required (approximately 10 ml).
Other advantages of the system
are the use of a rather low concentration of counter ion, it does not require a gradient and it can be used to separate mixed conjugates of estriol, 178-estradiol and estrone simultaneously.
592
S
r?lEIEOfDll AC~O~~LEDGE~&NT
This research was supported by NIH Grants ROl-HL-16394 and h, AM-13468. Dr. Collins is a recipient of U.S. Public Health Service Career Research Development Award l-K04-AM-70381 from the National Institute of Arthritis, DigestSve and Metabolic Diseases.
1. 2. 3. 4. f: 7. 8. 9. 10. ::: 2::
15. if:
Mail reprint requests to: Dr. P. I. Musey, Department of Medicine, Emory University School of Medicine, 69 Butler St., S. E., Atlanta, Georgia 30303. Paul, W., C. Stitt, W. 3. Dignam and S, Kushinsky, 3 CHRO~~TOG 45:381 (1969). Taul, w., c. stftt, W.J. Dignam and S. Kushinsky, J CHROMATOG 45:31j2(1962). Jirku, ii.and M: Levitz, 3 CLIN Er~DOCR~NO~METAEI=:615 (1969). Hahnel, R., ANAL BIOCHEM &184 (j965). tfo+o~rk~_Ri'P. I. Musey.and M. Nilsen, STEROIDS H:191 (1969). 0. C. Collins and J. R. K. Preedy, STEROIDS g:657 (19773. l ' Peterson, E. A+ jn +nratnrv Tpy_hniams __.___. .__...- _ ___ in . -Eliachemistrv ___..-...-___ and -. L ~~~~~ Bioloav. Vo~1 2, T. S. Work and E. Work, eds., Elsevier, i970, p: 228. Id, A. G., 8. A. Lodge and N. J. Pound, J CHROMATGG SC1 Butterfii' 11:401 (1973). Henry, R. A., J. A. Schmit and J. F. Die&man, 3 CHROMATOG SC1 9_z513 (1971). Beyer, W. and W. Korozowich* ANN NEW YORK ACAD SC1 E393 (1968). Mororowich, W. 3 CHROMATOG SC1 12:453 (1974). van Del Wal, S. and J. F. K. HaEr, J CHROMATOG E:353 (1974). The following trivial names and abbreviations for steroid conjugates have been used in this paper: ~strone-3-glucosiduronate (ElG~=l7-oxoestra-l,3,5(lO)-trien-3-y1B-D-glucop>ranosiduronate; 17%estradiol-3-glucosiduronate (E23G)=178-hydroxyestra-1,3,5(10t&en-3-yl-B-D-giucopyranosiduronate; 17B-estradiol-17-glucosiduronate(E217G)= 3- hydroxyestra-1,3,5 ~lO)~trien-178-yl-8-D-glucopyranosiduronate; estriol-3-glucosiduronate.(E33G)=l~,l78-dihydroxyestra-l,3,5~10~trien-3-yl-B-D-glucopyranosiduronafe; estriol-16 .-glucosiduronate(E3166)=3,178-dihydroxyestra-1,3,5 (IO)-trien-16a-yl-B-D-glucopyranosiduronate; estriol-17-glucosiduronate (E3176)=3,16a-dihydroxyestra-1,3,5(10)trien-17$-yl-8-D-~lucopyranosiduronate; 17B-estradiol-3-sulfate (E23S)=17f3-hydroxyestra=l,3,5(lO)-trien3-yl-sulfate; estrone-3-sulfate (E~3S)=17-oxoestra-l,3,5(lO)-trien-3-yl-sulfate. Preedy, 3. R. K. and E. H. Atken, J 610 CHEM 23:1300 (1961). Bradlow, H. t., STEROIDS 1?:265 (1968). Smith, E. R. and A. E. Keme, B~OCHEM J 105:1047 (1967).