- Email: [email protected]
HPLC separation of corticosterone metabolites and its use in following corticosterone metabolism and transport in the isolated rat hepatocyte
J. steroid Biochem. Vol. 29, No. 1, pp. 145-148, 1988 Printed in Great Britain. All rights reserved
0022-4731/88 $3.00 + 0.00
Copyright Q 1988Pergamon Journals Ltd
PRELIMINARY NOTE
HPLC SEPARATION OF CORTICOSTERONE METABOLITES AND ITS USE IN FOLLOWING CORTICOSTERONE METABOLISM AND TRANSPORT IN THE ISOLATED RAT HEPATOCYTE STEVE BOTTOMLEY*$, CHRIS LOVELL-SMITH? and PETER GARCIA-WEBB* *Department of Clinical Biochemistry, Queen Elizabeth II Medical Centre, Nedlands 6009, Perth, Western Australia and? Department of Chemical Pathology, University of Otago, Dunedin, New Zealand (Receiued 16 Junuary 1987) Snnnnnry-Incubating isolated rat hepatocytes with tritium labelled and unlabelled corticosterone at 37°C resulted in the rapid appearance of at least nine corticosterone metabolites. A quick, relatively easy, and quantitative high performance liquid chromatographic (HPLC) method was used to separate these metabolites and follow their rate of appearance both intra- and extracellularly. We found different intraand extracellular amounts of each metabolite at a particular time and this suggested that some metabolites were more available for transport than others.
INTRODUCTION It
is known that corticosterone is rapidly taken up, extensively metabolised, and that its more polar metabolites are secreted by rat liver tissue in uivo [l]. In isolated rat hepatocytes, however, corticosterone uptake, metabolism andin particular-secretion (or transport) of its metabolites out of the cell are less well characterised. Recently we have shown, using HPLC, the rapid appearance of both cellular and extracellular corticosterone metabolites in isolated rat hepatocytes [2]. In that study, however, metabolism was investigated over only 10min and the metabolites were not fully separated from one another. Consequently, the aim of this study was to modify the HPLC method to resolve the corticosterone metabolites and, moreover, use the method to follow the rate of appearance of individual metabolites both intra- and extracellularly. EXPERIMENTAL [ 1,2,6,7-‘HlCorticosterone
(84 Ci/mmol) was bought from Amersham Australia Pty Ltd. Aliquots of the steroid solution were dissolved in 5 ~1 of ethylene glycol, after evaporation of the stock solvent under nitrogen, and then taken up in protein-free Krebs-Ringer bicarbonate buffer, pH 7.35. containine 25 mM Henes (KRBH buffer). Unlabelled corticosterone and all otherbiochemicals were’bought from Sigma Chemical Co. St Louis, MO, U.S.A. Hepatocytes were isolated from the livers of male Wista-Furth rats by a three step perfusion based on Seglen’s method [3]. Briefly, the first and second steps were a non-recirculating perfusion with a calcium-free Earles’ Balanced Salts solution buffered with 20mM Hepes (pH 7.40) but the first step also contained 0.63 mM EGTA and 5 U/ml heparin. The third step was a recirculating perfusion with 0.5 g/l collagenase in the same salts buffer as This work was presented, in part, at the 4th FAOB Congress, 30th November-5th December, 1986, Singapore. $To whom correspondence should be addressed. 145
earlier but also containing 5 mM CaCl,. All buffers were gassed with O&O, (95:5; v/v) and entered the liver at 37°C. Hepatocytes were isolated by differential centrifugation at 37°C and finally suspended in KRBH buffer containing 3Og/l bovine serum albumin (fatty acid and globulin-free). Viability, assessed by trypan blue exclusion, was 95% or better. Suspensions of cells OS-l.0 x IO6 cells/ml were incubated at 37°C in 50 ml flasks containing a total of 61 nM labelled (to give about 1.4 x IO6dpm/ml) and unlabelled corticosterone in a total of 5 ml. Final albumin concentration was 2.4g/l. After 1, 10, 20, 35 and 60min incubation, a 0.9ml aliquot of the cell suspension was removed and rapidly filtered: the filter was then washed twice with 4ml of ice-cold saline. Both filter (containing cells) and filtrate (incubation medium) were then separately extracted for metabolites, in saline, by SepPak (Waters Associates, U.S.A.) using C-18 cartridges [2]. A 50 ~1 aliquot of both the cell and incubation medium extract was taken for liquid scintillation counting. The incubation medium extract and the remaining cell extract were then separately analysed by HPLC. A ‘standard’ or ‘control flask, in which buffer replaced cells, was also incubated at 37°C and was used to determine the lability of corticosterone over 60 min. To determine the extent of possible ‘contamination’ of the filtrate (incubation medium) by radioactivity from ruptured cells on the filter the following experiment was carried out: Suspensions of cells were incubated, as above, for 35 min after which a 0.9 ml aliquot was pipetted into test-tubes containing ice-cold saline. The cells were washed 3 times by centrifugation and resuspended in 1 ml of saline. A 0.9 ml aliquot was then rapidly filtered and washed, as earlier, and the filtrate assayed for radioactivity. The calculated amount of radioactivity found in the filtrate (after volume corrections) was assumed to be from ruptured cells. To determine the extent of ‘contamination’ of the filter (or cells) by radioactivity from the incubation medium the following experiment was carried out: Suspensions of cells were incubated, as above but without the labelled and unlabelled corticosterone, at 0°C for at least 60min. Labelled and unlabelled corticosterone was then added and a 0.9 ml aliquot of the cell suspension immediately filtered, and
Preliminary Note
146 100
60 r .=
5
20
1
1
100
a
\
80 15 9
60
40
1 E 8
20
. I----
OI 0
10
20
30
40
50
60
Time(min) Fig. I. Time-course of cell-associated (m) and incubation medium (0) radioactivity. Isolated hepatocytes were incubated with a total of 61 nM tritium-labelled and unlabelled corticosterone over 60min. A 0.9 ml aliquot of the cell suspension was taken at the indicated times, filtered and processed as described in Experimental, and assayed for radioactivity in the cells and in the surrounding incubation medium. The data have been corrected for contamination of both cells and incubation medium (see Experimental), of total radioactivity expressed as a percentage (1187000 _+32600 dpm/0.9 ml aliquot), and are means k SD of four experiments.
Incubation
medium
80
u) 15 = 3 .c 10
RESULTS
There was no break down of labelled corticosterone, at 37”C, over 60 min incubation in buffer without cells (results not shown). The extent of contamination of the filtrate (incubation medium) by radioactivity from cells ruptured during filtering was 10.5 + 1.8% (mean &-SD, n = 4) of the total radioactivity. The extent of contamination of the filter (cells) by radioactivity from the incubation medium was 0.9 k 0.1% (mean k SD, n = 3) of the total radioactivity. Total radioactivity was calculated as the amount of radioactivity in a 0.9 ml aliquot of the cell suspension at a particular time, and was 1,187,OOOf 32,600 dpm (mean f SEM, n = 22). Figure 1 shows that cell-associated radioactivity increased non-linearly with time up to 20min and then decreased slightly by 60 min incubation; this corresponded to an initial decrease, followed by an increase, in the radioactivity of the incubation medium. The data for the figure were corrected for the contamination of both filter and filtrate as described above. Radioactivity in the cells or in the incubation medium was found to be metabolised and unmetabolised corticosterone. Figure 2 shows that corticosterone metabolites could be resolved into nine separate peaks by HPLC. Unmetabolised
E
60 40
5
20
0
0 0
10
20
Fraction washed, as earlier. The filter was resuspended in saline, sonicated [2], and the sonicate assayed for radioactivity. The calculated amount of radioactivity found in the filter sonicate was assumed to be from the incubation medium. Some cell extracts were enzymatically hydrolysed by Helix pomalia extract (1705 U/ml) in 50 mM acetate, pH 5.0, for 18 h at 42°C [12] before further extraction by Sep-Pak and analysis by HPLC. Three successive linear gradients of 0.15 M ammonium acetate @H 5.0 at 25°C) and acetonitrile were used to eluate 50~1 samples from a Spherisorb C-18 5pm column, at 4O”C, on a Hewlett-Packard HP1090 liquid chromatograph. Flow rate was 1.8 ml/min and 0.6 min fractions were collected over 30 min. The gradient was increased from O-20% acetonitrile over the first 3 min after sample injection and then increased from 20-33% for 23 min and lastly from 33-100% over 1 min. Results are expressed as the means of four experiments (n = 4).
extract
s s
30
40
50
number
Fig. 2. HPLC chromatograms of tritiated corticosterone metabolites in cell and incubation medium extracts 35 min after incubating hepatocytes with a total of 61 nM tritium-labelled and unlabelled corticosterone. Resolved metabolites are numbered l-8, and 10. Unmetabolised corticosterone is peak 9. The results are expressed as a percentage of the radioactivity in the cell or incubation medium at 35 min and plotted as the mean chromatogram of four experiments.
corticosterone eluted at peak 9, all metabolites except metabolite 10, were more polar and eluted earlier. The area under the curve for each eluted peak, expressed as a percentage of the radioactivity in the cell or incubation medium extract at a particular time, was used to quantitate the relative amount of the metabolite. Table I shows that the appearance and relative amount of the metabolites depended upon the time of incubation for both cells and incubation medium. There appeared to be a progressive increase in the polarity of the metabolites over time both in cells and incubation medium. Metabolites 1, 2 and 3 increased over 60 min and metabolite 1 appeared to be the major end product of corticosterone metabolism. Metabolite 8 production was maximal over I min and decreased thereafter both in cells and incubation medium. This suggests that metabolite 8 was one of the earliest corticosterone metabolites in the cell. All metabolites appeared in the incubation medium over 60 min indicating their availability for transport. However, some metabolites were more readily transported than others. For example: at 60 min metabolite I and metabolite 3 comprised 26 and 23% respectively of the total radioactivity in the cell; in the incubation medium, however, metabolite 1 comprised 46% and metabolite 3 only I1 % of the total radioactivity. Helix pomatiu extract, which contains both sulfatase and P-glucuronidase, is used to deconjugate steroid sulfates and glucuronides. Enzymatic hydrolysis of the 35min cell extract with Helix pomafia extract significantly decreased the relative amounts of metabolites 1-5 by an average of 72% and concomitantly increased metabolites 6-8, 10 and corticosterone.
Note
Preliminary Table 1. Appearance
of corticosterone
Time (mm) 20
35
60
7.0 rt 2.3 4.4 * 0.4 5.5 + 4.0 18.9 + 4.7 10.4 + 4.0 2.4 * 0.7 2.4 f 0.6 16.4 13.8 29.9 It 2.6 2.1 & 0.7
11.5 +_2.5 4.9 c 0.6 19.2 f 3.9 20.1 i 6.9 10.9 f 2.8 3.0 + 0.8 2.6 i 0.2 12.7 * 3.4 19.9 f 2.1 2.7 + 1.1
16.8 _t 4.8 6.0 f 0.8 19.4 + 6.2 22.6 + 12.2 13.4 i 5.3 4.1 +0.9 2.4 + 0.3 9.2 F 2.5 20.0 * 10.3 2.2 + 0.8
25.8 f 2.0 7.8 + 1.0 22.8 + 1.4 IS.9 it 6.2 12.4 + 3.0 3.8 + 1.2 2.0 f 0.4 6.2 i_ 1.8 14.1 +_7.4 0.9 * 0.2
Incubation medium 1.5 kO.2 1.7 fr 2.7 I.5 +0.4 6.9 t_ 0.6 0.8 fO.l 5.5 rt 1.1 1.4 + 0.2 5.0 + 0.3 1.9 kO.2 3.4 * I .o 1.4kO.l 3.7 & 2.2 l.OkO.1 I.6 rtO.3 7.7 +_2.4 7.6 rfr0.9 89.2 f 5.9 65.5 + 9.0 3.2 F 0.3 2.1 rt 0.3
19.2 f 2.5 10.6 + 0.4 9.2 & 0.6 6.4 k 1.1 4.1 50.9 5.2 +_4.0 1.9ri_0.4 6.0 + 0.5 41.8 k 7.7 2.1 f0.2
31.0 f 3.2 15.2 * 2.3 10.8 + 0.5 6.1 + 1.1 4.2 LtrI .2 5.9 + 5.1 2.1 f0.4 4.9 + 0.6 24.7 f 7.1 1.9kO.9
45.9 * 1.4 16.7 k4.1 ll.0i0.8 5.6 it 0.4 4.6 + 1.6 5.5 +-4.9 l.7kO.3 3.0 f 0.4 10.5 * 3.7 1.0t0.1
1 Peak no.* Cells 2.2 f 0.4 I 3.0 i 0.4 2 3.0 * 1.0 3 7.2 + 0.8 4 7.4 2 5.2 5 2.0 + 0.5 6 2.5 + 0.9 7 30.3 + 5.1 8 48.7 + 1.9 9 5.0 f 0.4 IO
2 3 4 S 6 7 8 9 10
metabolites in cells and incubation medium over 60 min
radioactivity in cells or incubation medium
Peak area-%
1
147
10
*Refer to Fig. 2 for elution profile of peaks. Note also that although no obvious ‘peaks’ were seen for some metabolites at a particular time we assumed that the area under the curve at the expected positions for those peaks represented the amount of those metabolites. The results are expressed as a percentage of the radioactivity, in either cell or incubation medium extract, at a particular time; and are means + SD (n = 4).
DISCUSSION
Corticosterone metabolites have been extracted and separated from rat bile [4J, whole liver [5], and isolated perfused liver [6]. However, the extraction and separation procedures used in these studies, although efficacious, were involved and time consuming. In this study we have described, for the first time, the separation of ~rti~sterone metaboiites from isolated rat hepatocytes by a rapid, relatively easy, and quantitative HPLC method. Using this method the results showed that cell-associated and extracellular radioactivity consisted of corticosterone and at least nine of its metabolites. The progressive increase in the polarity of the metabolites and the effect of Helixpomatia extract on metabolites l-5 agrees with the view that catabolism of steroids leads to more polar compounds [l] and that metabolites l-5 were probably either glucuronide or sulphate conjugates. The extracellular decrease in the percentage of metabolite 8 suggests that it is either recycled back into the cell or metabolised extracellularly, Furthermore, the presence of different intra- and extracellular percentages of metabolites at each time suggested that some metabolites were more available for transport than others. The uptake of corticosterone by hepatocytes is by diffusion and is thought to depend on the solubility of the non-polar steroid in lipid containing cell membranes [ 1I]. Given the marked increase in polarity of the metabolites it is unlikely that the exit from hepatocytes of conjugated metabolites depends on diffusion through a lipid cell membrane. Thus transport out of the cell could be by diffusion through pores in the membrane or by a carrier-mediated process. It was not possible in this study to distinguish between these two mechanisms. Cortisol binding protein and albumin could act as carriers by binding corticosteroid metabolites, or there may be specific binding proteins for the metabolites similar to the proteins identified by Litwack and co-workers which bind metabolites of cortisol[7-91. The rate of transport of a metabolite out of the cell could also be influenced by the availability of the metabolite for transport. Thus, metabolite transport could be restricted by
compartmentation within the cell. In support of this Carlstedt-Duke et a1.[.5] reported differences in the subcellular distribution of corticosterone metabolites in rat liver. The rapid metabolism of corticosterone, and the differences in transport of metabolites out of the cell may contribute to the apparent non-linear uptake of corticosterone as assessed by cell-associated radioactivity seen in this study and previously [Z]. Further work is required to identify the different metabolites and to assess the mechanisms responsible for the different rates of transport of metabolites out of the cell. Acknowledgemenfs-This research was supported by grants from the University of Western Australia and the Oueen Elizabeth II Medical Centre. Steve Bottomley is a recipient of a Biomedical Postgraduate Scholarship from the National Health and Medical Research Council of Australia. We are grateful to Dr Anne Bonser for her helpful comments and to May tan Lee for her secretarial assistance. REFERENCES I.
2. 3. 4.
5.
Samuels L. T. and Eik-Nes K. B.: Metabolism of steroid hormones. In Merubo& Pathways, 3rd cdn (Edited by D. M. Greenberg). Academic Press, New York, Vol. 2 (1968) pp. 169-220. Lovell-Smith C. J. and Garcia-Webb P.: Glucocorticoids and the isolated rat hepatocyte. Biochem. biophys. Res. Commun. 135 119861 160-165. Seglen P. 0.: Preparation of isolated rat liver cells. Meth. Cell Biol. 8 119761 29-81 Chronholm TI: Steroid me&b&m in rats given [l-*Hz] ethanol. Biliary metabolites of corticosterone and administered 4-androstene-3,17 dione. Eur. J. Biochem. ~27 (1972) 10-22. ~rlst~t-Duke J., Gustafsson J.-A. and Gus~f~on S. A.: Sexual differences in hepatic metabolism and
148
Preliminary intracellular distribution of corticosterone studied by pulse labelling with [1,2,6,7-3H]corticosterone. Biochemistry 14 (1975) 639-648. Eriksson H. and Gustafsson J.-A.: Metabolism of corticosterone in the isolated perfused rat liver. Eur. J. Biochem. 20 (1971) 231-236. Litwack G. and Singer S.: Subcellular actions of glucocorticoids. In Biochemical Actions of Hormones (Edited by G. Litwack). Academic Press, New York, Vol. 2 (1972) pp. 113-163. Litwack G., Filler R., Rosenfield S. A., Lichtash N., Wishman C. A. and Singer S.: Liver cytosol corticosteroid binder II, a hormone receptor. J. biol. Chem. 248 (1973) 7481-7486.
Note 9. Steeger J. R. and Litwack G.: Isolation and characterization of binder IIIA a new protein which binds corticosteroid anions. Eur. J. Biochem. 114 (1981) 145-152. 10. Morey K. S. and Litwack G.: Isolation and properties of cortisol metabolite binding proteins of rat liver cytosol. Biochemisrry 8 (1969) 48134821. 11. Giorgi E. P. and Stein W. D.: The transport of steroids into animal cells in culture. Endocrinology 108 (1981) 688-697. 12. Sanghvi A., Wight C., Parikh B. and Desai H.: Urinary 17-hydroxycorticosteroid determination with p-hydrazinobenzenesulfonic acid-phosphoric acid. Am. J. c/in. Path. 60 (1973) 684-690.
0022-4731/88 $3.00 + 0.00
Copyright Q 1988Pergamon Journals Ltd
PRELIMINARY NOTE
HPLC SEPARATION OF CORTICOSTERONE METABOLITES AND ITS USE IN FOLLOWING CORTICOSTERONE METABOLISM AND TRANSPORT IN THE ISOLATED RAT HEPATOCYTE STEVE BOTTOMLEY*$, CHRIS LOVELL-SMITH? and PETER GARCIA-WEBB* *Department of Clinical Biochemistry, Queen Elizabeth II Medical Centre, Nedlands 6009, Perth, Western Australia and? Department of Chemical Pathology, University of Otago, Dunedin, New Zealand (Receiued 16 Junuary 1987) Snnnnnry-Incubating isolated rat hepatocytes with tritium labelled and unlabelled corticosterone at 37°C resulted in the rapid appearance of at least nine corticosterone metabolites. A quick, relatively easy, and quantitative high performance liquid chromatographic (HPLC) method was used to separate these metabolites and follow their rate of appearance both intra- and extracellularly. We found different intraand extracellular amounts of each metabolite at a particular time and this suggested that some metabolites were more available for transport than others.
INTRODUCTION It
is known that corticosterone is rapidly taken up, extensively metabolised, and that its more polar metabolites are secreted by rat liver tissue in uivo [l]. In isolated rat hepatocytes, however, corticosterone uptake, metabolism andin particular-secretion (or transport) of its metabolites out of the cell are less well characterised. Recently we have shown, using HPLC, the rapid appearance of both cellular and extracellular corticosterone metabolites in isolated rat hepatocytes [2]. In that study, however, metabolism was investigated over only 10min and the metabolites were not fully separated from one another. Consequently, the aim of this study was to modify the HPLC method to resolve the corticosterone metabolites and, moreover, use the method to follow the rate of appearance of individual metabolites both intra- and extracellularly. EXPERIMENTAL [ 1,2,6,7-‘HlCorticosterone
(84 Ci/mmol) was bought from Amersham Australia Pty Ltd. Aliquots of the steroid solution were dissolved in 5 ~1 of ethylene glycol, after evaporation of the stock solvent under nitrogen, and then taken up in protein-free Krebs-Ringer bicarbonate buffer, pH 7.35. containine 25 mM Henes (KRBH buffer). Unlabelled corticosterone and all otherbiochemicals were’bought from Sigma Chemical Co. St Louis, MO, U.S.A. Hepatocytes were isolated from the livers of male Wista-Furth rats by a three step perfusion based on Seglen’s method [3]. Briefly, the first and second steps were a non-recirculating perfusion with a calcium-free Earles’ Balanced Salts solution buffered with 20mM Hepes (pH 7.40) but the first step also contained 0.63 mM EGTA and 5 U/ml heparin. The third step was a recirculating perfusion with 0.5 g/l collagenase in the same salts buffer as This work was presented, in part, at the 4th FAOB Congress, 30th November-5th December, 1986, Singapore. $To whom correspondence should be addressed. 145
earlier but also containing 5 mM CaCl,. All buffers were gassed with O&O, (95:5; v/v) and entered the liver at 37°C. Hepatocytes were isolated by differential centrifugation at 37°C and finally suspended in KRBH buffer containing 3Og/l bovine serum albumin (fatty acid and globulin-free). Viability, assessed by trypan blue exclusion, was 95% or better. Suspensions of cells OS-l.0 x IO6 cells/ml were incubated at 37°C in 50 ml flasks containing a total of 61 nM labelled (to give about 1.4 x IO6dpm/ml) and unlabelled corticosterone in a total of 5 ml. Final albumin concentration was 2.4g/l. After 1, 10, 20, 35 and 60min incubation, a 0.9ml aliquot of the cell suspension was removed and rapidly filtered: the filter was then washed twice with 4ml of ice-cold saline. Both filter (containing cells) and filtrate (incubation medium) were then separately extracted for metabolites, in saline, by SepPak (Waters Associates, U.S.A.) using C-18 cartridges [2]. A 50 ~1 aliquot of both the cell and incubation medium extract was taken for liquid scintillation counting. The incubation medium extract and the remaining cell extract were then separately analysed by HPLC. A ‘standard’ or ‘control flask, in which buffer replaced cells, was also incubated at 37°C and was used to determine the lability of corticosterone over 60 min. To determine the extent of possible ‘contamination’ of the filtrate (incubation medium) by radioactivity from ruptured cells on the filter the following experiment was carried out: Suspensions of cells were incubated, as above, for 35 min after which a 0.9 ml aliquot was pipetted into test-tubes containing ice-cold saline. The cells were washed 3 times by centrifugation and resuspended in 1 ml of saline. A 0.9 ml aliquot was then rapidly filtered and washed, as earlier, and the filtrate assayed for radioactivity. The calculated amount of radioactivity found in the filtrate (after volume corrections) was assumed to be from ruptured cells. To determine the extent of ‘contamination’ of the filter (or cells) by radioactivity from the incubation medium the following experiment was carried out: Suspensions of cells were incubated, as above but without the labelled and unlabelled corticosterone, at 0°C for at least 60min. Labelled and unlabelled corticosterone was then added and a 0.9 ml aliquot of the cell suspension immediately filtered, and
Preliminary Note
146 100
60 r .=
5
20
1
1
100
a
\
80 15 9
60
40
1 E 8
20
. I----
OI 0
10
20
30
40
50
60
Time(min) Fig. I. Time-course of cell-associated (m) and incubation medium (0) radioactivity. Isolated hepatocytes were incubated with a total of 61 nM tritium-labelled and unlabelled corticosterone over 60min. A 0.9 ml aliquot of the cell suspension was taken at the indicated times, filtered and processed as described in Experimental, and assayed for radioactivity in the cells and in the surrounding incubation medium. The data have been corrected for contamination of both cells and incubation medium (see Experimental), of total radioactivity expressed as a percentage (1187000 _+32600 dpm/0.9 ml aliquot), and are means k SD of four experiments.
Incubation
medium
80
u) 15 = 3 .c 10
RESULTS
There was no break down of labelled corticosterone, at 37”C, over 60 min incubation in buffer without cells (results not shown). The extent of contamination of the filtrate (incubation medium) by radioactivity from cells ruptured during filtering was 10.5 + 1.8% (mean &-SD, n = 4) of the total radioactivity. The extent of contamination of the filter (cells) by radioactivity from the incubation medium was 0.9 k 0.1% (mean k SD, n = 3) of the total radioactivity. Total radioactivity was calculated as the amount of radioactivity in a 0.9 ml aliquot of the cell suspension at a particular time, and was 1,187,OOOf 32,600 dpm (mean f SEM, n = 22). Figure 1 shows that cell-associated radioactivity increased non-linearly with time up to 20min and then decreased slightly by 60 min incubation; this corresponded to an initial decrease, followed by an increase, in the radioactivity of the incubation medium. The data for the figure were corrected for the contamination of both filter and filtrate as described above. Radioactivity in the cells or in the incubation medium was found to be metabolised and unmetabolised corticosterone. Figure 2 shows that corticosterone metabolites could be resolved into nine separate peaks by HPLC. Unmetabolised
E
60 40
5
20
0
0 0
10
20
Fraction washed, as earlier. The filter was resuspended in saline, sonicated [2], and the sonicate assayed for radioactivity. The calculated amount of radioactivity found in the filter sonicate was assumed to be from the incubation medium. Some cell extracts were enzymatically hydrolysed by Helix pomalia extract (1705 U/ml) in 50 mM acetate, pH 5.0, for 18 h at 42°C [12] before further extraction by Sep-Pak and analysis by HPLC. Three successive linear gradients of 0.15 M ammonium acetate @H 5.0 at 25°C) and acetonitrile were used to eluate 50~1 samples from a Spherisorb C-18 5pm column, at 4O”C, on a Hewlett-Packard HP1090 liquid chromatograph. Flow rate was 1.8 ml/min and 0.6 min fractions were collected over 30 min. The gradient was increased from O-20% acetonitrile over the first 3 min after sample injection and then increased from 20-33% for 23 min and lastly from 33-100% over 1 min. Results are expressed as the means of four experiments (n = 4).
extract
s s
30
40
50
number
Fig. 2. HPLC chromatograms of tritiated corticosterone metabolites in cell and incubation medium extracts 35 min after incubating hepatocytes with a total of 61 nM tritium-labelled and unlabelled corticosterone. Resolved metabolites are numbered l-8, and 10. Unmetabolised corticosterone is peak 9. The results are expressed as a percentage of the radioactivity in the cell or incubation medium at 35 min and plotted as the mean chromatogram of four experiments.
corticosterone eluted at peak 9, all metabolites except metabolite 10, were more polar and eluted earlier. The area under the curve for each eluted peak, expressed as a percentage of the radioactivity in the cell or incubation medium extract at a particular time, was used to quantitate the relative amount of the metabolite. Table I shows that the appearance and relative amount of the metabolites depended upon the time of incubation for both cells and incubation medium. There appeared to be a progressive increase in the polarity of the metabolites over time both in cells and incubation medium. Metabolites 1, 2 and 3 increased over 60 min and metabolite 1 appeared to be the major end product of corticosterone metabolism. Metabolite 8 production was maximal over I min and decreased thereafter both in cells and incubation medium. This suggests that metabolite 8 was one of the earliest corticosterone metabolites in the cell. All metabolites appeared in the incubation medium over 60 min indicating their availability for transport. However, some metabolites were more readily transported than others. For example: at 60 min metabolite I and metabolite 3 comprised 26 and 23% respectively of the total radioactivity in the cell; in the incubation medium, however, metabolite 1 comprised 46% and metabolite 3 only I1 % of the total radioactivity. Helix pomatiu extract, which contains both sulfatase and P-glucuronidase, is used to deconjugate steroid sulfates and glucuronides. Enzymatic hydrolysis of the 35min cell extract with Helix pomafia extract significantly decreased the relative amounts of metabolites 1-5 by an average of 72% and concomitantly increased metabolites 6-8, 10 and corticosterone.
Note
Preliminary Table 1. Appearance
of corticosterone
Time (mm) 20
35
60
7.0 rt 2.3 4.4 * 0.4 5.5 + 4.0 18.9 + 4.7 10.4 + 4.0 2.4 * 0.7 2.4 f 0.6 16.4 13.8 29.9 It 2.6 2.1 & 0.7
11.5 +_2.5 4.9 c 0.6 19.2 f 3.9 20.1 i 6.9 10.9 f 2.8 3.0 + 0.8 2.6 i 0.2 12.7 * 3.4 19.9 f 2.1 2.7 + 1.1
16.8 _t 4.8 6.0 f 0.8 19.4 + 6.2 22.6 + 12.2 13.4 i 5.3 4.1 +0.9 2.4 + 0.3 9.2 F 2.5 20.0 * 10.3 2.2 + 0.8
25.8 f 2.0 7.8 + 1.0 22.8 + 1.4 IS.9 it 6.2 12.4 + 3.0 3.8 + 1.2 2.0 f 0.4 6.2 i_ 1.8 14.1 +_7.4 0.9 * 0.2
Incubation medium 1.5 kO.2 1.7 fr 2.7 I.5 +0.4 6.9 t_ 0.6 0.8 fO.l 5.5 rt 1.1 1.4 + 0.2 5.0 + 0.3 1.9 kO.2 3.4 * I .o 1.4kO.l 3.7 & 2.2 l.OkO.1 I.6 rtO.3 7.7 +_2.4 7.6 rfr0.9 89.2 f 5.9 65.5 + 9.0 3.2 F 0.3 2.1 rt 0.3
19.2 f 2.5 10.6 + 0.4 9.2 & 0.6 6.4 k 1.1 4.1 50.9 5.2 +_4.0 1.9ri_0.4 6.0 + 0.5 41.8 k 7.7 2.1 f0.2
31.0 f 3.2 15.2 * 2.3 10.8 + 0.5 6.1 + 1.1 4.2 LtrI .2 5.9 + 5.1 2.1 f0.4 4.9 + 0.6 24.7 f 7.1 1.9kO.9
45.9 * 1.4 16.7 k4.1 ll.0i0.8 5.6 it 0.4 4.6 + 1.6 5.5 +-4.9 l.7kO.3 3.0 f 0.4 10.5 * 3.7 1.0t0.1
1 Peak no.* Cells 2.2 f 0.4 I 3.0 i 0.4 2 3.0 * 1.0 3 7.2 + 0.8 4 7.4 2 5.2 5 2.0 + 0.5 6 2.5 + 0.9 7 30.3 + 5.1 8 48.7 + 1.9 9 5.0 f 0.4 IO
2 3 4 S 6 7 8 9 10
metabolites in cells and incubation medium over 60 min
radioactivity in cells or incubation medium
Peak area-%
1
147
10
*Refer to Fig. 2 for elution profile of peaks. Note also that although no obvious ‘peaks’ were seen for some metabolites at a particular time we assumed that the area under the curve at the expected positions for those peaks represented the amount of those metabolites. The results are expressed as a percentage of the radioactivity, in either cell or incubation medium extract, at a particular time; and are means + SD (n = 4).
DISCUSSION
Corticosterone metabolites have been extracted and separated from rat bile [4J, whole liver [5], and isolated perfused liver [6]. However, the extraction and separation procedures used in these studies, although efficacious, were involved and time consuming. In this study we have described, for the first time, the separation of ~rti~sterone metaboiites from isolated rat hepatocytes by a rapid, relatively easy, and quantitative HPLC method. Using this method the results showed that cell-associated and extracellular radioactivity consisted of corticosterone and at least nine of its metabolites. The progressive increase in the polarity of the metabolites and the effect of Helixpomatia extract on metabolites l-5 agrees with the view that catabolism of steroids leads to more polar compounds [l] and that metabolites l-5 were probably either glucuronide or sulphate conjugates. The extracellular decrease in the percentage of metabolite 8 suggests that it is either recycled back into the cell or metabolised extracellularly, Furthermore, the presence of different intra- and extracellular percentages of metabolites at each time suggested that some metabolites were more available for transport than others. The uptake of corticosterone by hepatocytes is by diffusion and is thought to depend on the solubility of the non-polar steroid in lipid containing cell membranes [ 1I]. Given the marked increase in polarity of the metabolites it is unlikely that the exit from hepatocytes of conjugated metabolites depends on diffusion through a lipid cell membrane. Thus transport out of the cell could be by diffusion through pores in the membrane or by a carrier-mediated process. It was not possible in this study to distinguish between these two mechanisms. Cortisol binding protein and albumin could act as carriers by binding corticosteroid metabolites, or there may be specific binding proteins for the metabolites similar to the proteins identified by Litwack and co-workers which bind metabolites of cortisol[7-91. The rate of transport of a metabolite out of the cell could also be influenced by the availability of the metabolite for transport. Thus, metabolite transport could be restricted by
compartmentation within the cell. In support of this Carlstedt-Duke et a1.[.5] reported differences in the subcellular distribution of corticosterone metabolites in rat liver. The rapid metabolism of corticosterone, and the differences in transport of metabolites out of the cell may contribute to the apparent non-linear uptake of corticosterone as assessed by cell-associated radioactivity seen in this study and previously [Z]. Further work is required to identify the different metabolites and to assess the mechanisms responsible for the different rates of transport of metabolites out of the cell. Acknowledgemenfs-This research was supported by grants from the University of Western Australia and the Oueen Elizabeth II Medical Centre. Steve Bottomley is a recipient of a Biomedical Postgraduate Scholarship from the National Health and Medical Research Council of Australia. We are grateful to Dr Anne Bonser for her helpful comments and to May tan Lee for her secretarial assistance. REFERENCES I.
2. 3. 4.
5.
Samuels L. T. and Eik-Nes K. B.: Metabolism of steroid hormones. In Merubo& Pathways, 3rd cdn (Edited by D. M. Greenberg). Academic Press, New York, Vol. 2 (1968) pp. 169-220. Lovell-Smith C. J. and Garcia-Webb P.: Glucocorticoids and the isolated rat hepatocyte. Biochem. biophys. Res. Commun. 135 119861 160-165. Seglen P. 0.: Preparation of isolated rat liver cells. Meth. Cell Biol. 8 119761 29-81 Chronholm TI: Steroid me&b&m in rats given [l-*Hz] ethanol. Biliary metabolites of corticosterone and administered 4-androstene-3,17 dione. Eur. J. Biochem. ~27 (1972) 10-22. ~rlst~t-Duke J., Gustafsson J.-A. and Gus~f~on S. A.: Sexual differences in hepatic metabolism and
148
Preliminary intracellular distribution of corticosterone studied by pulse labelling with [1,2,6,7-3H]corticosterone. Biochemistry 14 (1975) 639-648. Eriksson H. and Gustafsson J.-A.: Metabolism of corticosterone in the isolated perfused rat liver. Eur. J. Biochem. 20 (1971) 231-236. Litwack G. and Singer S.: Subcellular actions of glucocorticoids. In Biochemical Actions of Hormones (Edited by G. Litwack). Academic Press, New York, Vol. 2 (1972) pp. 113-163. Litwack G., Filler R., Rosenfield S. A., Lichtash N., Wishman C. A. and Singer S.: Liver cytosol corticosteroid binder II, a hormone receptor. J. biol. Chem. 248 (1973) 7481-7486.
Note 9. Steeger J. R. and Litwack G.: Isolation and characterization of binder IIIA a new protein which binds corticosteroid anions. Eur. J. Biochem. 114 (1981) 145-152. 10. Morey K. S. and Litwack G.: Isolation and properties of cortisol metabolite binding proteins of rat liver cytosol. Biochemisrry 8 (1969) 48134821. 11. Giorgi E. P. and Stein W. D.: The transport of steroids into animal cells in culture. Endocrinology 108 (1981) 688-697. 12. Sanghvi A., Wight C., Parikh B. and Desai H.: Urinary 17-hydroxycorticosteroid determination with p-hydrazinobenzenesulfonic acid-phosphoric acid. Am. J. c/in. Path. 60 (1973) 684-690.