- Email: [email protected]
A simple method for the isolation of vitamin D metabolites from plasma extracts
2881
149
A SIMPLE METHOD FOR THE ISOLATION OF VITAMIN D METABOLITES FROM PLASNA EXTRACTS # A.A. Redhwi+, D.C. Anderson* and G.N. Smith University of Manchester Departments of Chemical Pathology+ and Medicine*, Hope Hospital, Salford M6 BHD and of Chemistry#, Oxford Road, Manchester Ml3 Received 10-26-81 ABSTRACT A simple method has been developed using 'SEP-PAK' disposable silica cartridges to separate the major endogenous vitamin D metabolites, namely vitamin D3, 25-hydroxy vitamin D3 (250HD3), 1,25 dihydroxy vitamin D3 (1,25(OH)2D3) and 24,25 dihydroxyvitamin D3 (24,25 (OHf2D3). After extraction of plasma in isopropanol-toluene (25:75) the dried extract is reconstituted in hexane; this is applied to a SEP-PAK column, and stepwise elution carried out under gravity with 0.1% isopropanol in hexanecneutral lipids), 1% isopropanol in hexane (D3), 3% isopropanol in hexane (250HD3), 3.125% ethanol in dichloromethane (24,25(OH)2D3) and 50% ethanol in toluene (1,25(OHJzD3). Complete separation of these D3 metabolites is achieved by this process and up to 40 samples can be handled at one time. If combined with a suitable ligand binding assay, the system appears to be suitable for preparation of samples prior to the routine assay of vitamin D metabolites. INTRODUCTION Measurement of vitamin D metabolites in human serum presents several problems.
Assay methods for 1,25-dihydroxy vitamin D3 (1,25-
(OH)2D3) still require its extensive purification before radioligand assay, particularly as that sterol circulates in very low concentrations, in the picogram per ml range [Il. Close analogues of this compound, notably 25-hydroxy vitamin D3 (250HD3) and 24,25 dihydroxy vitamin D3 (24,25(OH~2D3) circulate at concentrations much higher than 1,25(OH)2D3 itself. Most current assays require prior extraction, a two-step chromatographic
separation on Sephadex LH-20 and high
performance liquid chromatography
(HPLC) [2,3,41.
We report here a
very simple stepwise separation system using a SEP-PAK silica cartridge, which achieves complete separation of vitamin D 3, 25OHD3, 24,25(OH)2D3 and 1,25(0H)2D3 on one cOllrmn and offers a significant advance aver other separation procedures.
Vdwne
39, Number 2
S
TIlXBOfDU
Febmary,
1982
s
150
WDEOXDI
MATERIALS AND METHODS Tritiated vitamin D3 (TRK 346,7.6 Ci/rsnol), 250HD3 (TRK 558,102 Ci/nunol), 24,25(08)2D3 (TRK 616,77 Ci/mmol) and 1,25(OH)2D3 (TRK 656, 163 Ci/mmol) were purchased from the radiochemical Centre, Amersham. Unlabelled 25OHD3, 24,25 (OH)2D3 and 1,25(OH)2D3 were obtained from Dr. M.Uskokovic, Hoffman La Roche. SEP-PAK silica cartridges (Part No 51900) were purchased from Waters Associates Ltd. The kidney binding protein for 250HD was generously provided by Dr. Barbara Mawer, after preparation by the method of Haddad and co-workers[5]. Foetal rat calvaria for radio-ligand assay of 1,25(OH)2D were prepared from 21-day foetal rats as described previously[6,7]. Solvents were Analar grade (BDH). Aqualuma was obtained from Lumac System AG, Reichenstomstrasse 14, CH 4000, Base1 2. Extraction and Chromatography of Vitamin D-metabolites Extraction:- A new extraction method for vitamin D metabolites was established as follows: An equal volume of saline (0.9%) was added to a serum pool to which had been added about 4000 cpm of vitamin D3 or its relevant metabolite. This was extracted into two volumes of an isopropanol-toluene mixture, which was varied from 25% to 100%; pure toluene was not used because of problems with emulsion formation. Samples were extracted by shaking for one minute, followed by centrifugation at 800x g for 15 minutes, and separation by freezinq the aqueous layer and decantinq the supernatant into polyethylene minivials. Samples were evacuated under a stream of N2 and radioactivity counted in a liquid scintillation Counter (LKB) after addition of 5 ml Aqualuma. The results are shown in figure 1. Isopropanol 25% in toluene was selected; single extractions gave recoveries of 86+3.5% for t3H1-250HD3, 95*3.2% for L3H1-24,25(OH)2D3, and 94?2.7% for [3H1-1,25(OH)2D3 (mean+SEM, n=5). 4oorJRecovery
Radioactivity (cpm)
NJ
3000-
.75
2000.
-50
-25
0’ 25 50 75 100 percentage of isopropanol in toluene 0
Figure 1.
3H 1.25 (OH),O,
0
3 H 24.25 (OH), D,
0
3H250H01
Recovery of tritium-labelled D3 metabolites from plasma.
Chromatography:- As recommended by the manufacturers, 5 ml hexane was first run through the SEP-PAK cartridges which were stoppered for 2 hours prior to chromatography. A suitable mobile phase was established for each vitamin D metabolite by trial and error, using solvent mixtures of increasing polarity and using labelled steroids. It was found that the following stepwise chromatography system was optimal: 10 ml isopropanol 0.1% in hexanecneutral lipids); 20 ml isopropanol 1% in hexane(D3); 20 ml isopropanol 3% in hexane(250HD3); 12 ml ethanol 3.125% in dichloromethane(24,25(OH)2Dj); and 4 ml ethanol 50% in toluene(l,25(0H)2D3). Their performance was then checked with serum extracts; equal amounts of radioactivity of each of the four steroids were added to four aliquots of pooled normal serum which were then extracted as described above. Each extract was dried under a stream of nitrogen, reconstituted in 0.1% isopropanol in hexane(2 ml) and applied to the top of a SEP-PAK cartridge; the cartridges were held in a purposebuilt rack. Solvents were dispensed into the barrel of a glass syringe attached to each SEP-PAK cartridge, and elution carried out under qravity; the elution rate was about 1 ml per minute. The elution profiles of radioactivity, collected in 2 ml fractions(except for 50% ethanol in toluene where six 1 ml fractions were collected)are shown in fit 2.
Figure 2. Elution profiles of tritium-labelled vitamin D3 and its metabolites. The fractions eluted with 0.1% isopropanol in hexane are omitted from the figure. 2 ml fractions were collected except for 50% ethanol in toluene where volume was 1 ml. Finally, a mixture of the three labelled vitamin D metabolites was added to 10 ml of serum from a normal male subject; this was extracted
and chromatographed as described above. One aliquot of each fraction was counted for radioactivity, and another assayed for unlabelled sterol in the appropriate ligand-binding system. Kidney binding protein was used to assay activity of 250HD and 24,25(OH)2D using [3B]-250HD3 as labelled ligand; the binding activity of each fraction was compared to a standard curve of 250HD3 and 24,25(OH)2D3 respectively. An intact calvaria incubation assay was used for 1,25(OH)2D3. Briefly, after preparation as described previously [6,71, four half-calvaria were incubated intact with 20 femtomoles (approx. 3000 cpm) [3Hl-1,25(OH>2D3 in 1 ml, and increasing amounts of unlabelled 1,25(OH)2D3 (O-1000 femtoales) or column eluate, reconstituted in tris-EDTA-dithiothreitol buffer. Incubation at 37OC for 60 min was followed by overnight incubation at 4"C, rinsing with cold saline, homogenisation at -7O'C on a Braun microdismembrator and reconstitution in 1 ml. Two 0.4 ml aliquots of reconstituted homogenate were taken, and 0.5 ml charcoal solution added to the tubes centrifuged at 2500 G for 15 minutes, and an aliquot of the supernatant counted for radioactivity. This detection system gives a standard curve with a sensitivity of about 100 femtomoles per tube, and cross-reactivity only with other vitamin D metabolites. 7OOJ Aadiorctirify (cpm)
n 250HD,
BGQ-
24.25 (OH), 0,
150
. /\
125
6W
24.25 & 1.25 lOti) in famtomole3
I 600
400
2w
0
Volume Ill?, a,uate,
.-@
Radioactivity
O--O
Ligand-binding activity Kidney DBP - for 25 OHD and 24,25(OHl,D Intact calvaria syr~m - for 1.25101ll~D
Figure 3. Elution profiles of radioactivity and ligand-binding activity of 250HD3, 24,25(OHj2D3 and 1,25(OH)2D3 from male plasma. 2 ml fractions were collected except for the 50% ethanol in toluene eluate, which was collected in 1 ml fractions. Results obtained for the three sterols are shown in figure 3; there was complete coincidence of peaks of ligand-binding activity and radioactivity of the three sterols. On this occasion there was a shoulder of radioactivity on the trailing edge of the 24,25(OH)2D3 peak, which may have been due to an impurity in the labelled sterol as it was not found In each case assayable activity fell abruptly to on other occasions.
S
TBEOIDl
153
baseline on either side of the peak. As described previously, this constant activity through the peak is a stringent criterion of the adequacy of a separation step prior to protein binding or radioimmunoassay 191, and suggests that our separation system can be used in combination with any reasonably specific end- assay system for these steroids. As a practical point, we have found that if desired some additional separation of 24,25(08)2D3 from 1,25(OH)2D3 can be obtained by reducing the ethanol content from 3.12% to 2.5% and increasing the volume of the ethanol-CH2Cl2 mixture to 2Oml. DISCUSSION This separation method for vitamin D metabolites appears to be a significant advance for the routine assay of these important compounds. Unfortunately
competitive protein-binding and rsdioimmuno-assays for
1,25(OHJ2D3 require this sterol to be separated from closely related vitamin D metabolites.
Our system provides excellent separation of
l,25(OHj2D3 from 250HD3 and 24,25(OH)2Dj.
Although we have not tested
25,26(OH12D3, another metabolite present in the circul‘ttion, it is probable from its polarity and close similarity to 24,2'j(OH)2D3 that this is also separated from 1,25(~)H!~D3. It
is expected that it would
elute between 24,25(OHj2D3 and 1,25(OH),D3. If necessary the concenL_ tration of ethanol in the ethanol-d'ichloromethane mixture may be decreased slightly in order to increase even further the separation of 24,25(0H12D3, 25,26(OHj2D3 and 1,25(OH)2U3.
It is unlikely that
the system would achieve the separation of D2 metabolites from those of D3*
but this is not usually of clinical importance. Evidence presented
here suggests that the degree of separation that can be achieved by this simple system is similar to that achieved by HPLC better than that on Sephadex LH-20 191.
121 and much
Moreover it requires no
expensive apparatus and can be used to fractionate extracts for 20-40 samples at a time. ACKNOWLEDGEMENTS A.A. Redhwi is supported by a scholarship from the Faculty of Medicine, King Abdul Aziz University, Jeddah, Saudi Arabia.
154
S
?cBBcOXDI
REFERENCES 1.
2. 3. 4. 5. 6. 7.
8. 9.
De Luca, H.F. Advances in Clinical Chemistry 19, 125-174 (1977) Eisman, J.A., Hamstra, A.J., Kream, B.E. and De Luca! H.F. Science 193, 1021-1023 (1976) Clemens, T.L., Hendy, G.N., Papapoulos, S.E., Fraher, L-J., Care, A.D. and O'Riordan, J.L.H. Clin. Endocrinol.~, 225-234 (1979) Lage, A. Clin. Chim. Acta 104, 133-146 (1980) Haddad, J.G. and Chyu, K.J, J. Clin. Endocrinol. Metab. 33, 992-995 (1971) Manolagas, S.C. and Anderson, D.C. J. Endocrinol. E, 379-380 (1978) Manolagas, S.C., Taylor, C.M. and Anderson, D.C. J, Endocrinol, 3, 35-39 (19791 Anderson, D.C., Hopper, B.R., Lasley, B.L. and Yen, S.S.C. Steroids 28, 179-196 (1976) Shimotsujc T., Heijima, T., Seino, T., Yamaoka, K., Ishii, 'I'., Ishida, M., Natsuda, S., Ikehara, C. and yabunchi, H. Clin. Chim. Acta, 106, 145-154 (1980)
149
A SIMPLE METHOD FOR THE ISOLATION OF VITAMIN D METABOLITES FROM PLASNA EXTRACTS # A.A. Redhwi+, D.C. Anderson* and G.N. Smith University of Manchester Departments of Chemical Pathology+ and Medicine*, Hope Hospital, Salford M6 BHD and of Chemistry#, Oxford Road, Manchester Ml3 Received 10-26-81 ABSTRACT A simple method has been developed using 'SEP-PAK' disposable silica cartridges to separate the major endogenous vitamin D metabolites, namely vitamin D3, 25-hydroxy vitamin D3 (250HD3), 1,25 dihydroxy vitamin D3 (1,25(OH)2D3) and 24,25 dihydroxyvitamin D3 (24,25 (OHf2D3). After extraction of plasma in isopropanol-toluene (25:75) the dried extract is reconstituted in hexane; this is applied to a SEP-PAK column, and stepwise elution carried out under gravity with 0.1% isopropanol in hexanecneutral lipids), 1% isopropanol in hexane (D3), 3% isopropanol in hexane (250HD3), 3.125% ethanol in dichloromethane (24,25(OH)2D3) and 50% ethanol in toluene (1,25(OHJzD3). Complete separation of these D3 metabolites is achieved by this process and up to 40 samples can be handled at one time. If combined with a suitable ligand binding assay, the system appears to be suitable for preparation of samples prior to the routine assay of vitamin D metabolites. INTRODUCTION Measurement of vitamin D metabolites in human serum presents several problems.
Assay methods for 1,25-dihydroxy vitamin D3 (1,25-
(OH)2D3) still require its extensive purification before radioligand assay, particularly as that sterol circulates in very low concentrations, in the picogram per ml range [Il. Close analogues of this compound, notably 25-hydroxy vitamin D3 (250HD3) and 24,25 dihydroxy vitamin D3 (24,25(OH~2D3) circulate at concentrations much higher than 1,25(OH)2D3 itself. Most current assays require prior extraction, a two-step chromatographic
separation on Sephadex LH-20 and high
performance liquid chromatography
(HPLC) [2,3,41.
We report here a
very simple stepwise separation system using a SEP-PAK silica cartridge, which achieves complete separation of vitamin D 3, 25OHD3, 24,25(OH)2D3 and 1,25(0H)2D3 on one cOllrmn and offers a significant advance aver other separation procedures.
Vdwne
39, Number 2
S
TIlXBOfDU
Febmary,
1982
s
150
WDEOXDI
MATERIALS AND METHODS Tritiated vitamin D3 (TRK 346,7.6 Ci/rsnol), 250HD3 (TRK 558,102 Ci/nunol), 24,25(08)2D3 (TRK 616,77 Ci/mmol) and 1,25(OH)2D3 (TRK 656, 163 Ci/mmol) were purchased from the radiochemical Centre, Amersham. Unlabelled 25OHD3, 24,25 (OH)2D3 and 1,25(OH)2D3 were obtained from Dr. M.Uskokovic, Hoffman La Roche. SEP-PAK silica cartridges (Part No 51900) were purchased from Waters Associates Ltd. The kidney binding protein for 250HD was generously provided by Dr. Barbara Mawer, after preparation by the method of Haddad and co-workers[5]. Foetal rat calvaria for radio-ligand assay of 1,25(OH)2D were prepared from 21-day foetal rats as described previously[6,7]. Solvents were Analar grade (BDH). Aqualuma was obtained from Lumac System AG, Reichenstomstrasse 14, CH 4000, Base1 2. Extraction and Chromatography of Vitamin D-metabolites Extraction:- A new extraction method for vitamin D metabolites was established as follows: An equal volume of saline (0.9%) was added to a serum pool to which had been added about 4000 cpm of vitamin D3 or its relevant metabolite. This was extracted into two volumes of an isopropanol-toluene mixture, which was varied from 25% to 100%; pure toluene was not used because of problems with emulsion formation. Samples were extracted by shaking for one minute, followed by centrifugation at 800x g for 15 minutes, and separation by freezinq the aqueous layer and decantinq the supernatant into polyethylene minivials. Samples were evacuated under a stream of N2 and radioactivity counted in a liquid scintillation Counter (LKB) after addition of 5 ml Aqualuma. The results are shown in figure 1. Isopropanol 25% in toluene was selected; single extractions gave recoveries of 86+3.5% for t3H1-250HD3, 95*3.2% for L3H1-24,25(OH)2D3, and 94?2.7% for [3H1-1,25(OH)2D3 (mean+SEM, n=5). 4oorJRecovery
Radioactivity (cpm)
NJ
3000-
.75
2000.
-50
-25
0’ 25 50 75 100 percentage of isopropanol in toluene 0
Figure 1.
3H 1.25 (OH),O,
0
3 H 24.25 (OH), D,
0
3H250H01
Recovery of tritium-labelled D3 metabolites from plasma.
Chromatography:- As recommended by the manufacturers, 5 ml hexane was first run through the SEP-PAK cartridges which were stoppered for 2 hours prior to chromatography. A suitable mobile phase was established for each vitamin D metabolite by trial and error, using solvent mixtures of increasing polarity and using labelled steroids. It was found that the following stepwise chromatography system was optimal: 10 ml isopropanol 0.1% in hexanecneutral lipids); 20 ml isopropanol 1% in hexane(D3); 20 ml isopropanol 3% in hexane(250HD3); 12 ml ethanol 3.125% in dichloromethane(24,25(OH)2Dj); and 4 ml ethanol 50% in toluene(l,25(0H)2D3). Their performance was then checked with serum extracts; equal amounts of radioactivity of each of the four steroids were added to four aliquots of pooled normal serum which were then extracted as described above. Each extract was dried under a stream of nitrogen, reconstituted in 0.1% isopropanol in hexane(2 ml) and applied to the top of a SEP-PAK cartridge; the cartridges were held in a purposebuilt rack. Solvents were dispensed into the barrel of a glass syringe attached to each SEP-PAK cartridge, and elution carried out under qravity; the elution rate was about 1 ml per minute. The elution profiles of radioactivity, collected in 2 ml fractions(except for 50% ethanol in toluene where six 1 ml fractions were collected)are shown in fit 2.
Figure 2. Elution profiles of tritium-labelled vitamin D3 and its metabolites. The fractions eluted with 0.1% isopropanol in hexane are omitted from the figure. 2 ml fractions were collected except for 50% ethanol in toluene where volume was 1 ml. Finally, a mixture of the three labelled vitamin D metabolites was added to 10 ml of serum from a normal male subject; this was extracted
and chromatographed as described above. One aliquot of each fraction was counted for radioactivity, and another assayed for unlabelled sterol in the appropriate ligand-binding system. Kidney binding protein was used to assay activity of 250HD and 24,25(OH)2D using [3B]-250HD3 as labelled ligand; the binding activity of each fraction was compared to a standard curve of 250HD3 and 24,25(OH)2D3 respectively. An intact calvaria incubation assay was used for 1,25(OH)2D3. Briefly, after preparation as described previously [6,71, four half-calvaria were incubated intact with 20 femtomoles (approx. 3000 cpm) [3Hl-1,25(OH>2D3 in 1 ml, and increasing amounts of unlabelled 1,25(OH)2D3 (O-1000 femtoales) or column eluate, reconstituted in tris-EDTA-dithiothreitol buffer. Incubation at 37OC for 60 min was followed by overnight incubation at 4"C, rinsing with cold saline, homogenisation at -7O'C on a Braun microdismembrator and reconstitution in 1 ml. Two 0.4 ml aliquots of reconstituted homogenate were taken, and 0.5 ml charcoal solution added to the tubes centrifuged at 2500 G for 15 minutes, and an aliquot of the supernatant counted for radioactivity. This detection system gives a standard curve with a sensitivity of about 100 femtomoles per tube, and cross-reactivity only with other vitamin D metabolites. 7OOJ Aadiorctirify (cpm)
n 250HD,
BGQ-
24.25 (OH), 0,
150
. /\
125
6W
24.25 & 1.25 lOti) in famtomole3
I 600
400
2w
0
Volume Ill?, a,uate,
.-@
Radioactivity
O--O
Ligand-binding activity Kidney DBP - for 25 OHD and 24,25(OHl,D Intact calvaria syr~m - for 1.25101ll~D
Figure 3. Elution profiles of radioactivity and ligand-binding activity of 250HD3, 24,25(OHj2D3 and 1,25(OH)2D3 from male plasma. 2 ml fractions were collected except for the 50% ethanol in toluene eluate, which was collected in 1 ml fractions. Results obtained for the three sterols are shown in figure 3; there was complete coincidence of peaks of ligand-binding activity and radioactivity of the three sterols. On this occasion there was a shoulder of radioactivity on the trailing edge of the 24,25(OH)2D3 peak, which may have been due to an impurity in the labelled sterol as it was not found In each case assayable activity fell abruptly to on other occasions.
S
TBEOIDl
153
baseline on either side of the peak. As described previously, this constant activity through the peak is a stringent criterion of the adequacy of a separation step prior to protein binding or radioimmunoassay 191, and suggests that our separation system can be used in combination with any reasonably specific end- assay system for these steroids. As a practical point, we have found that if desired some additional separation of 24,25(08)2D3 from 1,25(OH)2D3 can be obtained by reducing the ethanol content from 3.12% to 2.5% and increasing the volume of the ethanol-CH2Cl2 mixture to 2Oml. DISCUSSION This separation method for vitamin D metabolites appears to be a significant advance for the routine assay of these important compounds. Unfortunately
competitive protein-binding and rsdioimmuno-assays for
1,25(OHJ2D3 require this sterol to be separated from closely related vitamin D metabolites.
Our system provides excellent separation of
l,25(OHj2D3 from 250HD3 and 24,25(OH)2Dj.
Although we have not tested
25,26(OH12D3, another metabolite present in the circul‘ttion, it is probable from its polarity and close similarity to 24,2'j(OH)2D3 that this is also separated from 1,25(~)H!~D3. It
is expected that it would
elute between 24,25(OHj2D3 and 1,25(OH),D3. If necessary the concenL_ tration of ethanol in the ethanol-d'ichloromethane mixture may be decreased slightly in order to increase even further the separation of 24,25(0H12D3, 25,26(OHj2D3 and 1,25(OH)2U3.
It is unlikely that
the system would achieve the separation of D2 metabolites from those of D3*
but this is not usually of clinical importance. Evidence presented
here suggests that the degree of separation that can be achieved by this simple system is similar to that achieved by HPLC better than that on Sephadex LH-20 191.
121 and much
Moreover it requires no
expensive apparatus and can be used to fractionate extracts for 20-40 samples at a time. ACKNOWLEDGEMENTS A.A. Redhwi is supported by a scholarship from the Faculty of Medicine, King Abdul Aziz University, Jeddah, Saudi Arabia.
154
S
?cBBcOXDI
REFERENCES 1.
2. 3. 4. 5. 6. 7.
8. 9.
De Luca, H.F. Advances in Clinical Chemistry 19, 125-174 (1977) Eisman, J.A., Hamstra, A.J., Kream, B.E. and De Luca! H.F. Science 193, 1021-1023 (1976) Clemens, T.L., Hendy, G.N., Papapoulos, S.E., Fraher, L-J., Care, A.D. and O'Riordan, J.L.H. Clin. Endocrinol.~, 225-234 (1979) Lage, A. Clin. Chim. Acta 104, 133-146 (1980) Haddad, J.G. and Chyu, K.J, J. Clin. Endocrinol. Metab. 33, 992-995 (1971) Manolagas, S.C. and Anderson, D.C. J. Endocrinol. E, 379-380 (1978) Manolagas, S.C., Taylor, C.M. and Anderson, D.C. J, Endocrinol, 3, 35-39 (19791 Anderson, D.C., Hopper, B.R., Lasley, B.L. and Yen, S.S.C. Steroids 28, 179-196 (1976) Shimotsujc T., Heijima, T., Seino, T., Yamaoka, K., Ishii, 'I'., Ishida, M., Natsuda, S., Ikehara, C. and yabunchi, H. Clin. Chim. Acta, 106, 145-154 (1980)