- Email: [email protected]
Free haemoglobin interferes with detection of endothelin peptides
Vol, 189, No. 2, 1992 December 15, 1992
BIOCHEMICAL
FREE HAEMOGLOBIN
AND BIOPHYSICAL
INTERFERES
ENDOTHELIN
Anette Hems&
WITH
RESEARCH COMMUNICATIONS Pages 777-781
DETECTION
OF
PEPTIDES
and
Jan M. Lundberg
Dept. of Pharmacology, Karolinska Institute, Box 60 400, S-104 01 Stockholm, Sweden Received
October
6,
1992
SUMMARY: High levels of endothelin-like immunoreactivity were detected in red blood cells from rat, pig and man. When characterized on HPLC the immunoreactivity coeluted with haemoglobin, however. Thus, the high levels of endothelin-like immunoreactivity did not reflect occurrence of endothelin peptides but rather the interference of haemoglobin in the RIA. Free haemoglobin >0.8 g/l (which may occur in haemolytic samples) increased measured plasma “endothelin-like immunoreactivity”. SepPak extraction of plasma samples markedly reduced this interference, although some effect still remained at high haemoglobin concentrations. The influence of microperoxidase in the RIA suggests that the interference is related to the haeme portion of haemoglobin and thus may be extended to other haeme-containing proteins, e.g. cytochrome c or guanylate cyclase. The present findings emphasize the importance of characterizing endothelin-like immunoreactivity with HPLC, especially in haemolytic samples. B 1992Academic Press,Inc.
Endothelin (ET), a peptide consisting of 21 amino acids, was originally isolated from the culture supematant of endothelial cells (1). Since then, ET have also been found in several other cell types, such as renal epithelial cells (2), pulmonary endocrine cells (3), astrocytes (4), neurons (5) and macrofages (6). Genes encoding for three different peptides of the ET family have been found in the rat, porcine and human genoma (7). The peptides were named ET-l, ET-2 and ET-3 of which ET- 1 is the peptide which was first sequenced. ET-l mRNA is widely expressed in rat, guinea-pig, porcine and human tissues. ET-l have a wide range of actions, including bronchoconstriction (8) and growth promotion (9) in addition to vasoconstriction. Elevated plasma levels of ET-like immunoreactivity (LI) has been reported in several pathological conditions, including acute myocardial infarction (lo), cardiogenic shock (11) and patients with uraemia undergoing haemodialysis (12). This study was initiated in order to compare the content of ET in different blood cell fractions of various species, but turned into an attempt to investigate the mechanisms for the interference of haemoglobin (Hb) in ET-RIA.
B1-J
METHODS . . cell fractlonatlon. a Peripheral blood cells were fractionated on Percoll gradients
(Pharmacia, Sweden) essentially according to the recommendations provided by the manufacturer. Briefly, 7 ml blood was collected into Falcon tubes containing 100 IE heparin and EDTA in a final concentration of 7 mM. The blood samples were diluted threefold in phosphate-buffered saline (PBS) ans subsequently layered over a preformed discontinous Percoll gradient. After centrifugation at 400g 0006-291X/92 777
$4.00
Copyright 0 I992 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol.
189,
No.
2,
1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
for 20 min at room temperature, cells were collected at various positions in the gradient according to their density. Platelets were collected at the 1.04-1.053 g/ml interface (less than 0.001% contamination with mononuclear cells as determined by cell counting). Mononuclear cells were collected at a density of 1.062-l .077 @ml. This fraction was slightly contaminated with platelets and contained approximately and equal number of platelets and mononuclear cells. However, the number of platelets in this fraction was four orders of magnitude less than in the highly enriched platelet fraction. Polymorphonuclear cells were collected at the 1.077- 1.09 g/ml interface (slightly contaminated with mononuclear and red blood cells). Erythrocytes were collected as cells pelleting The enriched cell fractions were extracted for 10 min with 20 volumes of 1 M acetic acid at 95oC, homogenized with a Polytron@ and centrifuged. The supernatants were lyophilized and stored at -2oOCuntil analyzed. Plasma samples were extracted with acid ethanol (1.5 ml HCl added to 1 1 ethanol). After centrifugation, the supernatants were dried under a nitrogen. stream at 54oC and stored at -2oOC until analyzed. . Radlolmmunoassav: The dry samples were dissolved in assay buffer (0.1 M phosphate buffer, pH 7.4, containing 0.05 M NaCl, 0.01 % Naazide and 0.1 % bovine serum albumine (BSA)). Plasma samples were concentrated two-fold. The samples and standard concentrations of synthetic ET-l were incubated with antiserum (El, see Hems&t et al 1991) at a final dilution of 1:5000 for 2 days at 4oC. Synthetic ET- 1 labelled with [12X] (Amersham, England) was then added at a concentration of 1000 cpm/ml. The final incubation volume was 300 ~1. The assay was incubated for additional 2 days at 4oC after which the bound and free fractions were separated using a second antibody-method (A-SAC-l, Riact Aps, Denmark). The radioactivity in the samples were counted in a LKB 1272 Gamma Counter (Pharmacia, Sweden) and the concentrations of ET as compared to the standard concentrations were calculated using the Spline function. The influence of increasing concentrations of human and porcine Hb, myoglobin from horse heart, bovine serum albumin (BSA) and microperoxidase-8 (MP-8, the haeme portion of cytochrome c including 8 amino acids) (Sigma, St. Louis, MO, USA) .on the . assay.was investigated. . C Extracted samples were dissolved in distilled water and centrifuged through a 0.45 p.m Millipore UltraFree MC-filter (Millipom, USA). The filtrate was separated on a reverse phase HPLC-column (SuperPac Cartridge packed with Pep-S, 5 l.trn, Pharmacia, Sweden) and eluted with a 40 min linear gradient of 20-50% acetonitrile in 0.1% trifluoroacetic acid. The moblie phase was continously degas& with helium and delivered by a LKB 2249 Gradient Pump (Pharmacia, Sweden) at a flow rate of 1 ml/min. Fractions of 0.5 ml was collected, evaporated and stored at -2OoC until analyzed by RIA. Synthetic ET peptides were used as standards in separate runs and their elution positions were determined by RIA. Purified Hb, myoglobin, BSA and MP-8 were diluted in assay buffer and subjected to RIA as above. The effect of Hb was also analyzed in RIAs for calcitonin gene eptide Y (NPY) (14). A standard curve for Hb in final concentrations of 0.1-3.2 gP was added to heparinized human plasma and pipetted in aliquots of 1 ml. The samples were then subjected to extraction with acid ethanol or purified on Sep-Pak Cartridges (Waters, USA), n=5 for each method and concentration. To reveal the threshold for Hb interference, the samples were then subjected to RIA for ET. Two ml acid ethanol (1.5 ml cont. HCl to 1 1 95% ethanol) was added to 1 ml plasma. The samples were shaken for 1 min, centrifuged at 1200 g for 20 min and the supematants were decanted and evaporated. The plasma samples was acidified by 3 volumes of 4% (v/v) acetic acid (I-IAc) (solution A). The cartridges were pretreated with 5 ml 4% HAc in 86% ethanol (solution B) and 5ml methanol, rinsed with 5ml distilled water and equilibrated with 5 ml solution A. The sample was applied on the cartridge which then was washed twice with 3 ml distilled water. The sample was then. eluted with 3 ml solution B and evaporated in a Savant Vacuum Centrifuge. Statlstlcs:. Statistical differences were calculated using Kruskal-Wallis analysis of variance with multiple comparisons (15).
RESULTS AND DISCUSSION In all species investigated, the content of ET-L1 was lower in the platelet, mononuclear and polymorphonuclear cell fractions than in plasma (‘Iable I). The content of ET-L1 in the red blood cell fraction, however, was surprisingly high in both the rat, pig and man (Table I). 778
Vol.
189,
No.
2,
1992
BIOCHEMICAL
AND
Table
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
I
Levels of ET-L1 in plasma and blood cell fractions from the rat, pig and man. PMNC = polymorphonuclear cells, RBC = red blood cells. Valuesare given as fmol/g. N.D. = not determined Rat
Pig
Human
Plasma Thrombocytes Monocytes
24 ZD.
35 43 2
79 18 43
RBC PMNC
<8 668
506
;:7
However, when the immunoreactivity was characterized on HPLC, it did not coelute with any ET-peptide (Fig 1). Instead, ET-L1 seemed to correlate with the red colour of Hb. In order to further evaluate this phenomenon, we injected purified Hb on the column and found ET-L1 in similar fractions as for the red blood cells. Hence, the ET-assay must in some way be influenced by the presence of Hb. When we analyzed a Hb concentration curve in the ET-assay, it was found that the levels of ET-L1 increased exponentially with the Hb concentration (Fig 2). The threshold Hb concentration for interference was 0.8 g/l. Hb is a haemoprotein with a molecular weight of 64500 Dalton and consists of four protein chains each containing an iron protoporphyrin
as prosthetic group. One
explanation of the effects on the ET assay could be that the antiserum crossreacted with the protein part of Hb. However, when we added MP-8 (the haeme portion of cytochrome C, including 8 amino acids which are not identical to the ones found in Hb) to the assay we found that MP-8 crossreacted almost to equal extent as Hb, when comparing molar concentrations (Fig 3). This indicates that it is the haeme portion of Hb which is responsible for this interference. The Fez+-ion by itself did not influence the ET-assay in similar concentrations as Hb (data not shown). BSA and myoglobin showed 1 and 2 orders of magnitude less interference than Hb, respectively (Fig 3).
20
60
40
Fraction
SO
Number
F~J& ReversedphaseHPLC chromatogram of human red blood cells extracted with acetic acid. ETLI was eluted with a linear gradient of 20-50% acetonitrile in 0.1% trifluoroacetic of purified Hb is shown (fractions 69-80). 779
acid. The position
Vol.
02
189,
No. 2, 1992
;01
BIOCHEMICAL
1
,I Hb
10
AND BIOPHYSICAL
100
0
RESEARCH COMMUNICATIONS
2ot 10'
3
(dl)
Fig 2. Influence of increasing concentrationsof haemoglobin (Hb) on levels of “ET-LI” in RIA. Data are given asmeansf SEM from 5 separateexperimentsperformed in duplicate. FJ& Specific binding of W-ET- 1 to antisernmE- 1with increasingconcentrationsof haemoglobin (Hb), myoglobin (Myo), microperoxidase 8 (MP-8) and bovine serum albumin (BSA). Data are given as mean&EM, n=4- 10.
That Hb interferes with the detection of ET may have severe consequences on the results when analyzing haemolytic samples. However, when we compared different extraction procedures we found that this problem was much less pronounced when using SepPak purification compared to ethanol extraction. In ethanol extracted plasma the levels of ET-L1 were significantly increased at Hb concentrations of 1.6 g/l , whereas in SepPak purified samples increased levels of ET-L1 was seen first at 5.0 g/l Hb (Fig 4). Hb also interfered with the detection of CGRP-LI, almost to the same extent as for ET whereas the NPY assay was less sensitive (Fig 5). Possibly this phenomenon is related to the presence of disulphide bridges within the peptide structure, since both ET and CGRP contain
120 0
f ao3 E 60.
q
--t -
EtOH SepPak
ET-RI.4 CCRP-RIA NPY.RIA
,P
=;
40.
5,
200
0
4
60
T ND ND b .“”
0
ND NO ’ ““,
0.8
Hb
z n
ND v”1
1.6
,O,%
k
3.2
Concentration
5.0
10
Hb
(g/l)
Concentration
(g/l)
Fig 4. Content of “ET-LI” in Hb-containing plasma sampleswhich have been subjectedto ethanol extraction (EtOH) or SepPakpurification. Values are given as means f SEM, n=5. ** pcO.01, ***
p
FJ@ Influence of increasing haemoglobin concentrations on specific binding in radioimmunoassaysfor ET, CGRP and NPY. Data are given as mean&EM of 3-5 experiments performed in duplicate. 780
Vo4.
189,
disulphide
No.
2,
1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
bridges (ET contains two) whereas NPY does not. Additional
COMMUNICATIONS
studies should be
performed, however, to further characterize this interference. In conclusion, falsely elevated levels of “ET-LI” are obtained when measuring severely haemolytic plasma samples. This interference could in some extent be avoided by purification of the samples on ClS-cartridges, such as SepPak.
ACKNOWLEDGMENTS The present study was supported by Wibergs Stiftelse, funds from the Karolinska Institute and the Swedish Medical Research Council1 (14X-6554). For expert technical assistance we thank MS Carina Siiderblom and MS Margareta Stensdotter.
REFERENCES 1. Yanagisawa, M., Kurihara, H., Kimura, S., Tomobe, Y., Kobayashi, M., Mitsui, Y., Yazaki, Y., Goto, K. and Masaki, T. (1988) Nature (Lond), 332,411-415. 2. Shichiri, M., Hirata, Y., Emori, T., Ohta, K., Nakajima, T., Sato, K., Sato, A. and Marumo, F. (1989) FEBS Lett. 253,203-206. 3. Giaid, A., Polak, J.M., Gaitonde, V., Hamid, Q.A., Moscoso, G., Legon, S., Uwanogho, D., Roncalli, M., Shinmi, O., Sawamura, T., Kimura, S., Yanagisawa, M., Masaki, and Springall, D.R. (1991) Am. J. Respir. Cell Mol. Biol. 4, 50-58. 4. Ehrenreich, H., Kehrl, J.H., Andersson, R.W., Rieckmann, P., Vitkovic, L., Coligan, J.E. and Fauci, A.S. (1991) Brain Res. 538, 54-58. 5. Giaid, A., Gibson, S.J., Ibrahim, N.B.N., Legon, S., Bloom, S.R., Yanagisawa, M., Masaki, T., Vamdell, M. and Polak, J.M. (1989) Proc. Natl. Acad. Sci. USA, 86, 7634-7638. 6. Ehrenreich, H., Anderson, R.W., Fox, C.H., Rieckmann, P., Hoffman, G.S., Travis, W.D., Coligan, J.E., Kehrl, J.H. and Fauci, A.S. (1990) J. Exp. Med. 172, 1741-1748. 7. Inoue, A., Yanagisawa, M., Kimura, S., Kasuya, Y., Miyauchi, T., Goto, K. and Masaki, T. (1989) Proc. Natl. Acad. Sci. USA, 86,2863-2867. 8. Uchida, Y., Ninomiya, H., Saotome, M., Nomura, A., Ohtsuka, M., Yanagisawa, M., Goto, K., Masaki, T. and Hasegawa, S. (1988) Eur. J. Pharmacol. 154,227-228. 9. Komuro, I., Kurihara, H., Sugiyama, T., Takuku, F. and Yazai, Y. (1988) FEBS Lett. 238, 249-252. 10. Miyauchi, T., Yanagisawa, M., Tomizawa, T., Sugishita, Y., Suzuki, N., Fujino, M.. Ajisaka, R., Goto, K. and Masaki, T. Lancet, 8653, 53-54. 11. Cemacek, P. and Stewart, D.J. (1989) Biochem. Biophys. Res. Commun. 161, 562-567. 12. Koyama, H., Nishzawa, Y., Mot-ii, H., Tabata, T., Inoue, T. and Yamaji, T. (1989) Lancet, 8645, 99 l-992. 13. France-Cereceda, A., Saria, A. and Lundberg, J.M. (1989) Acta Physiol. Stand. 135, 173-187. 14. Theodorsson-Norheim, E., A. Hems& and Lundberg, J.M. (1985) Stand. J. Clin. Lab. Invest. 45, 355-365. 15. Theodorsson-Norheim, E. (1986) Computer Programs Biomed. 23,57-62.
781
BIOCHEMICAL
FREE HAEMOGLOBIN
AND BIOPHYSICAL
INTERFERES
ENDOTHELIN
Anette Hems&
WITH
RESEARCH COMMUNICATIONS Pages 777-781
DETECTION
OF
PEPTIDES
and
Jan M. Lundberg
Dept. of Pharmacology, Karolinska Institute, Box 60 400, S-104 01 Stockholm, Sweden Received
October
6,
1992
SUMMARY: High levels of endothelin-like immunoreactivity were detected in red blood cells from rat, pig and man. When characterized on HPLC the immunoreactivity coeluted with haemoglobin, however. Thus, the high levels of endothelin-like immunoreactivity did not reflect occurrence of endothelin peptides but rather the interference of haemoglobin in the RIA. Free haemoglobin >0.8 g/l (which may occur in haemolytic samples) increased measured plasma “endothelin-like immunoreactivity”. SepPak extraction of plasma samples markedly reduced this interference, although some effect still remained at high haemoglobin concentrations. The influence of microperoxidase in the RIA suggests that the interference is related to the haeme portion of haemoglobin and thus may be extended to other haeme-containing proteins, e.g. cytochrome c or guanylate cyclase. The present findings emphasize the importance of characterizing endothelin-like immunoreactivity with HPLC, especially in haemolytic samples. B 1992Academic Press,Inc.
Endothelin (ET), a peptide consisting of 21 amino acids, was originally isolated from the culture supematant of endothelial cells (1). Since then, ET have also been found in several other cell types, such as renal epithelial cells (2), pulmonary endocrine cells (3), astrocytes (4), neurons (5) and macrofages (6). Genes encoding for three different peptides of the ET family have been found in the rat, porcine and human genoma (7). The peptides were named ET-l, ET-2 and ET-3 of which ET- 1 is the peptide which was first sequenced. ET-l mRNA is widely expressed in rat, guinea-pig, porcine and human tissues. ET-l have a wide range of actions, including bronchoconstriction (8) and growth promotion (9) in addition to vasoconstriction. Elevated plasma levels of ET-like immunoreactivity (LI) has been reported in several pathological conditions, including acute myocardial infarction (lo), cardiogenic shock (11) and patients with uraemia undergoing haemodialysis (12). This study was initiated in order to compare the content of ET in different blood cell fractions of various species, but turned into an attempt to investigate the mechanisms for the interference of haemoglobin (Hb) in ET-RIA.
B1-J
METHODS . . cell fractlonatlon. a Peripheral blood cells were fractionated on Percoll gradients
(Pharmacia, Sweden) essentially according to the recommendations provided by the manufacturer. Briefly, 7 ml blood was collected into Falcon tubes containing 100 IE heparin and EDTA in a final concentration of 7 mM. The blood samples were diluted threefold in phosphate-buffered saline (PBS) ans subsequently layered over a preformed discontinous Percoll gradient. After centrifugation at 400g 0006-291X/92 777
$4.00
Copyright 0 I992 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol.
189,
No.
2,
1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
for 20 min at room temperature, cells were collected at various positions in the gradient according to their density. Platelets were collected at the 1.04-1.053 g/ml interface (less than 0.001% contamination with mononuclear cells as determined by cell counting). Mononuclear cells were collected at a density of 1.062-l .077 @ml. This fraction was slightly contaminated with platelets and contained approximately and equal number of platelets and mononuclear cells. However, the number of platelets in this fraction was four orders of magnitude less than in the highly enriched platelet fraction. Polymorphonuclear cells were collected at the 1.077- 1.09 g/ml interface (slightly contaminated with mononuclear and red blood cells). Erythrocytes were collected as cells pelleting The enriched cell fractions were extracted for 10 min with 20 volumes of 1 M acetic acid at 95oC, homogenized with a Polytron@ and centrifuged. The supernatants were lyophilized and stored at -2oOCuntil analyzed. Plasma samples were extracted with acid ethanol (1.5 ml HCl added to 1 1 ethanol). After centrifugation, the supernatants were dried under a nitrogen. stream at 54oC and stored at -2oOC until analyzed. . Radlolmmunoassav: The dry samples were dissolved in assay buffer (0.1 M phosphate buffer, pH 7.4, containing 0.05 M NaCl, 0.01 % Naazide and 0.1 % bovine serum albumine (BSA)). Plasma samples were concentrated two-fold. The samples and standard concentrations of synthetic ET-l were incubated with antiserum (El, see Hems&t et al 1991) at a final dilution of 1:5000 for 2 days at 4oC. Synthetic ET- 1 labelled with [12X] (Amersham, England) was then added at a concentration of 1000 cpm/ml. The final incubation volume was 300 ~1. The assay was incubated for additional 2 days at 4oC after which the bound and free fractions were separated using a second antibody-method (A-SAC-l, Riact Aps, Denmark). The radioactivity in the samples were counted in a LKB 1272 Gamma Counter (Pharmacia, Sweden) and the concentrations of ET as compared to the standard concentrations were calculated using the Spline function. The influence of increasing concentrations of human and porcine Hb, myoglobin from horse heart, bovine serum albumin (BSA) and microperoxidase-8 (MP-8, the haeme portion of cytochrome c including 8 amino acids) (Sigma, St. Louis, MO, USA) .on the . assay.was investigated. . C Extracted samples were dissolved in distilled water and centrifuged through a 0.45 p.m Millipore UltraFree MC-filter (Millipom, USA). The filtrate was separated on a reverse phase HPLC-column (SuperPac Cartridge packed with Pep-S, 5 l.trn, Pharmacia, Sweden) and eluted with a 40 min linear gradient of 20-50% acetonitrile in 0.1% trifluoroacetic acid. The moblie phase was continously degas& with helium and delivered by a LKB 2249 Gradient Pump (Pharmacia, Sweden) at a flow rate of 1 ml/min. Fractions of 0.5 ml was collected, evaporated and stored at -2OoC until analyzed by RIA. Synthetic ET peptides were used as standards in separate runs and their elution positions were determined by RIA. Purified Hb, myoglobin, BSA and MP-8 were diluted in assay buffer and subjected to RIA as above. The effect of Hb was also analyzed in RIAs for calcitonin gene eptide Y (NPY) (14). A standard curve for Hb in final concentrations of 0.1-3.2 gP was added to heparinized human plasma and pipetted in aliquots of 1 ml. The samples were then subjected to extraction with acid ethanol or purified on Sep-Pak Cartridges (Waters, USA), n=5 for each method and concentration. To reveal the threshold for Hb interference, the samples were then subjected to RIA for ET. Two ml acid ethanol (1.5 ml cont. HCl to 1 1 95% ethanol) was added to 1 ml plasma. The samples were shaken for 1 min, centrifuged at 1200 g for 20 min and the supematants were decanted and evaporated. The plasma samples was acidified by 3 volumes of 4% (v/v) acetic acid (I-IAc) (solution A). The cartridges were pretreated with 5 ml 4% HAc in 86% ethanol (solution B) and 5ml methanol, rinsed with 5ml distilled water and equilibrated with 5 ml solution A. The sample was applied on the cartridge which then was washed twice with 3 ml distilled water. The sample was then. eluted with 3 ml solution B and evaporated in a Savant Vacuum Centrifuge. Statlstlcs:. Statistical differences were calculated using Kruskal-Wallis analysis of variance with multiple comparisons (15).
RESULTS AND DISCUSSION In all species investigated, the content of ET-L1 was lower in the platelet, mononuclear and polymorphonuclear cell fractions than in plasma (‘Iable I). The content of ET-L1 in the red blood cell fraction, however, was surprisingly high in both the rat, pig and man (Table I). 778
Vol.
189,
No.
2,
1992
BIOCHEMICAL
AND
Table
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
I
Levels of ET-L1 in plasma and blood cell fractions from the rat, pig and man. PMNC = polymorphonuclear cells, RBC = red blood cells. Valuesare given as fmol/g. N.D. = not determined Rat
Pig
Human
Plasma Thrombocytes Monocytes
24 ZD.
35 43 2
79 18 43
RBC PMNC
<8 668
506
;:7
However, when the immunoreactivity was characterized on HPLC, it did not coelute with any ET-peptide (Fig 1). Instead, ET-L1 seemed to correlate with the red colour of Hb. In order to further evaluate this phenomenon, we injected purified Hb on the column and found ET-L1 in similar fractions as for the red blood cells. Hence, the ET-assay must in some way be influenced by the presence of Hb. When we analyzed a Hb concentration curve in the ET-assay, it was found that the levels of ET-L1 increased exponentially with the Hb concentration (Fig 2). The threshold Hb concentration for interference was 0.8 g/l. Hb is a haemoprotein with a molecular weight of 64500 Dalton and consists of four protein chains each containing an iron protoporphyrin
as prosthetic group. One
explanation of the effects on the ET assay could be that the antiserum crossreacted with the protein part of Hb. However, when we added MP-8 (the haeme portion of cytochrome C, including 8 amino acids which are not identical to the ones found in Hb) to the assay we found that MP-8 crossreacted almost to equal extent as Hb, when comparing molar concentrations (Fig 3). This indicates that it is the haeme portion of Hb which is responsible for this interference. The Fez+-ion by itself did not influence the ET-assay in similar concentrations as Hb (data not shown). BSA and myoglobin showed 1 and 2 orders of magnitude less interference than Hb, respectively (Fig 3).
20
60
40
Fraction
SO
Number
F~J& ReversedphaseHPLC chromatogram of human red blood cells extracted with acetic acid. ETLI was eluted with a linear gradient of 20-50% acetonitrile in 0.1% trifluoroacetic of purified Hb is shown (fractions 69-80). 779
acid. The position
Vol.
02
189,
No. 2, 1992
;01
BIOCHEMICAL
1
,I Hb
10
AND BIOPHYSICAL
100
0
RESEARCH COMMUNICATIONS
2ot 10'
3
(dl)
Fig 2. Influence of increasing concentrationsof haemoglobin (Hb) on levels of “ET-LI” in RIA. Data are given asmeansf SEM from 5 separateexperimentsperformed in duplicate. FJ& Specific binding of W-ET- 1 to antisernmE- 1with increasingconcentrationsof haemoglobin (Hb), myoglobin (Myo), microperoxidase 8 (MP-8) and bovine serum albumin (BSA). Data are given as mean&EM, n=4- 10.
That Hb interferes with the detection of ET may have severe consequences on the results when analyzing haemolytic samples. However, when we compared different extraction procedures we found that this problem was much less pronounced when using SepPak purification compared to ethanol extraction. In ethanol extracted plasma the levels of ET-L1 were significantly increased at Hb concentrations of 1.6 g/l , whereas in SepPak purified samples increased levels of ET-L1 was seen first at 5.0 g/l Hb (Fig 4). Hb also interfered with the detection of CGRP-LI, almost to the same extent as for ET whereas the NPY assay was less sensitive (Fig 5). Possibly this phenomenon is related to the presence of disulphide bridges within the peptide structure, since both ET and CGRP contain
120 0
f ao3 E 60.
q
--t -
EtOH SepPak
ET-RI.4 CCRP-RIA NPY.RIA
,P
=;
40.
5,
200
0
4
60
T ND ND b .“”
0
ND NO ’ ““,
0.8
Hb
z n
ND v”1
1.6
,O,%
k
3.2
Concentration
5.0
10
Hb
(g/l)
Concentration
(g/l)
Fig 4. Content of “ET-LI” in Hb-containing plasma sampleswhich have been subjectedto ethanol extraction (EtOH) or SepPakpurification. Values are given as means f SEM, n=5. ** pcO.01, ***
p
FJ@ Influence of increasing haemoglobin concentrations on specific binding in radioimmunoassaysfor ET, CGRP and NPY. Data are given as mean&EM of 3-5 experiments performed in duplicate. 780
Vo4.
189,
disulphide
No.
2,
1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
bridges (ET contains two) whereas NPY does not. Additional
COMMUNICATIONS
studies should be
performed, however, to further characterize this interference. In conclusion, falsely elevated levels of “ET-LI” are obtained when measuring severely haemolytic plasma samples. This interference could in some extent be avoided by purification of the samples on ClS-cartridges, such as SepPak.
ACKNOWLEDGMENTS The present study was supported by Wibergs Stiftelse, funds from the Karolinska Institute and the Swedish Medical Research Council1 (14X-6554). For expert technical assistance we thank MS Carina Siiderblom and MS Margareta Stensdotter.
REFERENCES 1. Yanagisawa, M., Kurihara, H., Kimura, S., Tomobe, Y., Kobayashi, M., Mitsui, Y., Yazaki, Y., Goto, K. and Masaki, T. (1988) Nature (Lond), 332,411-415. 2. Shichiri, M., Hirata, Y., Emori, T., Ohta, K., Nakajima, T., Sato, K., Sato, A. and Marumo, F. (1989) FEBS Lett. 253,203-206. 3. Giaid, A., Polak, J.M., Gaitonde, V., Hamid, Q.A., Moscoso, G., Legon, S., Uwanogho, D., Roncalli, M., Shinmi, O., Sawamura, T., Kimura, S., Yanagisawa, M., Masaki, and Springall, D.R. (1991) Am. J. Respir. Cell Mol. Biol. 4, 50-58. 4. Ehrenreich, H., Kehrl, J.H., Andersson, R.W., Rieckmann, P., Vitkovic, L., Coligan, J.E. and Fauci, A.S. (1991) Brain Res. 538, 54-58. 5. Giaid, A., Gibson, S.J., Ibrahim, N.B.N., Legon, S., Bloom, S.R., Yanagisawa, M., Masaki, T., Vamdell, M. and Polak, J.M. (1989) Proc. Natl. Acad. Sci. USA, 86, 7634-7638. 6. Ehrenreich, H., Anderson, R.W., Fox, C.H., Rieckmann, P., Hoffman, G.S., Travis, W.D., Coligan, J.E., Kehrl, J.H. and Fauci, A.S. (1990) J. Exp. Med. 172, 1741-1748. 7. Inoue, A., Yanagisawa, M., Kimura, S., Kasuya, Y., Miyauchi, T., Goto, K. and Masaki, T. (1989) Proc. Natl. Acad. Sci. USA, 86,2863-2867. 8. Uchida, Y., Ninomiya, H., Saotome, M., Nomura, A., Ohtsuka, M., Yanagisawa, M., Goto, K., Masaki, T. and Hasegawa, S. (1988) Eur. J. Pharmacol. 154,227-228. 9. Komuro, I., Kurihara, H., Sugiyama, T., Takuku, F. and Yazai, Y. (1988) FEBS Lett. 238, 249-252. 10. Miyauchi, T., Yanagisawa, M., Tomizawa, T., Sugishita, Y., Suzuki, N., Fujino, M.. Ajisaka, R., Goto, K. and Masaki, T. Lancet, 8653, 53-54. 11. Cemacek, P. and Stewart, D.J. (1989) Biochem. Biophys. Res. Commun. 161, 562-567. 12. Koyama, H., Nishzawa, Y., Mot-ii, H., Tabata, T., Inoue, T. and Yamaji, T. (1989) Lancet, 8645, 99 l-992. 13. France-Cereceda, A., Saria, A. and Lundberg, J.M. (1989) Acta Physiol. Stand. 135, 173-187. 14. Theodorsson-Norheim, E., A. Hems& and Lundberg, J.M. (1985) Stand. J. Clin. Lab. Invest. 45, 355-365. 15. Theodorsson-Norheim, E. (1986) Computer Programs Biomed. 23,57-62.
781