- Email: [email protected]
CD13 in serum from cholestatic patients
Clinica Chimica Acta 330 (2003) 141 – 149 www.elsevier.com/locate/clinchim
High-molecular-mass isoform of aminopeptidase N/CD13 in serum from cholestatic patients Makoto Kawai *, Yuya Otake, Yukichi Hara Department of Biochemistry and Biophysics, Graduate School of Allied Health Sciences, Tokyo Medical and Dental University, 5-45, Yushima 1-chome, Bunkyo-ku, Tokyo 113-8519, Japan Received 15 October 2002; received in revised form 19 December 2002; accepted 20 December 2002
Abstract Background: Because non-denaturing electrophoresis and aminopeptidase activity staining often detect noncovalent multienzyme complexes, we adopted procedures to specifically detect the aminopeptidase N (APN) molecule itself in liver disease serum. Methods: Sera or their immunoprecipitate with anti-APN monoclonal antibody were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) or two-dimensional electrophoresis and subsequent Western blotting with rabbit anti-APN serum. Results: In all the patient sera examined, the 140-kDa APN isoform was predominant. In all the sera from 10 patients with cholestatic diseases (8 with extra-hepatic cholestasis and 2 with primary biliary cirrhosis), we observed the 260-kDa isoform that was immunoprecipitated with monoclonal APN antibodies and had a similar isoelectric point to the 140-kDa isoform. However, the 260-kDa isoform was observed faintly in 2 out of 12 patients with other liver diseases, including chronic hepatitis and cirrhosis. Conclusions: We found a novel high-molecular-mass APN isoform (260-kDa) in serum, which is highly likely to be a homodimer of APNs bound covalently and a promising marker of cholestasis. This suggests increased crosslinking reaction between two APN molecules in cholestatic patients. D 2003 Elsevier Science B.V. All rights reserved. Keywords: Aminopeptidase N; Human serum; Isoform; SDS-PAGE; Homodimer; Cholestasis
1. Introduction Aminopeptidase N (APN) (EC 3.4.11.2), of which the cDNA has already been cloned [1,2], is an ectoAbbreviations: APN, aminopeptidase N; LAP, leucine aminopeptidase; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; ALP, alkaline phosphatase; GGT, gamma-glutamyl transferase; PVDF, polyvinylidene difluoride; TBS, Tris-buffered saline; HOS, hydroxyl radical. * Corresponding author. Tel.: +81-3-5803-5375; fax: +81-35803-0161. E-mail address: [email protected] (M. Kawai).
enzyme anchoring its N-terminus to the cell membrane and facing its catalytic domain outside the cell [3,4]. In laboratory medicine, APN is a marker of liver dysfunction, and is also called alanine aminopeptidase, leucine aminopeptidase (LAP), or arylamidase. High serum APN activity, like alkaline phosphatase (ALP) or gamma-glutamyl transferase (GGT), indicates cholestasis. However, serum APN activity is not as sensitive as ALP or GGT activity in detection of cholestasis [5]. On the other hand, isoforms of APN have been investigated. It is well known that, in cholestasis,
0009-8981/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0009-8981(03)00002-0
142
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
liver plasma membrane fragments containing ALP, GGT, and APN are shed into serum as high molecular weight multi-enzyme complexes that are detectable by gel chromatography in a void volume fraction. The gel chromatography of liver plasma membrane fragments, although useful to detect obstructive cholestasis or hepatocellular carcinoma sensitively and specifically, is time-consuming and inconvenient to run on many samples as a laboratory test [6 – 8]. Electrophoresis can analyze many samples at once, thus being more practical as a laboratory test. The electrophoretic techniques used in previous reports to analyze serum APN isoforms, however, were nondenaturing procedures, in which the noncovalent bond between protein molecules remained [9 –11]. Accordingly, two APN molecules possibly continue to be associated noncovalently with each other to form a homodimer [3,4,12] as produced in hepatocytes. Additionally, liver plasma membrane fragments containing APN would remain to some extent. Therefore, in these non-denaturing electrophoretic methods, too many bands attributable to various types of noncovalent, multi-molecular complexes of APN are observed [9]. In addition, the patterns of APN bands observed in respective patients with an identical cholestatic disease are often different from one another [9]. The presence of extra many bands and different patterns observed in the same disease make it difficult to analyze APN isoforms and to diagnose the diseases [9]. Moreover, non-denaturing polyacrylamide gel electrophoresis does not show molecular mass accurately, and enzyme activity staining also detects nonAPN enzymes, including cysteine aminopeptidase [10,11]. In the present study, we analyzed serum APN by SDS-PAGE or two-dimensional electrophoresis with Western blotting in which APN is completely denatured and specifically detected. In this condition, the protein molecule complex formed by a noncovalent bond dissociates completely, and the observed bands should show their molecular mass accurately. We observed an identical pattern of serum APN isoforms among cholestatic diseases and found a novel highmolecular-mass isoform that is likely to be a covalently formed homodimer and a promising marker of cholestasis (extra-hepatic cholestasis and primary biliary cirrhosis).
2. Materials and methods 2.1. Serum samples Serum samples were collected from 11 healthy subjects and 23 patients with liver diseases who came to our university hospital. The samples were frozen at 70 jC until analysis. The diseases were as follows: 5 patients with carcinoma of the pancreas head, 2 with carcinoma of the bile duct (choledochus), 1 with stones of the bile duct (choledochus), (in total, 8 patients with extra-hepatic bile duct obstruction), 1 with stones of the gall bladder, 2 with primary biliary cirrhosis, 4 with chronic hepatitis (C-type), 5 with liver cirrhosis (C-type), 2 with drug-induced liver dysfunction, and 1 with autoimmune hepatitis. Extra-hepatic cholestasis, carcinoma of the pancreas, carcinoma of the bile duct, stones of the bile duct, and stones of the gall bladder were diagnosed by ultrasonography, computed tomography, endoscopic retrograde cholangiopancreatography, and/or percutaneous transhepatic cholangiography. Chronic hepatitis, liver cirrhosis, drug-induced liver dysfunction, primary biliary cirrhosis, and autoimmune hepatitis were diagnosed with serum aminotransferases, bilirubin, ALP, GGT, albumin, gamma-globulin, platelet counts, hepatitis virus tests, autoantibodies (antinuclear antibody, antimitochondrial antibody), and ultrasonography. To diagnose primary biliary cirrhosis, liver biopsy was also performed. 2.2. Immunoprecipitation The immunoprecipitation procedure was described elsewhere [13]. For one-dimensional SDS-PAGE, 30 Al of serum from a cholestatic disease patient and 20 Al of 20 mmol/l sodium phosphate buffer, pH 7.4, 150 mmol/l NaCl, and 50 ml/l Triton X-100 were mixed and then incubated with 7.5 Al of undiluted monoclonal APN antibody or 0.1 g/l normal mouse IgG as a mock (control) antibody for 2 h at 4 jC. As a monoclonal APN antibody, MCS-2 (0.1 g/l) (Nichirei, Tokyo, Japan) or WM15 (Code: CMO307, Ylem, Rome, Italy; diluted 1:20 – 30 to stain frozen tissue according to Ylem) was added. The mixture of sample and antibody was then incubated with 0.5 Al of 28.7 g/l rabbit anti-mouse IgG g-globulin overnight at 4 jC. After centrifugation at 7000 g
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
for 5 min, the precipitate was washed twice with Trisbuffered saline (TBS) (25 mmol/l Tris, pH 7.4, 150 mmol/l NaCl) containing 1 ml/l Triton X-100 and once with TBS. For two-dimensional electrophoresis, 300 Al of serum was mixed with 200 Al of 20 mmol/l sodium phosphate buffer, pH 7.4, 150 mmol/l NaCl, and 50 ml/l Triton X-100 and immunoprecipitated with 75 Al of 0.1 g/l monoclonal APN antibody, MCS-2 and 5 Al of 28.7 g/l rabbit anti-mouse IgG g-globulin and washed as described above. 2.3. Electrophoresis For one-dimensional SDS-PAGE, 1 or 2 Al of serum or liver membrane suspension was dissolved in 10 Al of sample buffer containing 58 mmol/l Tris, pH 6.8, 60 g/l SDS, 50 g/l glycerol, and 0.1 mol/l dithiothreitol and heated at 56 jC for 5 min unless described otherwise (some samples were also heated at 100 jC for 5 min and electrophoresed). The samples were then subjected to SDS-PAGE performed on a 30-g/l polyacrylamide stacking gel (1cm length, 1-mm thickness) and a 50-g/l separating gel (6-cm length) according to Laemmli’s procedure [14]. An immunoprecipitated sample from 30 Al of serum was dissolved in 20 Al of the same sample buffer and heated at 56 jC for 5 min unless described otherwise. Four microliters of this sample was applied to one-dimensional SDS-PAGE. For combined isoelectric focusing and SDSPAGE (two-dimensional electrophoresis), an immunoprecipitated sample from 300 Al of serum was dissolved in 20 Al of 8 mol/l urea, 1.8 g/l Pharmalyte (preblended, pH 3.5– 9.5) (Amersham Pharmacia Biotech, Buckinghamshire, UK), and 20 ml/l Triton X-100. With a CoolPhoreStar IPG-IEF type P apparatus (Anatech, Tokyo, Japan), first-dimensional isoelectric focusing was performed at 20 jC on an Immobiline Strip (pH 3 –10, 7 cm, Amersham Pharmacia Biotech) swollen with the Pharmalyte solution containing the above components. The sample absorbed in a piece of small filter paper (Anatech) was applied on the swollen gel strip near its cathode side. The gel strip was immersed in silicon oil and subjected to the voltage programmed as follows: 200 V for 20 min, 500 V for 1 h 30 min,
143
700 V for 15 min, 1000 V for 15 min, 1500 V for 15 min, 2000 V for 15 min, 2500 V for 15 min, and 3000 V for 3 h. The electrophoresed immobiline gel strip was equilibrated with the buffer containing 6 mol/l urea, 50 mmol/l Tris, pH 6.8, 10 g/l SDS, 300 g/l glycerol, 16 mmol/l dithiothreitol at room temperature for 10 min and then subjected to seconddimensional SDS-PAGE. 2.4. Western blotting The electrophoresed proteins in a polyacrylamide gel were transferred to an Immobilon polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA) with 25 mmol/l Tris, 192 mmol/l glycine at 2 mA/cm2 for 30 min. The PVDF membrane was blocked with 50 g/l skim milk, TBS at 4 jC overnight and then washed with TBS for 1 min. The blocked PVDF membrane was incubated with rabbit anti-APN serum (donated by Prof. Yasuhiro Watanabe of Hokkaido College of Pharmacy [15]) diluted 1:500 with TBS at room temperature for 40 min and then incubated with goat anti-rabbit IgGperoxidase conjugate (Code: A4914, Sigma, St. Louis, MO) diluted 1:5000 with TBS at room temperature for 30 min. After each incubation with antibody, the membrane was washed with 1 g/l Tween-20, TBS for 3 min three times and rinsed with TBS for 1 min. Bound antibody was detected with ECL reagents (Amersham Pharmacia Biotech) and an X-ray film according to the manufacturer’s procedure. 2.5. Human liver membrane Human liver tissue was obtained at autopsy from a patient who died of ureteral carcinoma but had no particular disease in the liver. The autopsy was performed after 2.5 h after the death of the patient. 0.83 g of the liver tissue was homogenized in 10 ml of solution containing 0.25 mol/l sucrose, 10 mmol/l sodium phosphate buffer, pH 7.4, and then centrifuged at 1000 g for 10 min. The supernatant was centrifuged at 350,000 g for 20 min and the precipitate obtained was suspended in 500 Al of 10 mmol/l sodium phosphate buffer, pH 7.4 as a sample for electrophoresis. All these procedures were performed at 4 jC.
144
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
3. Results 3.1. Serum APN isoforms from patients with extrahepatic cholestasis and primary biliary cirrhosis As shown in Fig. 1A, SDS-PAGE followed by Western blotting with the APN antiserum detected four bands in each serum sample from cholestatic patients. The molecular masses of these bands were 130-, 140-, 165-, and 260-kDa. The 140- and 260-kDa bands were predominant. All 10 sera from cholestatic patients showed the same pattern of bands. These cholestatic patients include 5 patients with carcinoma of the pancreas head, 2 patients with carcinoma of the bile duct (choledochus), 1 patient with stones of the bile duct (choledochus), (in total, 8 patients of obvious extra-hepatic cholestasis demonstrated by ultrasonography), and 2 patients of primary biliary cirrhosis (intra-hepatic cholestasis). To avoid destruction of sample proteins, we heated serum with sample buffer at 56 jC for 5 min before SDS-PAGE. We also performed the same experiment as that in Fig. 1A except that the five samples were heated with sample buffer at 100 jC for 5 min to ensure dissociation of APN molecules. In this experiment, we obtained the same results as shown in Fig. 1A (data not shown). Besides the 10 patients mentioned above, 1 patient with stones of the gall bladder was considered to have latent extra-hepatic cholestasis because she had transiently shown bile duct dilatation and a high serum bilirubin concentration 2 years before the blood sampling for this study. Fig. 1B shows the analysis of this patient serum by SDS-PAGE and Western blotting. The 260-kDa band was clearly demonstrated. This result suggests that 260-kDa band is sensitive enough to detect extra-hepatic cholestasis even before its mechanical obstruction of the bile duct becomes morphologically clear. 3.2. Serum APN isoform from the patients with the diseases other than extra-hepatic cholestasis or primary biliary cirrhosis Besides extra-hepatic cholestasis and primary biliary cirrhosis, we examined 12 patients, including 4 patients with chronic hepatitis (C-type), 5 patients of cirrhosis (C-type), 2 patients with drug-induced liver
Fig. 1. APN isoforms in sera of patients with extra-hepatic cholestasis. (A) Sera taken from extra-hepatic cholestasis patients (1 Al each) were analyzed by SDS-PAGE followed by Western blotting with anti-APN serum. The diseases were: carcinoma of the bile duct in lane 1; carcinoma of the pancreas in lanes 2 – 4; stones of the bile duct in lane 5. This experiment showed 130-, 140-, 165-, and 260-kDa bands in each serum that are considered to be APN isoforms due to the results shown in Figs. 3 and 4. The 140- and 260-kDa isoforms were predominant. (B) Serum (2 Al) from a patient with gall bladder stones was analyzed. This patient was considered to have latent obstructive cholestasis because she had transiently shown bile duct dilatation and a high serum bilirubin 2 years before the blood sampling for this study. The 260- and the 140-kDa APN isoform were recognized.
dysfunction, and 1 patient with autoimmune hepatitis (taking prednisolone 12.5 mg per day). We clearly recognized the 140-kDa isoform in all these cases although its intensity was considerably lower than in cholestasis cases (extra-hepatic cholestasis and pri-
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
Fig. 2. APN isoform in sera from patients with chronic hepatitis or liver cirrhosis. Serum samples (2 Al each) taken from chronic hepatitis or liver cirrhosis patients were analyzed by SDS-PAGE, followed by Western blotting with anti-APN serum. The diseases were: C-type liver cirrhosis in lanes 1 – 3; C-type chronic hepatitis in lanes 4 – 6. The 140-kDa band is identified.
145
monoclonal APN antibody was compared with original serum samples in simultaneous electrophoresis (SDS-PAGE). Fig. 3 shows that 130-, 140-, 165-, and 260-kDa bands found in serum of an extra-hepatic cholestasis patient (number 1 patient in Fig. 1A) were definitely immunoprecipitated with two different monoclonal APN antibodies: MCS-2 and WM15. We also performed the same experiment as that in Fig. 3 except that the samples were heated with sample buffer at 100 jC for 5 min to ensure dissociation of APN molecules. In this experiment, we obtained the same results as shown in Fig. 3 (data not shown). Fig. 4 shows the two-dimensional electrophoresis of the same patient serum where 130-, 140-, 165-, and 260-kDa bands were immunoprecipitated with MCS-
mary biliary cirrhosis). Fig. 2 shows the results of 5 patients as the examples. The 260-kDa band was faintly visible in only 2 patients: 1 with cirrhosis (Ctype) and 1 with drug-induced liver dysfunction (data not shown). The results of patients with cholestasis and other diseases suggest that the 260-kDa band is a promising marker of cholestasis caused by extra-hepatic obstruction or primary biliary cirrhosis. 3.3. Serum APN isoform from healthy subjects In serum from 11 healthy subjects, the predominantly recognized band was a 140-kDa APN isoform whose electrophoretic mobility (molecular mass) was identical with that in chronic hepatitis or liver cirrhosis (C-type) patient serum. The intensity of the band of the healthy subjects was also similar to that of chronic hepatitis or cirrhosis patients (data not shown). 3.4. Immunoprecipitation of serum APN If the specificity of the APN antiserum is insufficient, there is the possibility that, besides APN bands, non-APN bands may appear in the Western blot. To exclude this possibility and confirm the bands to be true APN isoforms, immunoprecipitated APN with
Fig. 3. SDS-PAGE of immunoprecipitated APN. Serum of one patient with extra-hepatic cholestasis (carcinoma of the bile duct, the serum in lane 1 of Fig. 1A) was used in this SDS-PAGE and Western blotting. The samples electrophoresed were: lane 1, the original serum (1 Al); lane 2, a sample (4 Al) of the same serum immunoprecipitated with MCS-2 monoclonal APN antibody; lane 3, a sample (4 Al) of the same serum immunoprecipitated with WM15 monoclonal APN antibody; lane 4, a sample (4 Al) of the same serum immunoprecipitated with mock antibody (normal mouse IgG). The 130-, 140-, 165-, and 260-kDa bands were immunoprecipitated with monoclonal APN antibodies but not with mock antibody.
146
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
Fig. 4. Two-dimensional electrophoresis of immunoprecipitated APN. An immunoprecipitate from 300 Al of cholestatic patient serum (Fig. 3) with MCS-2 monoclonal APN antibody was subjected to isoelectric focusing (first-dimensional electrophoresis) on an Immobile gel strip. The electrophoresed gel strip and the original serum, indicated as ‘IEF’ and ‘serum’, respectively, were simultaneously subjected to SDS-PAGE (second-dimensional electrophoresis). The 130-, 140-, 165-, and 260-kDa bands (arrowheads) were immunoprecipitated with MCS-2. Additionally, their isoelectric points were similar. In particular, the predominant 140- and 260-kDa bands showed almost identical isoelectric points.
2. Additionally, the isoelectric points of these four bands in Fig. 4 were similar to one another and consistent with those of APN obtained from normal subjects that we previously reported [16]. In particular, the predominant 140- and 260-kDa bands showed almost identical isoelectric points. We also confirmed that the serum of another extrahepatic cholestasis patient (number 2 in Fig. 1A) gave the same results (data not shown) as shown in Figs. 3 and 4. The molecular mass of 260-kDa is substantially higher than the others. We termed this 260-kDa isoform ‘high-molecular-mass APN isoform’. This isoform is resistant to denaturation by SDS and reduction by dithiothreitol. 3.5. APN isoforms of human liver membrane Fig. 5 shows that the 140- and 260-kDa isoforms are also contained in the membrane fragments of human liver tissue, although the proportion of 140to 260-kDa band intensity is different from that of serum.
Fig. 5. APN isoforms in liver membrane. The same serum as in Fig. 3 (2 Al) was electrophoresed in lane 1 and liver membrane suspension (2 Al) obtained as described in Section 2.5 was in lane 2. The 140- and 260-kDa bands were also recognized in human liver membrane.
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
4. Discussion Although it has been already reported that serum leucine aminopeptidase and alanine aminopeptidase activity are mainly derived from APN/CD13 [13, 17,18], electrophoretic analysis of serum APN isoforms with anti-APN antibody has not been previously reported. In the past report of non-denaturing electrophoresis, serum APN isoform patterns observed in the same disease were often different from one another [9]. By one-dimensional SDS-PAGE with anti-APN antiserum, we recognized the same pattern of APN bands in cholestatic patients, consisting of 130-, 140-, 165-, and 260-kDa bands. All four bands were immunoprecipitated with two monoclonal APN antibodies, thus indicating that each of them has two epitopes of APN; they were authentic APN isoforms. Two-dimensional electrophoresis also confirmed the immunoprecipitation of the four bands with a monoclonal APN antibody. The high-molecular-mass APN isoform (260-kDa) was observed in all the sera of extra-hepatic cholestasis and primary biliary cirrhosis. In the cases of liver dysfunction of other causes, however, no or very faint bands of high-molecular-mass APN were recognized. Additionally, we observed high-molecular-mass APN in a case of gall stones where no dilatation of bile ducts was recognized by ultrasonographic examination at the time of blood sampling. These results suggest that the high-molecular-mass APN isoform is closely related to extra-hepatic cholestasis and primary biliary cirrhosis and is a promising marker of cholestasis of these diseases. Considering the molecular weight of human and porcine APNs previously reported, the bands of 130-, 140-, and 165-kDa must be monomers of APN [3,19] with variations of their N-glycosylation [3] and the 260-kDa band, the high-molecular-mass APN isoform, is highly likely to be an APN homodimer [19]. Human APN molecules are known to be produced in cells as a homodimer whose subunits (APNs) are bound to each other with a noncovalent bond [3]. The noncovalently formed homodimer should dissociate to be presented as monomers under reducing and denaturing SDS-PAGE conditions. Our results suggest that the subunits (APN molecules) of the 260-kDa isoform (high-molecular-mass isoform) are bound covalently by a non-disulfide bond. The
147
results of the 260-kDa isoform immunoprecipitated with two monoclonal APN antibodies show that this isoform contains two APN epitopes. Thus, the alternative hypothesis that should be considered is that the 260-kDa isoform is a complex consisting of an APN and another different protein molecule. The result that the 260-kDa isoform has an almost identical isoelectric point to the 140-kDa APN isoform supports the hypothesis of the APN homodimer rather than that of APN and another protein complex. Additionally, 260-kDa is approximately two times 140-kDa. Therefore, we consider that the high-molecular-mass APN isoform of 260-kDa in cholestatic serum is an APN homodimer whose subunits are covalently bound to each other. Being compatible with the assumption that the 260-kDa APN isoform is released from liver, Fig. 5 shows that human liver definitely contains the highmolecular-mass (260-kDa) APN, although the intensity of 260-kDa band was much less than 140-kDa band as compared with serum. The liver tissue sample used was obtained from a patient without cholestasis. To further support the above assumption, liver tissue from cholestatic patients should also be examined because 260-kDa isoform is supposed to increase in cholestasis from our results. It was previously observed that some cross-linking agents, e.g., dimethyl adipimidate, bind covalently the two APN molecules associated by noncovalent bond [19,20]. However, it has not been reported that human serum samples that were not exposed to any artificial or experimental cross-linking treatment contain a covalently formed homodimer of APN. Our results indicate that the covalently formed APN homodimer increases in cholestatic serum. Therefore, the occurrence of this APN dimer in serum and the mechanism of its production in cholestatic cases are of interest. The mechanism should include increased cross-linking reaction between two APN molecules in cholestatic condition. It was previously reported that, in extra-hepatic cholestasis, oxide free radicals in the liver, including hydroxyl radical (HOS), increase due to hydrophobic bile acid accumulation in the liver and act as a major pathogenic factor in cholestatic liver injury [21 –23]. Also, in primary biliary cirrhosis, lipid peroxidation was reported to increase in the liver, suggesting
148
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
increased production of free radicals [24]. Hydroxyl free radical (HOS) is known to dimerize proteins; the proteins exposed to hydroxyl radical undergo progressive cross-linking to form dimers because the tyrosine residues of the proteins are oxidized by hydroxyl radical and intermolecular bityrosine are formed [25]. During the physiological synthesis in endoplasmic reticulum, two APN molecules are associated noncovalently. In cholestatic cases, increased free radicals should bind covalently the two APN molecules associated in the synthetic process of APN. Therefore, the high-molecular-mass APN isoform possibly reflects the increased production of free radicals in cholestatic cases that plays a pivotal role in the pathogenesis of cholestatic liver injury. Further study is necessary to clarify the sensitivity and specificity of high-molecular-mass APN isoform for diagnosing cholestasis and to reveal the mechanism of covalent APN homodimer formation.
Acknowledgements We thank Prof. Yasuhiro Watanabe (Department of Pathological Biochemistry, Hokkaido College of Pharmacy) for his donation of rabbit anti-aminopeptidase N serum.
References [1] Look AT, Ashmun RA, Shapiro LH, Peiper SC. Human myeloid plasma membrane glycoprotein CD13 (gp150) is identical to aminopeptidase N. J Clin Invest 1989;83:1299 – 307. [2] Olsen J, Cowell GM, Konigshofer E, Danielsen EM, Moller J, Laustsen L, et al. Complete amino acid sequence of human intestinal aminopeptidase N as deduced from cloned cDNA. FEBS Lett 1988;238:307 – 14. [3] Shipp MA, Look AT. Hematopoietic differentiation antigens that are membrane-associated enzymes: cutting is the key! Blood 1993;82:1052 – 70. [4] Riemann D, Kehlen A, Langner J. CD13-not just a marker in leukemia typing. Immunol Today 1999;20:83 – 8. [5] Reichling JJ, Kaplan MM. Clinical use of serum enzymes in liver disease. Dig Dis Sci 1988;33:1601 – 14. [6] Shinkai K, Akedo H. A multienzyme complex in serum of hepatic cancer. Cancer Res 1972;32:2307 – 13. [7] Ideo G, Ronchi G. Sephadex gel filtration of g-glutamyl-transpeptidase, alkaline phosphatase and leucine aminopeptidase in the serum of patients affected by various liver diseases. Z Klin Chem Klin Biochem 1972;10:S211 – 4.
[8] Deng JT, Hoylaerts MF, Nouwen EJ, De Broe ME, Van Hoof VO. Purification of circulating liver plasma membrane fragments using a monoclonal antileucine aminopeptidase antibody. Hepatology 1996;23:445 – 54. [9] Dingjan PG, Jongeneel J, Postma T, Stroes JAP. Evaluation of the determination of alanine naphthyl-amidase isoenzymes by agar and polyacrylamide disc electrophoresis for the diagnosis of pancreatic diseases. Clin Chim Acta 1973;49: 215 – 24. [10] Sanderink GJCM, Artur Y, Galteau MM, Wellman-Bednawska M, Siest G. Multiple forms of serum aminopeptidases separated by micro two-dimensional electrophoresis under non-denaturing conditions. Electrophoresis 1986;7:471 – 5. [11] Sanderink GJ, Artur Y, Paille F, Siest G. Clinical significance of a new isoform of serum alanine aminopeptidase; relationship of liver disease and alcohol consumption. Clin Chim Acta 1989;179:23 – 32. [12] Hussain MM, Tranum-Jensen J, Noren O, Sjostrom H, Christiansen K. Reconstitution of purified amphiphilic pig intestinal microvillus aminopeptidase. Mode of membrane insertion and morphology. Biochem J 1981;199:179 – 86. [13] Kawai M, Hara Y, Kubota T, Shiba K, Hosaki S. A family with high serum leucine aminopeptidase activity derived from a novel variant CD13. Clin Chem 1998;44:215 – 20. [14] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227: 680 – 5. [15] Watanabe Y, Iwaki-Egawa S, Mizukoshi H, Fujimoto Y. Identification of an alanine aminopeptidase in human maternal serum as a membrane-bound aminopeptidase N. Biol Chem Hoppe-Seyler 1995;376:397 – 400. [16] Kawai M, Hara Y, Miyazato I, Hosaki S. Novel, aberrantly truncated isoform of serum CD13 in a family with high serum aminopeptidase N (CD13) activity. Clin Chem 2001;47: 223 – 30. [17] Favaloro EJ, Browning T, Nandurkar H. The hepatobiliary disease marker serum alanine aminopeptidase predominantly comprises an isoform of the haematological myeloid differentiation antigen and leukaemia marker CD-13/gp150. Clin Chim Acta 1993;220:81 – 90. [18] Watanabe Y, Ito K, Iwaki-Egawa S, Yamaguchi R, Fujimoto Y. Aminopeptidase N in sera of healthy subjects is a different N-terminal processed derivative from the one obtained from maternal serum. Mol Genet Metab 1998;63:289 – 94. [19] Danielsen EM. Biosynthesis of intestinal microvillar proteins. Dimerization of aminopeptidase N and lactase-phlorizin hydrolase. Biochemistry 1990;29:305 – 8. [20] Danielsen EM. Perturbation of intestinal microvillar enzyme biosynthesis by amino acid analogs. J Biol Chem 1990;265: 14566 – 71. [21] Sokol RJ, Devereaux M, Khandwala R, O’Brien K. Evidence for involvement of oxygen free radicals in bile acid toxicity to isolated rat hepatocytes. Hepatology 1993;17:869 – 81. [22] Sokol RJ, Winklhofer-Roob BM, Devereaux MW, McKim Jr JM. Generation of hydroperoxides in isolated rat hepatocytes and hepatic mitochondria exposed to hydrophobic bile acids. Gastroenterology 1995;109:1249 – 56.
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149 [23] Tsai LY, Lee KT, Liu TZ. Evidence for accelerated generation of hydroxyl radicals in experimental obstructive jaundice of rats. Free Radic Biol Med 1998;24:732 – 7. [24] Kawamura K, Kobayashi Y, Kageyama F, Kawasaki T, Nagasawa M, Toyokuni S, et al. Enhanced hepatic lipid peroxida-
149
tion in patients with primary biliary cirrhosis. Am J Gastroenterol 2000;95:3596 – 601. [25] Davies KJA, Delsignore ME. Protein damage and degradation by oxygen radicals. III. Modification of secondary and tertiary structure. J Biol Chem 1987;262:9908 – 13.
High-molecular-mass isoform of aminopeptidase N/CD13 in serum from cholestatic patients Makoto Kawai *, Yuya Otake, Yukichi Hara Department of Biochemistry and Biophysics, Graduate School of Allied Health Sciences, Tokyo Medical and Dental University, 5-45, Yushima 1-chome, Bunkyo-ku, Tokyo 113-8519, Japan Received 15 October 2002; received in revised form 19 December 2002; accepted 20 December 2002
Abstract Background: Because non-denaturing electrophoresis and aminopeptidase activity staining often detect noncovalent multienzyme complexes, we adopted procedures to specifically detect the aminopeptidase N (APN) molecule itself in liver disease serum. Methods: Sera or their immunoprecipitate with anti-APN monoclonal antibody were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) or two-dimensional electrophoresis and subsequent Western blotting with rabbit anti-APN serum. Results: In all the patient sera examined, the 140-kDa APN isoform was predominant. In all the sera from 10 patients with cholestatic diseases (8 with extra-hepatic cholestasis and 2 with primary biliary cirrhosis), we observed the 260-kDa isoform that was immunoprecipitated with monoclonal APN antibodies and had a similar isoelectric point to the 140-kDa isoform. However, the 260-kDa isoform was observed faintly in 2 out of 12 patients with other liver diseases, including chronic hepatitis and cirrhosis. Conclusions: We found a novel high-molecular-mass APN isoform (260-kDa) in serum, which is highly likely to be a homodimer of APNs bound covalently and a promising marker of cholestasis. This suggests increased crosslinking reaction between two APN molecules in cholestatic patients. D 2003 Elsevier Science B.V. All rights reserved. Keywords: Aminopeptidase N; Human serum; Isoform; SDS-PAGE; Homodimer; Cholestasis
1. Introduction Aminopeptidase N (APN) (EC 3.4.11.2), of which the cDNA has already been cloned [1,2], is an ectoAbbreviations: APN, aminopeptidase N; LAP, leucine aminopeptidase; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; ALP, alkaline phosphatase; GGT, gamma-glutamyl transferase; PVDF, polyvinylidene difluoride; TBS, Tris-buffered saline; HOS, hydroxyl radical. * Corresponding author. Tel.: +81-3-5803-5375; fax: +81-35803-0161. E-mail address: [email protected] (M. Kawai).
enzyme anchoring its N-terminus to the cell membrane and facing its catalytic domain outside the cell [3,4]. In laboratory medicine, APN is a marker of liver dysfunction, and is also called alanine aminopeptidase, leucine aminopeptidase (LAP), or arylamidase. High serum APN activity, like alkaline phosphatase (ALP) or gamma-glutamyl transferase (GGT), indicates cholestasis. However, serum APN activity is not as sensitive as ALP or GGT activity in detection of cholestasis [5]. On the other hand, isoforms of APN have been investigated. It is well known that, in cholestasis,
0009-8981/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0009-8981(03)00002-0
142
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
liver plasma membrane fragments containing ALP, GGT, and APN are shed into serum as high molecular weight multi-enzyme complexes that are detectable by gel chromatography in a void volume fraction. The gel chromatography of liver plasma membrane fragments, although useful to detect obstructive cholestasis or hepatocellular carcinoma sensitively and specifically, is time-consuming and inconvenient to run on many samples as a laboratory test [6 – 8]. Electrophoresis can analyze many samples at once, thus being more practical as a laboratory test. The electrophoretic techniques used in previous reports to analyze serum APN isoforms, however, were nondenaturing procedures, in which the noncovalent bond between protein molecules remained [9 –11]. Accordingly, two APN molecules possibly continue to be associated noncovalently with each other to form a homodimer [3,4,12] as produced in hepatocytes. Additionally, liver plasma membrane fragments containing APN would remain to some extent. Therefore, in these non-denaturing electrophoretic methods, too many bands attributable to various types of noncovalent, multi-molecular complexes of APN are observed [9]. In addition, the patterns of APN bands observed in respective patients with an identical cholestatic disease are often different from one another [9]. The presence of extra many bands and different patterns observed in the same disease make it difficult to analyze APN isoforms and to diagnose the diseases [9]. Moreover, non-denaturing polyacrylamide gel electrophoresis does not show molecular mass accurately, and enzyme activity staining also detects nonAPN enzymes, including cysteine aminopeptidase [10,11]. In the present study, we analyzed serum APN by SDS-PAGE or two-dimensional electrophoresis with Western blotting in which APN is completely denatured and specifically detected. In this condition, the protein molecule complex formed by a noncovalent bond dissociates completely, and the observed bands should show their molecular mass accurately. We observed an identical pattern of serum APN isoforms among cholestatic diseases and found a novel highmolecular-mass isoform that is likely to be a covalently formed homodimer and a promising marker of cholestasis (extra-hepatic cholestasis and primary biliary cirrhosis).
2. Materials and methods 2.1. Serum samples Serum samples were collected from 11 healthy subjects and 23 patients with liver diseases who came to our university hospital. The samples were frozen at 70 jC until analysis. The diseases were as follows: 5 patients with carcinoma of the pancreas head, 2 with carcinoma of the bile duct (choledochus), 1 with stones of the bile duct (choledochus), (in total, 8 patients with extra-hepatic bile duct obstruction), 1 with stones of the gall bladder, 2 with primary biliary cirrhosis, 4 with chronic hepatitis (C-type), 5 with liver cirrhosis (C-type), 2 with drug-induced liver dysfunction, and 1 with autoimmune hepatitis. Extra-hepatic cholestasis, carcinoma of the pancreas, carcinoma of the bile duct, stones of the bile duct, and stones of the gall bladder were diagnosed by ultrasonography, computed tomography, endoscopic retrograde cholangiopancreatography, and/or percutaneous transhepatic cholangiography. Chronic hepatitis, liver cirrhosis, drug-induced liver dysfunction, primary biliary cirrhosis, and autoimmune hepatitis were diagnosed with serum aminotransferases, bilirubin, ALP, GGT, albumin, gamma-globulin, platelet counts, hepatitis virus tests, autoantibodies (antinuclear antibody, antimitochondrial antibody), and ultrasonography. To diagnose primary biliary cirrhosis, liver biopsy was also performed. 2.2. Immunoprecipitation The immunoprecipitation procedure was described elsewhere [13]. For one-dimensional SDS-PAGE, 30 Al of serum from a cholestatic disease patient and 20 Al of 20 mmol/l sodium phosphate buffer, pH 7.4, 150 mmol/l NaCl, and 50 ml/l Triton X-100 were mixed and then incubated with 7.5 Al of undiluted monoclonal APN antibody or 0.1 g/l normal mouse IgG as a mock (control) antibody for 2 h at 4 jC. As a monoclonal APN antibody, MCS-2 (0.1 g/l) (Nichirei, Tokyo, Japan) or WM15 (Code: CMO307, Ylem, Rome, Italy; diluted 1:20 – 30 to stain frozen tissue according to Ylem) was added. The mixture of sample and antibody was then incubated with 0.5 Al of 28.7 g/l rabbit anti-mouse IgG g-globulin overnight at 4 jC. After centrifugation at 7000 g
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
for 5 min, the precipitate was washed twice with Trisbuffered saline (TBS) (25 mmol/l Tris, pH 7.4, 150 mmol/l NaCl) containing 1 ml/l Triton X-100 and once with TBS. For two-dimensional electrophoresis, 300 Al of serum was mixed with 200 Al of 20 mmol/l sodium phosphate buffer, pH 7.4, 150 mmol/l NaCl, and 50 ml/l Triton X-100 and immunoprecipitated with 75 Al of 0.1 g/l monoclonal APN antibody, MCS-2 and 5 Al of 28.7 g/l rabbit anti-mouse IgG g-globulin and washed as described above. 2.3. Electrophoresis For one-dimensional SDS-PAGE, 1 or 2 Al of serum or liver membrane suspension was dissolved in 10 Al of sample buffer containing 58 mmol/l Tris, pH 6.8, 60 g/l SDS, 50 g/l glycerol, and 0.1 mol/l dithiothreitol and heated at 56 jC for 5 min unless described otherwise (some samples were also heated at 100 jC for 5 min and electrophoresed). The samples were then subjected to SDS-PAGE performed on a 30-g/l polyacrylamide stacking gel (1cm length, 1-mm thickness) and a 50-g/l separating gel (6-cm length) according to Laemmli’s procedure [14]. An immunoprecipitated sample from 30 Al of serum was dissolved in 20 Al of the same sample buffer and heated at 56 jC for 5 min unless described otherwise. Four microliters of this sample was applied to one-dimensional SDS-PAGE. For combined isoelectric focusing and SDSPAGE (two-dimensional electrophoresis), an immunoprecipitated sample from 300 Al of serum was dissolved in 20 Al of 8 mol/l urea, 1.8 g/l Pharmalyte (preblended, pH 3.5– 9.5) (Amersham Pharmacia Biotech, Buckinghamshire, UK), and 20 ml/l Triton X-100. With a CoolPhoreStar IPG-IEF type P apparatus (Anatech, Tokyo, Japan), first-dimensional isoelectric focusing was performed at 20 jC on an Immobiline Strip (pH 3 –10, 7 cm, Amersham Pharmacia Biotech) swollen with the Pharmalyte solution containing the above components. The sample absorbed in a piece of small filter paper (Anatech) was applied on the swollen gel strip near its cathode side. The gel strip was immersed in silicon oil and subjected to the voltage programmed as follows: 200 V for 20 min, 500 V for 1 h 30 min,
143
700 V for 15 min, 1000 V for 15 min, 1500 V for 15 min, 2000 V for 15 min, 2500 V for 15 min, and 3000 V for 3 h. The electrophoresed immobiline gel strip was equilibrated with the buffer containing 6 mol/l urea, 50 mmol/l Tris, pH 6.8, 10 g/l SDS, 300 g/l glycerol, 16 mmol/l dithiothreitol at room temperature for 10 min and then subjected to seconddimensional SDS-PAGE. 2.4. Western blotting The electrophoresed proteins in a polyacrylamide gel were transferred to an Immobilon polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA) with 25 mmol/l Tris, 192 mmol/l glycine at 2 mA/cm2 for 30 min. The PVDF membrane was blocked with 50 g/l skim milk, TBS at 4 jC overnight and then washed with TBS for 1 min. The blocked PVDF membrane was incubated with rabbit anti-APN serum (donated by Prof. Yasuhiro Watanabe of Hokkaido College of Pharmacy [15]) diluted 1:500 with TBS at room temperature for 40 min and then incubated with goat anti-rabbit IgGperoxidase conjugate (Code: A4914, Sigma, St. Louis, MO) diluted 1:5000 with TBS at room temperature for 30 min. After each incubation with antibody, the membrane was washed with 1 g/l Tween-20, TBS for 3 min three times and rinsed with TBS for 1 min. Bound antibody was detected with ECL reagents (Amersham Pharmacia Biotech) and an X-ray film according to the manufacturer’s procedure. 2.5. Human liver membrane Human liver tissue was obtained at autopsy from a patient who died of ureteral carcinoma but had no particular disease in the liver. The autopsy was performed after 2.5 h after the death of the patient. 0.83 g of the liver tissue was homogenized in 10 ml of solution containing 0.25 mol/l sucrose, 10 mmol/l sodium phosphate buffer, pH 7.4, and then centrifuged at 1000 g for 10 min. The supernatant was centrifuged at 350,000 g for 20 min and the precipitate obtained was suspended in 500 Al of 10 mmol/l sodium phosphate buffer, pH 7.4 as a sample for electrophoresis. All these procedures were performed at 4 jC.
144
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
3. Results 3.1. Serum APN isoforms from patients with extrahepatic cholestasis and primary biliary cirrhosis As shown in Fig. 1A, SDS-PAGE followed by Western blotting with the APN antiserum detected four bands in each serum sample from cholestatic patients. The molecular masses of these bands were 130-, 140-, 165-, and 260-kDa. The 140- and 260-kDa bands were predominant. All 10 sera from cholestatic patients showed the same pattern of bands. These cholestatic patients include 5 patients with carcinoma of the pancreas head, 2 patients with carcinoma of the bile duct (choledochus), 1 patient with stones of the bile duct (choledochus), (in total, 8 patients of obvious extra-hepatic cholestasis demonstrated by ultrasonography), and 2 patients of primary biliary cirrhosis (intra-hepatic cholestasis). To avoid destruction of sample proteins, we heated serum with sample buffer at 56 jC for 5 min before SDS-PAGE. We also performed the same experiment as that in Fig. 1A except that the five samples were heated with sample buffer at 100 jC for 5 min to ensure dissociation of APN molecules. In this experiment, we obtained the same results as shown in Fig. 1A (data not shown). Besides the 10 patients mentioned above, 1 patient with stones of the gall bladder was considered to have latent extra-hepatic cholestasis because she had transiently shown bile duct dilatation and a high serum bilirubin concentration 2 years before the blood sampling for this study. Fig. 1B shows the analysis of this patient serum by SDS-PAGE and Western blotting. The 260-kDa band was clearly demonstrated. This result suggests that 260-kDa band is sensitive enough to detect extra-hepatic cholestasis even before its mechanical obstruction of the bile duct becomes morphologically clear. 3.2. Serum APN isoform from the patients with the diseases other than extra-hepatic cholestasis or primary biliary cirrhosis Besides extra-hepatic cholestasis and primary biliary cirrhosis, we examined 12 patients, including 4 patients with chronic hepatitis (C-type), 5 patients of cirrhosis (C-type), 2 patients with drug-induced liver
Fig. 1. APN isoforms in sera of patients with extra-hepatic cholestasis. (A) Sera taken from extra-hepatic cholestasis patients (1 Al each) were analyzed by SDS-PAGE followed by Western blotting with anti-APN serum. The diseases were: carcinoma of the bile duct in lane 1; carcinoma of the pancreas in lanes 2 – 4; stones of the bile duct in lane 5. This experiment showed 130-, 140-, 165-, and 260-kDa bands in each serum that are considered to be APN isoforms due to the results shown in Figs. 3 and 4. The 140- and 260-kDa isoforms were predominant. (B) Serum (2 Al) from a patient with gall bladder stones was analyzed. This patient was considered to have latent obstructive cholestasis because she had transiently shown bile duct dilatation and a high serum bilirubin 2 years before the blood sampling for this study. The 260- and the 140-kDa APN isoform were recognized.
dysfunction, and 1 patient with autoimmune hepatitis (taking prednisolone 12.5 mg per day). We clearly recognized the 140-kDa isoform in all these cases although its intensity was considerably lower than in cholestasis cases (extra-hepatic cholestasis and pri-
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
Fig. 2. APN isoform in sera from patients with chronic hepatitis or liver cirrhosis. Serum samples (2 Al each) taken from chronic hepatitis or liver cirrhosis patients were analyzed by SDS-PAGE, followed by Western blotting with anti-APN serum. The diseases were: C-type liver cirrhosis in lanes 1 – 3; C-type chronic hepatitis in lanes 4 – 6. The 140-kDa band is identified.
145
monoclonal APN antibody was compared with original serum samples in simultaneous electrophoresis (SDS-PAGE). Fig. 3 shows that 130-, 140-, 165-, and 260-kDa bands found in serum of an extra-hepatic cholestasis patient (number 1 patient in Fig. 1A) were definitely immunoprecipitated with two different monoclonal APN antibodies: MCS-2 and WM15. We also performed the same experiment as that in Fig. 3 except that the samples were heated with sample buffer at 100 jC for 5 min to ensure dissociation of APN molecules. In this experiment, we obtained the same results as shown in Fig. 3 (data not shown). Fig. 4 shows the two-dimensional electrophoresis of the same patient serum where 130-, 140-, 165-, and 260-kDa bands were immunoprecipitated with MCS-
mary biliary cirrhosis). Fig. 2 shows the results of 5 patients as the examples. The 260-kDa band was faintly visible in only 2 patients: 1 with cirrhosis (Ctype) and 1 with drug-induced liver dysfunction (data not shown). The results of patients with cholestasis and other diseases suggest that the 260-kDa band is a promising marker of cholestasis caused by extra-hepatic obstruction or primary biliary cirrhosis. 3.3. Serum APN isoform from healthy subjects In serum from 11 healthy subjects, the predominantly recognized band was a 140-kDa APN isoform whose electrophoretic mobility (molecular mass) was identical with that in chronic hepatitis or liver cirrhosis (C-type) patient serum. The intensity of the band of the healthy subjects was also similar to that of chronic hepatitis or cirrhosis patients (data not shown). 3.4. Immunoprecipitation of serum APN If the specificity of the APN antiserum is insufficient, there is the possibility that, besides APN bands, non-APN bands may appear in the Western blot. To exclude this possibility and confirm the bands to be true APN isoforms, immunoprecipitated APN with
Fig. 3. SDS-PAGE of immunoprecipitated APN. Serum of one patient with extra-hepatic cholestasis (carcinoma of the bile duct, the serum in lane 1 of Fig. 1A) was used in this SDS-PAGE and Western blotting. The samples electrophoresed were: lane 1, the original serum (1 Al); lane 2, a sample (4 Al) of the same serum immunoprecipitated with MCS-2 monoclonal APN antibody; lane 3, a sample (4 Al) of the same serum immunoprecipitated with WM15 monoclonal APN antibody; lane 4, a sample (4 Al) of the same serum immunoprecipitated with mock antibody (normal mouse IgG). The 130-, 140-, 165-, and 260-kDa bands were immunoprecipitated with monoclonal APN antibodies but not with mock antibody.
146
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
Fig. 4. Two-dimensional electrophoresis of immunoprecipitated APN. An immunoprecipitate from 300 Al of cholestatic patient serum (Fig. 3) with MCS-2 monoclonal APN antibody was subjected to isoelectric focusing (first-dimensional electrophoresis) on an Immobile gel strip. The electrophoresed gel strip and the original serum, indicated as ‘IEF’ and ‘serum’, respectively, were simultaneously subjected to SDS-PAGE (second-dimensional electrophoresis). The 130-, 140-, 165-, and 260-kDa bands (arrowheads) were immunoprecipitated with MCS-2. Additionally, their isoelectric points were similar. In particular, the predominant 140- and 260-kDa bands showed almost identical isoelectric points.
2. Additionally, the isoelectric points of these four bands in Fig. 4 were similar to one another and consistent with those of APN obtained from normal subjects that we previously reported [16]. In particular, the predominant 140- and 260-kDa bands showed almost identical isoelectric points. We also confirmed that the serum of another extrahepatic cholestasis patient (number 2 in Fig. 1A) gave the same results (data not shown) as shown in Figs. 3 and 4. The molecular mass of 260-kDa is substantially higher than the others. We termed this 260-kDa isoform ‘high-molecular-mass APN isoform’. This isoform is resistant to denaturation by SDS and reduction by dithiothreitol. 3.5. APN isoforms of human liver membrane Fig. 5 shows that the 140- and 260-kDa isoforms are also contained in the membrane fragments of human liver tissue, although the proportion of 140to 260-kDa band intensity is different from that of serum.
Fig. 5. APN isoforms in liver membrane. The same serum as in Fig. 3 (2 Al) was electrophoresed in lane 1 and liver membrane suspension (2 Al) obtained as described in Section 2.5 was in lane 2. The 140- and 260-kDa bands were also recognized in human liver membrane.
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
4. Discussion Although it has been already reported that serum leucine aminopeptidase and alanine aminopeptidase activity are mainly derived from APN/CD13 [13, 17,18], electrophoretic analysis of serum APN isoforms with anti-APN antibody has not been previously reported. In the past report of non-denaturing electrophoresis, serum APN isoform patterns observed in the same disease were often different from one another [9]. By one-dimensional SDS-PAGE with anti-APN antiserum, we recognized the same pattern of APN bands in cholestatic patients, consisting of 130-, 140-, 165-, and 260-kDa bands. All four bands were immunoprecipitated with two monoclonal APN antibodies, thus indicating that each of them has two epitopes of APN; they were authentic APN isoforms. Two-dimensional electrophoresis also confirmed the immunoprecipitation of the four bands with a monoclonal APN antibody. The high-molecular-mass APN isoform (260-kDa) was observed in all the sera of extra-hepatic cholestasis and primary biliary cirrhosis. In the cases of liver dysfunction of other causes, however, no or very faint bands of high-molecular-mass APN were recognized. Additionally, we observed high-molecular-mass APN in a case of gall stones where no dilatation of bile ducts was recognized by ultrasonographic examination at the time of blood sampling. These results suggest that the high-molecular-mass APN isoform is closely related to extra-hepatic cholestasis and primary biliary cirrhosis and is a promising marker of cholestasis of these diseases. Considering the molecular weight of human and porcine APNs previously reported, the bands of 130-, 140-, and 165-kDa must be monomers of APN [3,19] with variations of their N-glycosylation [3] and the 260-kDa band, the high-molecular-mass APN isoform, is highly likely to be an APN homodimer [19]. Human APN molecules are known to be produced in cells as a homodimer whose subunits (APNs) are bound to each other with a noncovalent bond [3]. The noncovalently formed homodimer should dissociate to be presented as monomers under reducing and denaturing SDS-PAGE conditions. Our results suggest that the subunits (APN molecules) of the 260-kDa isoform (high-molecular-mass isoform) are bound covalently by a non-disulfide bond. The
147
results of the 260-kDa isoform immunoprecipitated with two monoclonal APN antibodies show that this isoform contains two APN epitopes. Thus, the alternative hypothesis that should be considered is that the 260-kDa isoform is a complex consisting of an APN and another different protein molecule. The result that the 260-kDa isoform has an almost identical isoelectric point to the 140-kDa APN isoform supports the hypothesis of the APN homodimer rather than that of APN and another protein complex. Additionally, 260-kDa is approximately two times 140-kDa. Therefore, we consider that the high-molecular-mass APN isoform of 260-kDa in cholestatic serum is an APN homodimer whose subunits are covalently bound to each other. Being compatible with the assumption that the 260-kDa APN isoform is released from liver, Fig. 5 shows that human liver definitely contains the highmolecular-mass (260-kDa) APN, although the intensity of 260-kDa band was much less than 140-kDa band as compared with serum. The liver tissue sample used was obtained from a patient without cholestasis. To further support the above assumption, liver tissue from cholestatic patients should also be examined because 260-kDa isoform is supposed to increase in cholestasis from our results. It was previously observed that some cross-linking agents, e.g., dimethyl adipimidate, bind covalently the two APN molecules associated by noncovalent bond [19,20]. However, it has not been reported that human serum samples that were not exposed to any artificial or experimental cross-linking treatment contain a covalently formed homodimer of APN. Our results indicate that the covalently formed APN homodimer increases in cholestatic serum. Therefore, the occurrence of this APN dimer in serum and the mechanism of its production in cholestatic cases are of interest. The mechanism should include increased cross-linking reaction between two APN molecules in cholestatic condition. It was previously reported that, in extra-hepatic cholestasis, oxide free radicals in the liver, including hydroxyl radical (HOS), increase due to hydrophobic bile acid accumulation in the liver and act as a major pathogenic factor in cholestatic liver injury [21 –23]. Also, in primary biliary cirrhosis, lipid peroxidation was reported to increase in the liver, suggesting
148
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149
increased production of free radicals [24]. Hydroxyl free radical (HOS) is known to dimerize proteins; the proteins exposed to hydroxyl radical undergo progressive cross-linking to form dimers because the tyrosine residues of the proteins are oxidized by hydroxyl radical and intermolecular bityrosine are formed [25]. During the physiological synthesis in endoplasmic reticulum, two APN molecules are associated noncovalently. In cholestatic cases, increased free radicals should bind covalently the two APN molecules associated in the synthetic process of APN. Therefore, the high-molecular-mass APN isoform possibly reflects the increased production of free radicals in cholestatic cases that plays a pivotal role in the pathogenesis of cholestatic liver injury. Further study is necessary to clarify the sensitivity and specificity of high-molecular-mass APN isoform for diagnosing cholestasis and to reveal the mechanism of covalent APN homodimer formation.
Acknowledgements We thank Prof. Yasuhiro Watanabe (Department of Pathological Biochemistry, Hokkaido College of Pharmacy) for his donation of rabbit anti-aminopeptidase N serum.
References [1] Look AT, Ashmun RA, Shapiro LH, Peiper SC. Human myeloid plasma membrane glycoprotein CD13 (gp150) is identical to aminopeptidase N. J Clin Invest 1989;83:1299 – 307. [2] Olsen J, Cowell GM, Konigshofer E, Danielsen EM, Moller J, Laustsen L, et al. Complete amino acid sequence of human intestinal aminopeptidase N as deduced from cloned cDNA. FEBS Lett 1988;238:307 – 14. [3] Shipp MA, Look AT. Hematopoietic differentiation antigens that are membrane-associated enzymes: cutting is the key! Blood 1993;82:1052 – 70. [4] Riemann D, Kehlen A, Langner J. CD13-not just a marker in leukemia typing. Immunol Today 1999;20:83 – 8. [5] Reichling JJ, Kaplan MM. Clinical use of serum enzymes in liver disease. Dig Dis Sci 1988;33:1601 – 14. [6] Shinkai K, Akedo H. A multienzyme complex in serum of hepatic cancer. Cancer Res 1972;32:2307 – 13. [7] Ideo G, Ronchi G. Sephadex gel filtration of g-glutamyl-transpeptidase, alkaline phosphatase and leucine aminopeptidase in the serum of patients affected by various liver diseases. Z Klin Chem Klin Biochem 1972;10:S211 – 4.
[8] Deng JT, Hoylaerts MF, Nouwen EJ, De Broe ME, Van Hoof VO. Purification of circulating liver plasma membrane fragments using a monoclonal antileucine aminopeptidase antibody. Hepatology 1996;23:445 – 54. [9] Dingjan PG, Jongeneel J, Postma T, Stroes JAP. Evaluation of the determination of alanine naphthyl-amidase isoenzymes by agar and polyacrylamide disc electrophoresis for the diagnosis of pancreatic diseases. Clin Chim Acta 1973;49: 215 – 24. [10] Sanderink GJCM, Artur Y, Galteau MM, Wellman-Bednawska M, Siest G. Multiple forms of serum aminopeptidases separated by micro two-dimensional electrophoresis under non-denaturing conditions. Electrophoresis 1986;7:471 – 5. [11] Sanderink GJ, Artur Y, Paille F, Siest G. Clinical significance of a new isoform of serum alanine aminopeptidase; relationship of liver disease and alcohol consumption. Clin Chim Acta 1989;179:23 – 32. [12] Hussain MM, Tranum-Jensen J, Noren O, Sjostrom H, Christiansen K. Reconstitution of purified amphiphilic pig intestinal microvillus aminopeptidase. Mode of membrane insertion and morphology. Biochem J 1981;199:179 – 86. [13] Kawai M, Hara Y, Kubota T, Shiba K, Hosaki S. A family with high serum leucine aminopeptidase activity derived from a novel variant CD13. Clin Chem 1998;44:215 – 20. [14] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227: 680 – 5. [15] Watanabe Y, Iwaki-Egawa S, Mizukoshi H, Fujimoto Y. Identification of an alanine aminopeptidase in human maternal serum as a membrane-bound aminopeptidase N. Biol Chem Hoppe-Seyler 1995;376:397 – 400. [16] Kawai M, Hara Y, Miyazato I, Hosaki S. Novel, aberrantly truncated isoform of serum CD13 in a family with high serum aminopeptidase N (CD13) activity. Clin Chem 2001;47: 223 – 30. [17] Favaloro EJ, Browning T, Nandurkar H. The hepatobiliary disease marker serum alanine aminopeptidase predominantly comprises an isoform of the haematological myeloid differentiation antigen and leukaemia marker CD-13/gp150. Clin Chim Acta 1993;220:81 – 90. [18] Watanabe Y, Ito K, Iwaki-Egawa S, Yamaguchi R, Fujimoto Y. Aminopeptidase N in sera of healthy subjects is a different N-terminal processed derivative from the one obtained from maternal serum. Mol Genet Metab 1998;63:289 – 94. [19] Danielsen EM. Biosynthesis of intestinal microvillar proteins. Dimerization of aminopeptidase N and lactase-phlorizin hydrolase. Biochemistry 1990;29:305 – 8. [20] Danielsen EM. Perturbation of intestinal microvillar enzyme biosynthesis by amino acid analogs. J Biol Chem 1990;265: 14566 – 71. [21] Sokol RJ, Devereaux M, Khandwala R, O’Brien K. Evidence for involvement of oxygen free radicals in bile acid toxicity to isolated rat hepatocytes. Hepatology 1993;17:869 – 81. [22] Sokol RJ, Winklhofer-Roob BM, Devereaux MW, McKim Jr JM. Generation of hydroperoxides in isolated rat hepatocytes and hepatic mitochondria exposed to hydrophobic bile acids. Gastroenterology 1995;109:1249 – 56.
M. Kawai et al. / Clinica Chimica Acta 330 (2003) 141–149 [23] Tsai LY, Lee KT, Liu TZ. Evidence for accelerated generation of hydroxyl radicals in experimental obstructive jaundice of rats. Free Radic Biol Med 1998;24:732 – 7. [24] Kawamura K, Kobayashi Y, Kageyama F, Kawasaki T, Nagasawa M, Toyokuni S, et al. Enhanced hepatic lipid peroxida-
149
tion in patients with primary biliary cirrhosis. Am J Gastroenterol 2000;95:3596 – 601. [25] Davies KJA, Delsignore ME. Protein damage and degradation by oxygen radicals. III. Modification of secondary and tertiary structure. J Biol Chem 1987;262:9908 – 13.