help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Journal of Clinical Endocrinology & Metabolism , doi:10.1210/jc.2004-1600
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hattori, N.
Right arrow Articles by Inagaki, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hattori, N.
Right arrow Articles by Inagaki, C.
Right arrowPubmed/NCBI databases
*Compound via MeSH
*Substance via MeSH
Hazardous Substances DB
*L-SERINE
Related Collections
Right arrow Neuroendocrinology and Pituitary
The Journal of Clinical Endocrinology & Metabolism Vol. 90, No. 5 3036-3044
Copyright © 2005 by The Endocrine Society

Immunoglobulin G Subclasses and Prolactin (PRL) Isoforms in Macroprolactinemia Due to Anti-PRL Autoantibodies

Naoki Hattori, Katsuji Ikekubo, Yasuhisa Nakaya, Kaori Kitagawa and Chiyoko Inagaki

Department of Pharmacology (N.H., Y.N., K.K., C.I.), Kansai Medical University, Osaka and Department of Nuclear Medicine (K.I.), Kobe City Hospital, Kobe 650, Japan

Address all correspondence and requests for reprints to: Naoki Hattori, Department of Pharmacology, Kansai Medical University, 10–15 Fumizono-cho, Moriguchi-City, Osaka 570-8506, Japan. E-mail: hattorin{at}takii.kmu.ac.jp.


   Abstract
Top
 Abstract
Introduction
 Subjects and Methods
 Results
 Discussion
 References
 
Although macroprolactinemia due to antiprolactin (anti-PRL) autoantibodies is not uncommon among hyperprolactinemic patients, the pathogenesis of such macroprolactinemia is still unknown. We examined IgG subclasses of anti-PRL autoantibodies by enzyme immunoassay, and PRL phosphorylation and isoforms by Western blotting, mass spectrometry, and two-dimensional electrophoresis in six patients with anti-PRL autoantibodies and in 29 controls. PRL-specific IgG subclasses in patients with anti-PRL autoantibodies were heterogeneous, but five of six patients showed IgG4 predominance, which is known to be produced by chronic antigen stimulation. Western blot and mass spectrometric analyses revealed that human pituitary PRL was phosphorylated at serine 194 and serine 163, whereas serine 163 in serum PRL was dephosphorylated. On two-dimensional electrophoresis, serum PRL mainly consisted of isoform with isoelectric point (pI) 6.58 in control hyperprolactinemic patients, whereas acidic isoforms (pIs 6.43 and 6.29) were also observed in patients with anti-PRL autoantibodies. Our data first demonstrate that human pituitary PRL is serine phosphorylated and partially dephosphorylated in serum, and suggest that the acidic isoforms may give rise to chronic antigen stimulation in patients with anti-PRL autoantibodies.


   Introduction
Top
 Abstract
 Introduction
Subjects and Methods
 Results
 Discussion
 References
 
SERUM LEVELS OF prolactin (PRL) increase during pregnancy, PRL-secreting pituitary adenoma, hypothyroidism, chest wall diseases, hepatorenal disorders, and taking pharmacological agents (1). PRL consists of several molecular forms: little PRL (molecular mass, 23 kDa), big PRL (50 kDa), and big-big PRL (>150 kDa) on gel chromatography (2). Although the major circulating form is little PRL, some patients with hyperprolactinemia show a higher proportion of big-big PRL in the serum, which is termed macroprolactinemia. The first case was reported by Whittaker et al. (3); the patient characteristically lacked clinical symptoms of hyperprolactinemia, such as amenorrhea and galactorrhea, and spontaneous pregnancy was possible despite hyperprolactinemia. Recently, the prevalence of macroprolactinemia has been investigated in a large population at several clinical centers and reported to range from 10–46% among patients with hyperprolactinemia (4, 5, 6). This value was unexpectedly high because a prevalence of idiopathic hyperprolactinemia of unknown cause had been reported to be 14–40% (7, 8), suggesting that macroprolactinemia is possibly a major cause of idiopathic hyperprolactinemia. This new entity of hyperprolactinemia is clinically important because repeated examinations by computerized axial tomography or magnetic resonance imaging of the pituitary gland have been performed to pursue a possible microadenoma in the pituitary gland in patients with idiopathic hyperprolactinemia, and these patients have been treated with dopamine agonist therapy over the long term. Because patients with macroprolactinemia typically lack clinical symptoms of hyperprolactinemia and spontaneous pregnancy is possible without any treatment, identification of macroprolactinemia gives clinical doctors useful information in the treatment of hyperprolactinemic patients.

Despite many reports about the prevalence, clinical manifestations, and laboratory diagnosis of macroprolactinemia (4, 5, 6, 9, 10, 11), little is known of its causes. Several investigators and we have identified anti-PRL autoantibodies in sera of patients with macroprolactinemia (12, 13, 14, 15, 16, 17, 18, 19). Although macroprolactinemia is heterogeneous (14), a recent study has demonstrated that most cases of macroprolactinemia possess PRL-IgG complexes (20). Because we previously showed that the autoantibody-bound PRL was more slowly cleared from the circulation than free PRL, leading to the accumulation of PRL in blood (15), presence of anti-PRL autoantibodies may be a primary cause of hyperprolactinemia. Autoantibodies to several other hormones, such as thyroid hormone (21) and insulin (22), have been reported, although the causes leading to the production of these autoantibodies are also unclear. Alterations of the antigenicity of the hormones might be involved besides disorders of the host immune system.

The aims of this study are: 1) to clarify whether chronic antigen stimulation exists in patients with anti-PRL autoantibodies by examining IgG subclasses of the autoantibodies; 2) to examine whether modification or demodification of pituitary PRL occurs in serum that may be related to the acquirement of antigenicity, focusing on PRL phosphorylation; and 3) to examine whether serum PRL isoforms on two-dimensional (2D)-electrophoresis are different between patients with and without anti-PRL autoantibodies, which may represent different modification on PRL molecules.


   Subjects and Methods
Top
 Abstract
 Introduction
 Subjects and Methods
Results
 Discussion
 References
 
Subjects

We studied six hyperprolactinemic patients with anti-PRL autoantibodies (age, 28–60 yr) and 29 controls including nine hyperprolactinemic subjects (six patients with prolactinoma and three pregnant women; age, 26–59 yr) without anti-PRL autoantibodies and 20 normoprolactinemic subjects (age, 24–61 yr) (Table 1Go). Patients with anti-PRL autoantibodies were screened from hyperprolactinemic sera that were sent to our laboratory for PRL measurement by the polyethylene glycol (PEG) method. The patients were diagnosed as having anti-PRL autoantibodies because of the following findings: 1) the recovery of PRL in the supernatant (free PRL/total PRL x 100) after treating the serum with 12.5% PEG, which precipitates {gamma}-globulin fractions, less than 40% (4); 2) substantial amounts of PRL in the fractions greater than 150 kDa (big-big PRL) on gel chromatography (15); and 3) substantial amounts of PRL trapped to a protein G column (Amersham Biosciences Corp., Uppsala, Sweden), which binds IgG (13). Medical records showed that all six patients with anti-PRL autoantibodies had been diagnosed as having idiopathic hyperprolactinemia and did not have any signs of systemic autoimmune disorders. Enzyme immunoassay for human PRL, PEG precipitation method, gel chromatography, and affinity purification of IgG with a protein G column were performed as previously described (13, 15). The study was approved by the Ethics Committee of Kobe City Hospital, and informed consent was obtained from all participants.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Clinical characteristics of patients with anti-PRL autoantibodies

 
Human pituitary PRL

Human pituitary PRL (hPRL-I-9) was kindly supplied by Dr. A. F. Parlow, a scientific director of the National Hormone and Pituitary Program. Human pituitary glands were collected within 24 h after death and subsequently preserved by freezing at –80 C. The glandular extraction and purification of the PRL was performed in the cold in the presence of proteolytic enzyme inhibitors. The purification procedures included solvent extraction, salt and ethanol fractionation, gel chromatography on Sephadex G100, and ion exchange chromatography. The human PRL isolated and purified in this way showed a single, sharp symmetrical peak on analytical gel filtration. Biopotency and immunopotency were in the range of 35 IU/mg (Dr. Parlow’s personal communication).

Measurement of total serum IgG subclasses

The concentrations of total serum IgG subclasses were measured with an ELISA kit (Zymed Laboratories, Inc., San Francisco, CA). Fifty microliters of serum (1:2500 dilution) was added to the antihuman IgG-coated 96-well microplate, and then 50 µl mouse monoclonal antihuman IgG1, IgG2, IgG3, or IgG4 antibody was added. The mixture was incubated at room temperature for 30 min and washed with PBS three times. Then 100 µl antimouse IgG-horseradish peroxidase (HRP) conjugate was added to the plate and incubated at room temperature for 30 min. After washing with PBS three times, 50 µl of 3,3',5,5'-tetramethylbenzidine solution and 50 µl of 0.02% H2O2 were added and incubated at room temperature for 10 min. The reaction was stopped with 100 µl of 1 mol/liter phosphoric acid and the absorbance was measured at a wavelength of 450 nm with a microplate reader model 550 (Bio-Rad Laboratories, Inc., Hercules, CA).

Measurement of IgG subclasses of anti-PRL autoantibodies

To evaluate the subclasses of IgG, which are bound to serum PRL, we established a sensitive and specific enzyme immunoassay system. Serum samples of 10 µl were incubated with antihuman PRL antiserum (NIDDK-anti-hPRL-IC-5)-coated polystyrene balls (Precision Plastic Ball Co. Ltd., Chicago, IL) in duplicate at 37 C for 6 h with continuous shaking. After washing with 150 mmol/liter saline, the balls were incubated with mouse antihuman IgG subtype-specific antisera (Zymed Laboratories, Inc.) at 4 C overnight. The balls were incubated with goat antimouse IgG-HRP (Cappel, Aurora, OH) at 20 C for another 4 h with continuous shaking. In this system, serum PRL is trapped to the antihuman PRL antiserum-coated balls, and IgG subclasses of the autoantibodies bound to PRL can be determined using mouse antihuman IgG subclass-specific antisera. After washing with 150 mmol/liter saline, the peroxidase activities bound to the balls were assayed by an enzyme reaction using 3-(p-hydroxyphenyl) propionic acid (Aldrich Chemical Co., Milwaukee, WI) as a substrate at 30 C for 90 min. HRP catalyzes the oxidation of 3-(p-hydroxyphenyl) propionic acid, a hydrogen-donating substrate, with hydrogen peroxide, forming a fluorescent substance. Fluorescence intensity, at 320 nm for excitation and 405 nm for emission, was measured by a fluorophotometer (F-2000; Hitachi, Ibaragi, Japan). The intra- and interassay coefficients of variation were less than 10%. The levels of PRL-bound IgG subclasses were shown as the ratios (percent) of fluorescence intensity of the samples to that of internal control serum, which was obtained from a healthy volunteer with normal serum PRL concentration and no anti-PRL autoantibodies. Serum samples from 29 controls were treated in the same way, and the mean ± 2 SD was determined as nonspecific binding.

Electrophoresis

All reagents for electrophoresis were purchased from Amersham Biosciences Corp. SDS-PAGE was performed as previously described (15). 2D-electrophoresis was performed with a Multiphor II system (Amersham Biosciences Corp.) according to the manufacturer’s instructions. PRL in the patient’s serum (1 ml) was immunoprecipitated with 5 µl rabbit antihuman PRL antiserum (NIDDK-anti-hPRL-IC-5) at room temperature for 3 h and with goat antirabbit IgG (radioisotope; Daiichi Pharmaceutical Company Ltd., Tokyo, Japan) for another 3 h. Then the serum was centrifuged at 14,000 rpm for 20 min, and the precipitate was dissolved in 80 µl sample buffer containing 8 mol/liter urea that was treated with Amberlite, 40 mmol/liter dithiothreitol, 2% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate, 0.5% immobilized pH gradient buffer, and 0.001% orange G. The samples were applied on the preswollen immobilized 18-cm strips of pH 4–7 with the same buffer at the position of pH 5. Focusing was carried out on a Multiphor II system at 500 V for 6 h and at 3500 V for 10 h at 15 C. After isoelectric focusing, the strip was incubated for 10 min in equilibration buffer (50 mmol/liter Tris-HCl, pH 6.8; 6 mol/liter urea; 30% glycerol, and 1% sodium dodecyl sulfate) containing 16.2 mmol/liter dithiothreitol, followed by 10 min incubation in equilibration buffer containing 243 mmol/liter iodoacetamide. The equilibrated strip was transferred onto the Excel Gel (gradient 8–18%, 245 x 110 mm; Amersham Biosciences Corp.), and electrophoresis was carried out at 15 C on a Multiphor II system at 600 V, 20 mA for 35 min and at 600 V, 50 mA for 75 min. pI of each spot was calculated based on the linearity of pH on the strip.

Western blot analysis

After electrophoresis, the proteins in the gel were electrophoretically transferred onto polyvinylidene difluoride membranes (Bio-Rad Laboratories, Inc.). The membranes were blocked with 10 mmol/liter Tris-buffered saline (pH 7.4) containing 150 mmol/liter NaCl, 0.05% Tween 20, and 10 mmol/liter sodium fluoride plus 5% skim milk for PRL or plus 5% BSA for phosphoserine, phosphothreonine, and phosphotyrosine at 4 C overnight. The membranes were then incubated with rabbit antihuman PRL antiserum (NIDDK-anti-hPRL-IC-5), mouse monoclonal antibody to phosphoserine (clone 7F12) (Affinity Research Products Ltd., Mamhead Castle, UK), mouse monoclonal antibody to phosphothreonine (Cell Signaling Technology, Inc., Beverly, MA), or mouse monoclonal antibody to phosphotyrosine (Transduction Laboratories, Inc. and PharMingen, San Diego, CA). After washing, the membranes were incubated with goat antirabbit F(ab')2-HRP conjugate (ICN Biomedicals, Inc., Aurora, OH) for human PRL or goat antimouse F(ab')2-HRP conjugate (Zymed Laboratories, Inc.) for phosphoserine, phosphothreonine, and phosphotyrosine. The signals were visualized by using chemiluminescence (PerkinElmer Life Sciences, Boston, MA) and x-ray film exposure (X-Omat; Eastman Kodak Co., Rochester, NY). The density of each spot was quantified by densitometry (DMU-33C; Advantec Toyo, Osaka, Japan). Jurkat T cell line was obtained from RIKEN Cell Bank (Tsukuba, Japan), and A549 cell line was provided by the Department of Respiratory Medicine, Kyoto University. Total cell lysate was prepared as previously described (23).

Matrix-assisted laser desorption ionization time of flight (MALDI-TOF) mass spectrometry

Mass spectrometric identification of peptides was performed with a Biospectrometry Workstation Voyager-DE PRO (Applied Biosystems, Foster City, CA). Mass spectra were recorded at an accelerating voltage of 20,000 V in the positive ion mode with a reflector. Typically, spectra from 300 shots were summed to obtain the final spectrum. After 2D-electrophoresis of human pituitary PRL or immunoprecipitated serum PRL, the gel was stained with Sypro Ruby (Molecular Probes, Inc., Eugene, OR) overnight. The stained PRL spots with 1.5-mm diameter were picked by spot picker (The Gel Company, San Francisco, CA) under UV, destained with 50% acetonitrile (Wako Pure Chemical Industries Ltd., Osaka, Japan)/25 mmol/liter ammonium bicarbonate, dehydrated with 100% acetonitrile, and dried with SpeedVac (Life Sciences International, Tokyo, Japan). The dried gel pieces were rehydrated with modified trypsin (10 µg/ml; Promega Corp., Madison, WI) in 50 mmol/liter ammonium bicarbonate. After overnight digestion, the peptides were extracted with 50% acetonitrile containing 5% trifluoroacetic acid (TFA) (Wako Pure Chemical Industries Ltd.). After evaporating the acetonitrile and TFA with a SpeedVac, the solution containing peptide fragments was desalted and concentrated with Zip Tip C18 (Millipore Corp., Bedford, MA) according to the manufacturer’s instructions. For MALDI-TOF mass spectrometric analysis, peptides were eluted with 1 µl saturated matrix (4-hydroxy-cyanocinnamic acid; Sigma, St. Louis, MO) solution containing 50% acetonitrile and 0.1% TFA. The eluate was spotted onto a sample plate. The mass spectra were externally calibrated with bradykinin, angiotensin I, fibrinopeptide B, neurotensin, ACTH, and insulin (Applied Biosystems). The spectra derived from peptides usually consisted of four peaks due to the presence of natural isotopes such as 13C. The monoisotopic first peak was used for the estimation of mass values. The peptide mass maps produced by the MALDI-TOF mass spectrometry were searched in a published database (National Center for Biotechnology Information) using Mascot program (http://www.matrixscience.com/; Matrix Science Inc., Boston, MA) with parameters including possible modifications of peptides such as: alkylation of cysteine; oxidation of methionine; mass tolerance, ± 0.2 Da; mass range, 800-4000; and allowing miscleavage by trypsin up to 1.

Statistical analyses

Statistical analyses were performed using Statview 4.5 software (SAS Institute, Inc., Cary, NC). Unpaired Student’s t tests were used to compare the densities of isoforms of PRL. All values are presented as mean ± SEM, and differences were considered to be statistically significant at P < 0.05.


   Results
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
Discussion
 References
 
Clinical characteristics of patients with anti-PRL autoantibodies

Clinical characteristics of patients with anti-PRL autoantibodies (n = 6) are shown in comparison with nine control hyperprolactinemic subjects (Table 1Go). All patients with anti-PRL autoantibodies lacked clinical symptoms of hyperprolactinemia, such as amenorrhea and galactorrhea, and none of them showed any signs suggestive of systemic autoimmune disorders. All the patients fulfilled the criteria of possessing anti-PRL autoantibodies: 1) the recovery of PRL after 12.5% PEG treatment, which is known to precipitate {gamma}-globulin fractions, was 7.9 ± 2.7%, whereas that in controls was 104 ± 12.7%; 2) gel chromatography disclosed that 80.4 ± 2.9% of PRL was eluted at a molecular mass greater than 150 kDa (big-big PRL), whereas less than 1% of PRL was eluted in such a high molecular mass fraction in the controls; and 3) when IgG was affinity purified from the sera using a protein G column, which is known to bind IgG fraction, 62.4 ± 6.4% of PRL was bound to the column, whereas no PRL was contained in the IgG fraction in the controls. Different results by the three techniques could be explained by a possible dissociation of PRL from IgG during gel filtration or affinity purification on a protein G column, or the other components than IgG-PRL complex such as aggregates of PRL may be involved in PEG precipitates or big-big PRL fraction on gel chromatography.

IgG subclasses of anti-PRL autoantibodies

Total serum levels of IgG subclasses in patients with anti-PRL autoantibodies are shown in Fig. 1AGo. The concentrations of total IgG1, IgG2, and IgG3 subclasses in the patients were within mean ± 2 SD in control subjects. However, two patients (cases 1 and 3) had total IgG4 subclass levels greater than the normal range. The levels of IgG subclasses bound to PRL are shown as the ratios of the fluorescence intensity of the samples to that of the internal control serum in Fig. 1BGo. When values greater than 2 SD above the means of the normal controls (nonspecific binding) were scored as positive, five of six patients showed positive for IgG4 subclass. Case 1 was positive for all subclasses, and case 2 was positive only for the IgG1 subclass.



View larger version (29K):
[in this window]
[in a new window]
  		FIG. 1. Serum IgG subclasses in patients with anti-PRL autoantibodies. A, Serum values of total IgG subclasses that were measured by ELISA are shown in comparison with the mean ± 2 SD in controls (n = 29, shadow). B, Serum levels of PRL-specific IgG subclasses in patients with anti-PRL autoantibodies are shown in comparison with nonspecific binding in controls (n = 29, shadow), using the fluorescence intensity of internal control serum as 100%. Data are expressed as mean ± SEM (n = 3–7 different samplings).

 
Phosphorylation of human pituitary PRL shown by Western blotting

Because PRL is known to possess serine, threonine, and tyrosine phosphorylation sites, we examined whether human pituitary PRL was phosphorylated on these sites by Western blotting. As shown in Fig. 2AGo, SDS-PAGE of human pituitary PRL followed by Western blotting with antiphosphoserine, antiphosphothreonine, and antiphosphotyrosine antibodies revealed that only serine residue was phosphorylated. Antibodies for phosphothreonine and phosphotyrosine were shown to work in positive control experiments using calyculin A for phosphothreonine in Jurkat T cell line and using epidermal growth factor (EGF) for phosphotyrosine in A549 alveolar epithelial cell line (Fig. 2BGo). Many kinds of proteins were phosphorylated in threonine residues after the treatment with calyculin A, a serine and threonine phosphatase inhibitor (left). EGF receptor (arrowhead) was phosphorylated in tyrosine residues after the treatment with EGF (right), as previously reported (24).



View larger version (37K):
[in this window]
[in a new window]
  		FIG. 2. Western blots of phosphorylated PRL. A, Human pituitary PRL (10 µg) was run on SDS-PAGE and blotted with antiphosphoserine (P-serine), phosphothreonine (P-threonine), and phosphotyrosine (P-tyrosine) antibodies (upper panel) and reprobed with antihuman PRL antiserum (lower panel). B, Positive control experiments of phosphothreonine and phosphotyrosine antibodies: Jurkat cells were treated with 0.1 mol/liter calyculin A for 30 min, and the whole-cell lysate (20 µg) was run on SDS-PAGE followed by Western blotting with antiphosphothreonine antibody (left). A549 cells were treated with 200 µg/liter EGF for 10 min, and the whole-cell lysate (20 µg) was run on SDS-PAGE followed by Western blotting with antiphosphotyrosine antibody (right). An arrow indicates the position of the EGF receptor.

 
Determination of serine phosphorylation sites in human pituitary PRL by mass spectrometry

Mass spectrometry is a powerful tool to determine not only peptides but also the modifications, such as phosphorylation. We then examined the position of serine phosphorylation in human pituitary PRL. Human pituitary PRL consisted of two isoforms (pIs 6.43 and 6.29) on 2D-electrophoresis (Fig. 3AGo). The gel spots stained with Sypro Ruby were subjected to in-gel digestion with trypsin and analyzed by MALDI-TOF mass spectrometry. Each peak originating from peptides usually consisted of four peaks due to the presence of isotopes such as 13C, and the first monoisotopic peak was used for the determination of mass values. In both isoforms, several peaks with mass values corresponding to trypsin-digested PRL fragments were observed and identified to derive from PRL by Mascot software (Fig. 3BGo). Because phosphorylation in one amino acid increases the mass value by 79.97, we examined whether there are peaks with mass values of PRL fragments plus 79.97 in the spectra. Two peaks with mass values of 1758.85 and 2018.99 were identified to be phosphorylated PRL fragments corresponding to the sequence 193–205 and 154–170, respectively. The former fragment contains serine194, and the latter one contains serine163, suggesting that these serine residues are phosphorylated (Fig. 3CGo).



View larger version (41K):
[in this window]
[in a new window]
  		FIG. 3. Analysis of trypsin-digested fragments of human pituitary PRL by MALDI-TOF mass spectrometry. A, Representative image of human pituitary PRL stained with Sypro Ruby on 2D-electrophoresis. B, Representative profiles of mass spectra of the digested PRL isoforms with pIs 6.43 and 6.29. Asterisks indicate the peaks with mass values corresponding to the trypsin-digested PRL fragments. C, Spectra of phosphorylated peptide fragments of PRL and the corresponding amino acid sequence. Each peak seen in B usually consisted of four peaks due to the presence of isotopes such as 13C, and the first monoisotopic peak (*) was used for the determination of mass value. The amino acid is numbered from the start of translation (methionine). Serine 194 and serine 163 correspond to serine 166 and serine 135, respectively, in mature PRL in which 28-amino-acid signal peptide is cleaved.

 
Determination of serine phosphorylation sites in human serum PRL by mass spectrometry

Because we could not confirm serine phosphorylation in serum by Western blotting because antiphosphoserine antiserum required large amounts of PRL for its use and the amounts of serum samples were limited, we tried to examine whether serum PRL was also phosphorylated using MALDI-TOF mass spectrometry. Serum PRL from patients with anti-PRL autoantibodies and that from control hyperprolactinemic patients without anti-PRL autoantibodies were immunoprecipitated with human PRL antiserum, separated on 2D-electrophoresis and stained with Sypro Ruby (Fig. 4AGo). Because the first (rabbit antihuman PRL) and the second (goat antirabbit {gamma}-globulin) antibodies for immunoprecipitation were carried over in the final 2D-gels, mass spectrometric analysis showed that most of the stained proteins around the area of PRL spots were IgG light chains. We picked 20–30 gel spots with 1.5-mm diameter along the lower border (molecular mass, 23 kDa) of the broad protein bands of IgG light chains. The gel spots were subjected to in-gel digestion with trypsin and analyzed by MALDI-TOF mass spectrometry. Using Mascot software, PRL signals were identified in the gel spots with pIs 6.58, 6.43, and 6.29 in serum from patients with anti-PRL autoantibodies, whereas they were only identified in the gel spot with pI 6.58 in serum from control hyperprolactinemic patients (Fig. 4BGo). However, because the signal intensity of phosphorylated PRL fragments was not so strong, relevant information about the phosphorylation states could be obtained only in four patients with anti-PRL autoantibodies and two control hyperprolactinemic patients. We found the same peak with mass value of 1758.85 corresponding to a phosphorylated PRL fragment (residues 193–205) as in the pituitary PRL in every spot. On the other hand, the peak with mass value of 2018.99, which corresponded to another phosphorylated PRL fragment (residues 154–170) in the pituitary PRL, could not be observed, and instead the same fragment without phosphorylation (mass value 1939.02) was found in every spot, suggesting that serine163 is likely to be dephosphorylated in serum (Fig. 4CGo).



View larger version (51K):
[in this window]
[in a new window]
  		FIG. 4. Analysis of trypsin-digested fragments of immunoprecipitated serum PRL by MALDI-TOF mass spectrometry. A, Representative images of immunoprecipitated serum PRL stained with Sypro Ruby on 2D-electrophoresis in patients with anti-PRL autoantibodies (left) and in control hyperprolactinemic patients (right). The broad protein bands were IgG light chains originating from the first and the second antibodies used for immunoprecipitation. The signals of PRL were obtained along the lower border (molecular mass, 23 kDa). B, Representative profiles of mass spectra of the digested PRL isoforms with pIs 6.58, 6.43, and 6.29. Asterisks indicate the peaks with mass values corresponding to the trypsin-digested PRL fragments. C, Spectra of phosphorylated and nonphosphorylated peptide fragments of PRL and the corresponding amino acid sequence. The amino acid is numbered from the start of translation (methionine). Serine 194 and serine 163 correspond to serine 166 and serine 135, respectively, in mature PRL in which 28-amino-acid signal peptide is cleaved.

 
Comparison of serum PRL isoforms between hyperprolactinemic patients with and without anti-PRL autoantibodies

Representative profiles of PRL isoforms in human pituitary PRL and serum PRL with and without anti-PRL autoantibodies are shown in Fig. 5AGo. PRL immunoprecipitated from hyperprolactinemic sera in controls was focused mainly on pI 6.58, whereas PRL in the sera from patients with anti-PRL autoantibodies contained not only isoform with pI 6.58 but also substantial amounts of isoforms with pIs 6.43 and 6.29 similar to pituitary PRL. In sera from patients with anti-PRL autoantibodies, the ratio of PRL isoform (pI 6.58) to total PRL was significantly lower, and the ratios of PRL isoforms (pIs 6.43 and 6.29) were significantly higher, than in control hyperprolactinemic sera (Fig. 5BGo).



View larger version (26K):
[in this window]
[in a new window]
  		FIG. 5. Comparison of the profiles of pituitary PRL and serum PRL with and without anti-PRL autoantibodies (Ab). A, Representative Western blots of PRL on 2D-electrophoresis. B, Quantitative analysis of the density of the spots in pituitary PRL (three separate experiments), and hyperprolactinemic sera from controls (n = 9) and patients with anti-PRL autoantibodies (n = 6; **, P < 0.0001; *, P = 0.022; serum PRL in patients with anti-PRL autoantibodies vs. serum PRL in controls). The ratio (percent) of the density of each spot to the sum of the densities is shown.

 

   Discussion
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
References
 
Mechanisms involved in the development of anti-PRL autoantibodies are unknown. Genetic susceptibility and environment may alter immunoreactivity in hosts as postulated in the other autoimmune disorders (25). On the other hand, some changes in PRL molecule may increase the antigenicity leading to the production of anti-PRL autoantibodies. To elucidate the underlying mechanisms, we first characterized IgG that binds to PRL molecule to examine which kind of immune response is taking place. Because IgG subclasses have different biological properties (26), determination of IgG subclasses of anti-PRL autoantibodies may help to clarify the pathogenic significance. In this study, we established an enzyme immunoassay system to detect IgG subclasses specific for anti-PRL autoantibodies. The pattern varied, but the IgG4 subclass was positive in five of six patients with anti-PRL autoantibodies. It has been reported that IgG4 response occurs in chronic antigen stimulation, that switch from IgG1 to IgG4 response is driven by repeated exposure to allergens, and that IgG4 is a predominant subclass in other autoantibodies (27, 28). Therefore, our observation that IgG4 antibodies were frequently present in greater amounts than in controls suggests that there is a chronic stimulation by PRL with possible altered antigenicity. Two patients with elevated total IgG4 subclass may be susceptible to a broad range of allergens, because IgG4 reportedly plays an important role in allergic conditions (26).

To elucidate the possible alteration of antigenicity of PRL, we first examined phosphorylation of human pituitary PRL that might be a target of autoantibody production (29). Phosphorylation is an important posttranslational modification for the regulation of structure and function of proteins in cells. Studies have shown that a significant proportion of rat (30, 31, 32), bovine (33, 34), and avian (35) PRL is phosphorylated, which accounts for much of the PRL charge heterogeneity observed. We found that human pituitary PRL was also serine-phosphorylated. Then, we examined whether the phosphorylation of pituitary PRL was altered, that is, further phosphorylation or dephosphorylation occurred in serum. We found that serine 194 remained phosphorylated, whereas serine 163 was dephosphorylated in serum. This is the first report suggesting that human pituitary PRL is phosphorylated and partially dephosphorylated when it is secreted into circulation.

Finally, we compared the profiles of serum PRL in hyperprolactinemic patients with and without anti-PRL autoantibodies. We previously reported that the molecular mass of PRL in sera from patients with anti-PRL autoantibodies was 23 kDa on SDS-PAGE, as in those with prolactinoma and in pituitary PRL standard (15). However, posttranslational modifications may produce PRL molecules with different net charges and the same molecular weight, and this prompted us to investigate PRL isoforms on 2D-electrophoresis in this study. The theoretical pI of human PRL is 6.50 according to ExPASy proteomics (http://us.expasy.org/), but it reportedly consists of more than two isoforms on isoelectric focusing (pI 5.7–7.2) in the sera (36), pituitary extract (37), culture medium of prolactinoma (38), and recombinant PRL (39). We confirmed that human PRL also consisted of three isoforms on 2D-electrophoresis. Moreover, we found, for the first time, that there was a difference in the profiles of serum PRL isoforms between control hyperprolactinemic patients and those with anti-PRL autoantibodies. PRL in sera from control hyperprolactinemic patients mainly consisted of isoform with pI 6.58, whereas that from patients with anti-PRL autoantibodies consisted of not only isoform with pI 6.58 but also more acidic ones with pIs 6.43 and 6.29, as seen in human pituitary PRL.

There are several possible causes for the appearance of different isoforms of PRL: cleavage, glycosylation, deamidation, phosphorylation, and sulfation (2). Cleavage can be ruled out to explain the present observation because this posttranslational modification decreases the molecular weight. Approximately 20% of serum PRL is reportedly glycosylated (2), and this modification can alter pIs due to the sialic acid residues on the N-linkage (40). However, apparent similar molecular sizes of the three isoforms on Western blotting make the possibility unlikely because molecular mass of glycosylated PRL (25 kDa) is different from that of nonglycosylated PRL (23 kDa). Deamidation, which involves the loss of ammonia from asparagine and glutamine residues, possibly occurs during purification procedures. To avoid deamidation during processing of the samples, we immediately stored serum samples at –80 C after the separation from blood; used electrophoresis grade urea treated with Amberlite, an ion exchange resin that removes trace ionic impurities; and performed electrophoresis at 15 C on a cooling plate. In addition, because serum samples were treated in the same way, it cannot explain the different profiles of PRL isoforms in the sera between hyperprolactinemic patients with anti-PRL autoantibodies and those without them. Because phosphorylation is known to shift the pI to the acidic side (31), concern has been raised that phosphorylation states may be different among the isoforms. However, we could not find any difference in the phosphorylation states among the three PRL isoforms in serum. Because mass spectrometric analysis has a limitation that all the trypsin-digested peptide fragments cannot be recovered, some of the other serine residues of pituitary PRL might be phosphorylated, and dephosphorylation might not occur properly when PRL is secreted into circulation in patients with anti-PRL autoantibodies, leading to an acquirement of antigenicity. This hypothesis is sustained, in part, by the observation that partial dephosphorylation occurred in the serum. It should also be recognized that the pattern of isoforms found in patients with anti-PRL autoantibodies reflects both free PRL and PRL bound to IgG, and there is a possibility that some of them are free and the others are originally autoantibody bound ones. Other possible modifications, such as sulfation, might be involved. Further studies are required to elucidate the different profiles of PRL isoforms in sera from patients with and without anti-PRL autoantibodies, which may be related to the development of anti-PRL autoantibodies.

In summary, we have demonstrated that: 1) IgG subclasses of anti-PRL autoantibodies are heterogeneous, but the predominant form is IgG4; 2) human pituitary PRL is serine-phosphorylated and partially dephosphorylated in serum; 3) human PRL consists of three isoforms with the same molecular weight and different pIs; and 4) the profiles of PRL isoforms are different between hyperprolactinemic sera with and without anti-PRL autoantibodies.


   Acknowledgments
 
We thank Dr. A. F. Parlow, a scientific director of the National Hormone and Pituitary Program, for supplying human pituitary PRL and antihuman PRL antiserum. We also thank the Kansai Medical University Laboratory Center for providing experimental instruments.


   Footnotes
 
This work was supported by grants from the Ministry of Education, Science, Sports and Culture of Japan; Japan Private School Promotion Foundation; and the Institute of Growth Science of Japan.

First Published Online February 1, 2005

Abbreviations: 2D, Two-dimensional; EGF, epidermal growth factor; HRP, horseradish peroxidase; MALDI-TOF, matrix-assisted laser desorption ionization time of flight; PEG, polyethylene glycol; pI, isoelectric point; PRL, prolactin; TFA, trifluoroacetic acid.

Received August 11, 2004.

Accepted January 24, 2005.


   References
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
 References
 

  1. Freeman ME, Kanyicska B, Lerant A, Nagy G 2000 Prolactin: structure, function, and regulation of secretion. Physiol Rev 80:1523–1631[Abstract/Free Full Text]
  2. Sinha YN 1995 Structural variants of prolactin: occurrence and physiological significance. Endocr Rev 16:354–369[Abstract/Free Full Text]
  3. Whittaker PG, Wilcox T, Lind T 1981 Maintained fertility in a patient with hyperprolactinemia due to big, big prolactin. J Clin Endocrinol Metab 53:863–866[Abstract/Free Full Text]
  4. Leslie H, Courtney CH, Bell PM, Hadden DR, McCance DR, Ellis PK, Sheridan B, Atkinson AB 2001 Laboratory and clinical experience in 55 patients with macroprolactinemia identified by a simple polyethylene glycol precipitation method. J Clin Endocrinol Metab 86:2743–2746[Abstract/Free Full Text]
  5. Kasic SV, Ramos IM, Selim A, Gunz G, Morange S, Enjalbert A, Martin PM, Jaquet P, Brue T 2002 Macroprolactinemia revisited: a study on 106 patients. J Clin Endocrinol Metab 87:581–588[Abstract/Free Full Text]
  6. Hauache OM, Rocha AJ, Maia Jr ACM, Maciel RMB, Vieira JGH 2002 Screening for macroprolactinemia and pituitary imaging studies. Clin Endocrinol (Oxf) 57:327–331[CrossRef][Medline]
  7. Kleinberg DL, Noel GL, Frantz AG 1977 Galactorrhea: a study of 235 cases, including 48 with pituitary tumors. N Engl J Med 296:589–600[Abstract]
  8. Sluijmer AV, Lappohn RE 1992 Clinical history and outcome of 59 patients with idiopathic hyperprolactinemia. Fertil Steril 58:72–77[Medline]
  9. Jackson RD, Wortsman J, Malarkey WB 1985 Characterization of a large molecular weight prolactin in women with idiopathic hyperprolactinemia and normal menses. J Clin Endocrinol Metab 61:258–264[Abstract/Free Full Text]
  10. Leite V, Cosby H, Sobrinho LG, Fresnoza A, Sanios MA, Friesen HG 1992 Characterization of big, big prolactin in patients with hyperprolactinemia. Clin Endocrinol (Oxf) 37:365–372[Medline]
  11. Carlson HE, Markoff E, Lee DW 1992 On the nature of serum prolactin in two patients with macroprolactinemia. Fertil Steril 58:78–87[Medline]
  12. Hattori N, Ishihara T, Ikekubo K, Moridera K, Hino M, Kurahachi H 1992 Autoantibody to human prolactin in patients with idiopathic hyperprolactinemia. J Clin Endocrinol Metab 75:1226–1229[Abstract]
  13. Hattori N, Ikekubo K, Ishihara T, Moridera K, Hino M, Kurahachi H 1994 Correlation of the antibody titers with serum prolactin levels and their clinical course in patients with anti-prolactin autoantibody. Eur J Endocrinol 130:438–445[Abstract/Free Full Text]
  14. Hattori N 1996 The frequency of macroprolactinemia in pregnant women and the heterogeneity of its etiologies. J Clin Endocrinol Metab 81:586–590[Abstract]
  15. Hattori N, Inagaki C 1997 Anti-prolactin (PRL) autoantibodies cause asymptomatic hyperprolactinemia: bioassay and clearance studies of PRL-immunoglobulin G complex. J Clin Endocrinol Metab 82:3107–3110[Abstract/Free Full Text]
  16. Olukoga AO, Kane J 1995 Anti-prolactin autoantibodies and hyperprolactinemia. Eur J Endocrinol 133:463–464[Abstract/Free Full Text]
  17. Cavaco B, Leite V, Santos MA, Arranhado E, Sobrinho LG 1995 Some forms of big big prolactin behave as a complex of monomeric prolactin with an immunoglobulin G in patients with macroprolactinemia or prolactinoma. J Clin Endocrinol Metab 80:2342–2346[Abstract]
  18. Blanco-Favela F, Chavez-Rueda K, Leanos-Miranda A 2001 Analysis of anti-prolactin autoantibodies in systemic lupus erythematosus. Lupus 10:757–761[Abstract/Free Full Text]
  19. Pascoe-Lira D, Duran-Reyes G, Contreras-Hernandez I, Manuel-Apolinar L, Blanco-Favela F, Leanos-Miranda A 2001 Frequency of macroprolactinemia due to autoantibodies against prolactin in pregnant women. J Clin Endocrinol Metab 86:924–929[Abstract/Free Full Text]
  20. Schepper JD, Schiettecatte J, VelkeniersB, Blumenfeld Z, Shteinberg M, Devroey P, Anckaert E, Smitz J, Verdood P, Hooghe R, Hooghe-Peters E 2003 Clinical and biological characterization of macroprolactinemia with and without prolactin-IgG complexes. Eur J Endocrinol 149:201–207[Abstract]
  21. Sakata S, Nakamura S, Miura K 1985 Autoantibodies against thyroid hormones or iodothyronine. Ann Intern Med 103:579–589
  22. Folling I, Norman N 1972 Hyperglycemia, hypoglycemic attacks, and production of anti-insulin antibodies without previous known immunization: immunological and functional studies in a patient. Diabetes 21:814–826[Medline]
  23. Jiang B, Hattori N, Liu B, Nakayama Y, Kitagawa K, Sumita K, Inagaki C 2004 Expression and roles of Cl– channel ClC-5 in cell cycles of myeloid cells. Biochem Biophys Res Commun 317:192–197[CrossRef][Medline]
  24. Rubin RA, Earp HS 1983 Solubilization of EGF receptor with triton X-100 alters stimulation of tyrosine residue phosphorylation by EGF and dimethyl sulfoxide. J Biol Chem 258:5177–5182[Abstract/Free Full Text]
  25. Marrack P, Kappler J, Kotzin BL 2001 Autoimmune disease: why and where it occurs. Nat Med 7:899–905[CrossRef][Medline]
  26. Maran R, Dueymes M, Le Corre R, Renaudineau Y, Shoenfeld Y, Youinou P 1997 IgG subclasses of human autoantibodies. Ann Med Interne (Paris) 148:29–38[Medline]
  27. Calvanico NJ, Swartz SJ, Diaz LA 1993 Affinity immunoblotting studies on the restriction of autoantibodies from endemic pemphigus foliaceus patients. J Autoimmun 6:145–157[CrossRef][Medline]
  28. Ohara M, Hibi T, Watanabe N, Kobayashi K, Takaishi H, Hayashi A, Hosoda Y, Toda K, Iwao Y, Watanabe M, Aiso S, Tsuchiya M, Ishii H 1995 Immunoglobulin G subclass distribution of human anticolon antibodies in ulcerative colitis. J Gastroenterol Hepatol 10:158–164[Medline]
  29. Utz PJ, Hottelet M, Schur PH, Anderson P 1997 Proteins phosphorylated during stress-induced apoptosis are common targets for autoantibody production in patients with systemic lupus erythematosus. J Exp Med 185:843–854[Abstract/Free Full Text]
  30. Oetting WS, Tuazon PT, Traugh JA, Walker AM 1986 Phosphorylation of prolactin. J Biol Chem 261:1649–1652
  31. Wang YF, Walker AM 1993 Dephosphorylation of standard prolactin produces a more biologically active molecule: evidence for antagonism between nonphosphorylated and phosphorylated prolactin in the stimulation of Nb2 cell proliferation. Endocrinology 133:2156–2160[Abstract/Free Full Text]
  32. Wang YF, Liu JW, Mamidi M, Walker AM 1996 Identification of the major site of rat prolactin phosphorylation as serine 177. J Biol Chem 271:2462–2469[Abstract/Free Full Text]
  33. Kim BG, Brooks CL 1993 Isolation and characterization of phosphorylated bovine prolactin. Biochem J 296:41–47
  34. Brooks CL, Saiduddin S 1998 Phosphorylation of bovine prolactin eliminates luteotropic activity in the rat. Life Sci 63:1281–1287[CrossRef][Medline]
  35. Aramburo C, Montiel JL, Proudman JA, Berghman LR, Scanes CG 1992 Phosphorylation of prolactin and growth hormone. J Mol Endocrinol 8:183–191[Abstract/Free Full Text]
  36. Guido R, Valenti S, Foppiani L, De Martini D, Cossu M, Giusti M 1997 Prolactin decrease and shift to a normal-like isoform profile during treatment with quinagolide in a patient affected by an invasive prolactinoma. J Endocrinol Invest 20:289–293[Medline]
  37. Hummel BCW, Brown GM, Hwang P, Friesen HG 1975 Human and monkey prolactin and growth hormone: separation of polymorphic forms by isoelectric focusing. Endocrinology 97:855–867[Abstract/Free Full Text]
  38. Hoffmann T, Gunz G, Brue T, Jaquet P, Ronin C 1992 Prolactin isoforms secreted by human prolactinomas. Horm Res 38:164–170[Medline]
  39. Price AE, Logvinenko KB, Higgins EA, Cole ES, Richards SM 1995 Studies on the microheterogeneity and in vitro activity of glycosylated and nonglycosylated recombinant human prolactin separated using a novel purification process. Endocrinology 136:4827–4833[Abstract]
  40. Shimomura K, Bremel RD 1988 Characterization of bovine placental lactogen as a glycoprotein with N-linked and O-linked carbohydrate side chains. Mol Endocrinol 2:845–853[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 EndocrinologyHome page
N. Hattori, Y. Nakayama, K. Kitagawa, T. Li, and C. Inagaki
Development of Anti-PRL (Prolactin) Autoantibodies by Homologous PRL in Rats: A Model for Macroprolactinemia
Endocrinology, May 1, 2007; 148(5): 2465 - 2470.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hattori, N.
Right arrow Articles by Inagaki, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hattori, N.
Right arrow Articles by Inagaki, C.
Right arrowPubmed/NCBI databases
*Compound via MeSH
*Substance via MeSH
Hazardous Substances DB
*L-SERINE
Related Collections
Right arrow Neuroendocrinology and Pituitary

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals