| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Medical Department II, Laboratory of Endocrine Research, Klinikum Grosshadern, University of Munich, 81377 Munich, Germany
Address all correspondence and requests for reprints to: PD Dr. Med. Matthias M. Weber, Medizinische Klinik II, Klinikum Grosshadern, Marchioninistrasse 15, 81377 Munich, Germany.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The principal method by which IGFBPs are detected is the Western ligand blot (WLB) analysis first described by Hossenlopp et al. (8). Using this technique, IGFBPs are separated according to molecular mass by a one-dimensional SDS-PAGE under nonreducing conditions, transfered to a nitrocellulose membrane, and incubated with 125I-labeled IGF-I or IGF-II. The specific IGFBP bands are then visualized by autoradiography and identified according to their apparent molecular masses. However, since IGFBP-1, glycosylated IGFBP-4, IGFBP-5, and IGFBP-6 migrate in close proximity at approximately 30 kDa, the identification of each individual IGFBP in complex biological fluids is very difficult. In this study we describe for the first time the identification and characterization of all six high affinity IGFBPs in human biological fluids by two-dimensional WLB (2D-WLB), a powerful method combining modified two-dimensional immobilized pH gradient (IPG) gel electrophoresis and WLB analysis.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recombinant human IGF-I and IGF-II were purchased from Boehringer Mannheim (Mannheim, Germany), and 125iodotyrosyl-IGF-I and -IGF-II (human recombinant; SA, 2000 Ci/mmol) were obtained from Amersham Buchler GmbH & CoKG (Braunschweig, Germany). Polyclonal rabbit antiserum against human IGFBP-1 was a gift from Dr. Hans Bohn (Behringwerke, Marburg, Germany), antibodies against human IGFBP-4 and -6 were obtained from Austral Biological (San Ramon, CA), polyclonal antibody against IGFBP-3 was purchased from Diagnostic Systems Laboratories, Inc. (Webster, TX), polyclonal anti IGFBP-5 was purchased from Upstate Biotechnology (Lake Placid, NY), and polyclonal antibody against IGFBP-2 was a gift from Dr. Elmlinger (Tubingen, Germany).
Samples
Serum samples were obtained from healthy adults (age, 20–50 yr). Immediately after the blood was collected, the serum fraction was stored individually or pooled (using an equal volume of five males and five females) at -80 C. Human cerebrospinal fluid was obtained from otherwise healthy adults who underwent a diagnostic cerebrospinal puncture due to headache. In all patients biochemical analysis of the cerebrospinal fluid showed no signs of inflammation or any other pathological finding. Human amniotic fluid was obtained from pregnant women in the first trimester who underwent diagnostic amniocentesis. The genetic analysis in all samples was normal. The sample collection was performed according to the guidelines of the local ethical committee, and all patients signed consent forms.
2D-WLB analysis of IGFBPs
Aliquots of serum, amniotic fluid, or cerebrospinal fluid were subjected to a modified two-dimensional PAGE under nonreducing conditions according to the method described by Görg (9). In contrast to conventional two-dimensional gel electrophoresis, no reducing agent such as dithiothreitol was used, so as to retain the binding capacity of the separated IGFBPs. For isoelectric focusing in the first dimension, immobilized pH gradient gels (IPG gels) with a pH range of 4–10 or 4–7 were prepared as previously described (10). Additional, precast gels with a pH range of 3–10 (0.5 x 3 x 110 mm, Immobiline DryStrip, Pharmacia, Freiburg, Germany) were used. Before the isoelectric focusing, the dry gels were reswollen for 16 h in IPG buffer (8 mol/L urea and 0.2% Pharmalyte 3–10; Pharmacia). Ten microliters of the samples were then mixed with an equal volume of IPG buffer, incubated for 1 h at room temperature, and loaded on a rehydrated IPG gelstrip. Isoelectric focusing was carried out at 20 C for 0.5 h with 200 V, followed by 6 h with 3000 V in a cooled horizontal electrophoresis unit (Multiphor II, Pharmacia), and the gelstrips were stored at -20 C. For separation in the second dimension (SDS-PAGE), the IPG strips were equilibrated for 15 min in 50 mmol/L Tris-HCl (pH 8.8), 6 mol/L urea, 30% glycerol, 2% SDS and a trace of bromophenol blue and placed on a horizontal SDS-PAGE gradient gel (10–15%). After protein transfer (45 min at 200 V, 10 C), the IPG strips were removed, and nonreducing SDS-PAGE electrophoresis was continued at 800 V. After two-dimensional gel electrophoresis, the IGFBPs were identified by WLB according to the method of Hossenlopp et al. (8), with the following modifications. Proteins were transfered to polyvinylidene difluoride (PVDF) membranes (Immobilone, Millipore Corp., Munich, Germany) using the semidry electroblotting technique described by Khyse-Andersen (11). After the transfer, the membranes were washed and blocked in Tris buffer (1.5% Tween-20, 0.01 mol/L Tris-HCl, and 0.15 mol/L NaCl, pH 7.4), incubated in the presence of [125I]IGF-I or [125I]IGF-II (50,000 cpm/mL) in the same buffer for 16 h in a shaking water bath at room temperature, washed again, air-dried, and subjected to autoradiography (BioMax MS, Eastman Kodak Co., Rochester, NY) for 16–36 h at -70 C with amplifying screens. The specificity of the spots was verified by incubation of duplicate gels with an excess of cold IGF-II or insulin (50 nmol/L). After scanning the autoradiographs, the specific spots were quantified, and their isoelectric points (pI) and molecular masses were calculated using a personal computer-based program (Gelreader 2.05, Macintosh, University of Illinois, Chicago, IL).
Deglycosylation of IGFBPs
For N-deglycosylation of IGFBPs, 10 µL of the sample were incubated with 3 µL endoglycosidase F-N-glycosidase F (Boehringer Mannheim) for 3 h at 37 C before 2D-WLB.
Immunoprecipitation of IGFBPs
Twenty milligrams of protein A-Sepharose (Pharmacia Biotech, Uppsala, Sweden) were reconstituted in 1 mL phosphate-buffered saline (PBS), washed three times with PBS, and preincubated together with 2–20 µL specific IGFBP antiserum for 1 h at room temperature. Then, 50 µL of the sample were added together with protease inhibitors (Complete, Boehringer Mannheim) and incubated at 4 C overnight. The complexed binding proteins were precipitated (5 min at 3000 x g), washed twice, and resuspended in 50 µL IPG buffer. After centrifugation at 3000 x g for 5 min, the supernatant was aspirated and subjected to 2D-WLB as described.
Immunoblotting
Samples were subjected to nonreducing two-dimensional gel electrophoresis and transferred to PVDF membranes as described above. The membrane was blocked with 5% BSA and 0.1% Tween-20 in PBS for 12 h before incubation with the indicated IGFBP antiserum in PBS containing 1% BSA and 0.1% Tween-20 for 3 h. The PVDF membranes were washed twice with PBS containing 0.1% Tween-20 and incubated with a peroxidase-labeled second antibody for another 2 h. The membranes were then washed extensively for 1 h with PBS containing 0.1% Tween-20, and the specific immune complexes were visualized using a chemiluminescence system (ECL, Amersham, Braunschweig, Germany).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
shows the results from a
pooled serum sample subjected to conventional one-dimensional (A) and
two-dimensional (B) WLB. Depending on the exposure time, a variety
of spots were evident on the 2D-WLB from adult human serum. These spots
were specific, as the labeled ligand could be effectively displaced by
unlabeled IGF-I or IGF-II (1 µg/mL) but not by equimolar
concentrations of insulin (data not shown). All serum blots shown are
representative of the results obtained from at least 20 individual and
5 pooled serum samples from healthy adults assayed by 2D-WLB. The
relative positions of all IGFBP variants were highly conserved between
different serum samples, and the results were highly reproducible when
the same sample was analyzed repeatedly (interassay coefficient of
variation of <5% for pI and molecular mass).
|
Most prominent were two broad bands of specific spots on the
2D-WLB of human serum with apparent molecular masses of 41 and 45 kDa
and pI ranging from 4.8–8.2 (Fig. 1
). When the serum sample was
subjected to deglycosylation before 2D-WLB, these bands were reduced to
a single band of approximately 36 kDa (pI 6.0–9.0). However, the
deglycosylation was incomplete, as two faint bands at 41 and 45 kDa
remained visible (Fig. 2). According to
their molecular masses and glycosylation status, these spots were
compatible with different glycosylation variants of IGFBP-3. The
identity of IGFBP-3 was confirmed by immunoprecipitation of the serum
sample with a specific IGFBP-3 antibody before 2D-WLB (Fig. 3) as well as by IGFBP-3 immunoblotting
(Fig. 4). Although IGFBP-3
immunoprecipitation of the serum sample showed two series of 41- and
45-kDa spots, which corresponded to the two 41/45-kDa bands in native
2D-WLB, the IGFBP-3 immunoblot additionally revealed a variety of spots
in the range of 29–31 kDa (pI 4.3–5.5; Fig. 4B). These immunoreactive
spots lacked binding capacity for labeled IGF-ligands and most likely
represent proteolytic fragments of IGFBP-3. In addition, smaller
IGFBP-3 fragments with molecular masses of 18 kDa (pI 4.3–5) and 14
kDa (pI 5 and 6.2) were found when concentrated serum samples were
subjected to two-dimensional immunoblotting (data not shown).
|
|
|
At a molecular mass of 34 kDa, four major spots with pI values of
6.2–7.1 were visible in the 2D-WLB of human serum (Fig. 1) and
cerebrospinal fluid (Fig. 5). These spots
could not be reduced by enzymatical N-deglycosylation (Fig. 2) and were confirmed to represent different isoforms of IGFBP-2 by
immunoprecipitation (Fig. 3).
|
In 2D-WLB of human serum, at least three prominent spots at a
molecular mass of 24 kDa were identified with pI values of 6.3, 6.5,
and 6.8 (Fig. 1). Furthermore, two prominent spots with a molecular
mass of 29 kDa and pI values of 6.1 and 6.3 could be completely
deglycosylated by treatment with endoglycosidase F before 2D-WLB,
thereby increasing the abundance of the 24-kDa spots (Fig. 2).
Therefore, these spots are compatible with different isoforms of
glycosylated and deglycosylated IGFBP-4, as confirmed by IGFBP-4
immunoprecipitation (Fig. 3).
IGFBP-5
The presence of IGFBP-5 as a series of doublet spots with a
molecular mass of 30/31 kDa and a pI between 6.4–8 could be shown
after immunoprecipitation of pooled human serum (Fig. 3). Although the
IGFBP-5 spots were only faintly visible with the short exposure of the
native serum 2D-WLB (Fig. 1), these spots could be clearly demarcated
after long exposures of the autoradiographs (blot not shown). The fact
that the IGFBP-5 spots were not affected by enzymatical
N-deglycosylation (Fig. 2) might be due to
O-glycosylation of IGFBP-5 or its low levels in human
serum.
IGFBP-1
When human amniotic fluid or conditioned medium from HepG2 cells,
which are known to be rich sources of IGFBP-1 (12), were subjected to
2D-WLB, a series of specific spots at 30 kDa with pI of 4.0, 4.2, 4.4,
4.6, and 4.8 were found (Fig. 6). In
amniotic fluid, which is known to contain mainly dephosphorylated
IGFBP-1, the most predominant IGFBP-1 isoform exhibited an isoelectric
point of 4.8, whereas in medium from HepG2 cells, which produce mainly
phosphorylated IGFBP-1 variants, the more acidic species of IGFBP-1 was
the most abundant. The authenticity of these spots as at least five
different different isoforms of IGFBP-1 (most likely different
phosphorylation variants) was confirmed by IGFBP-1
immunoprecipitation (Fig. 3) and immunoblotting (Fig. 6). Due to the
low abundance of IGFBP-1 in human serum, only a faint spot, which
presumably represents IGFBP-1, could be observed in the 2D-WLB of
pooled serum (Fig. 1). However, in individual serum samples with
elevated IGFBP-1 levels, these spots were clearly visible (blots not
shown).
|
For the identification of IGFBP-6, human cerebrospinal fluid,
which is known to contain large amounts of IGFBP-2 and IGFBP-6 (13),
was analyzed by 2D-WLB (Fig. 5). With IGF-I as radioligand for the
autoradiography, mainly the characteristic spots of IGFBP-2 and, to a
lesser extent, of IGFBP-4 were found (Fig. 5A). In contrast, when the
same blot was incubated with labeled IGF-II, a series of strong
additional spots with a molecular mass around 30 kDa and pI values
between 4.8–5.8 were found (Fig. 5B). In accordance with the high
affinity of IGFBP-6 to IGF-II, the identities of these spots as
variants of IGFBP-6 were verified by immunoblotting the same blot with
a specific IGFBP-6 antibody (Fig. 5C). In analogy to the finding in
cerebrospinal fluid, IGFBP-6 could be visualized in human serum by
2D-WLB as faint spots with pI values from 4.8–5.8 and molecular masses
between 29–31 kDa (Fig. 1), as confirmed by IGFBP-6
immunoprecipitation (Fig. 3) of human serum.
According to the data presented above, we have been able to identify
all six human high affinity binding proteins in a pooled human serum
sample. Based on the ligand blot shown in Fig. 1, the relative
positions of the isoforms of all six IGFBPs on a 2D-WLB in human serum
are shown schematically in Fig. 7.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Two-dimensional WLB analysis revealed a molecular heterogeneity of IGFBPs that was more complex than had previously been suspected. In adult human serum, more than 45 specific spots of IGFBPs were detected by 2D-WLB. The IGFBP pattern after 2D-WLB was highly conserved among 30 different individual serum samples and 5 pooled serum samples from healthy adults. These multiple variants of IGFBPs may reflect differences in the degree of phosphorylation, variances in their carbohydrate side-chains, as well as minor amino acid changes. Most prominent in human serum are two broad bands of IGFBP-3 spots that can be reduced by enzymatical N-deglycosylation to a series of spots at 36 kDa with a more basic pI. In addition, a series of smaller spots is found by IGFBP-3 immunoblotting, which presumably represents proteolytic fragments of IGFBP-3 and exhibits a more acidic pI than the 41/45-kDa variants of IGFBP-3. Furthermore, at least 5 different isoforms of IGFBP-2 and IGFBP-4 could be readily detected by 2D-WLB of normal adult human serum. When the 2D-WLB of human serum was exposed for a longer time, additional spots, representing IGFBP-1, IGFBP-5, and IGFBP-6, were observed. Although these spots exhibit a similar molecular mass between 29–33 kDa, the different binding proteins can be clearly separated from each other according to their characteristic range of isoelectric points. Due to the preferred affinity for IGF-II, IGFBP-6 could only be detected when IGF-II was used as a radioligand. Although we were not able to precipitate IGFBP-1 from adult human serum, we identified five different isoforms of IGFBP-1 (30 kDa; pI 4.0–4.8) in human amniotic fluid and in conditioned medium from the human hepatoma cell line HepG2 by immunoprecipitation and immunoblotting. In amniotic fluid, which is known to contain mainly dephosphorylated IGFBP-1, the most predominant IGFBP-1 isoform exhibited an isoelectric point of 4.8, whereas in medium from HepG2 cells, which produce mainly phosphorylated IGFBP-1 variants, the more acidic species of IGFBP-1 were the most abundant. This is in accordance to the findings of different phosphorylation variants after one-dimensional, non-SDS ligand blot analysis of IGFBP-1 (12) and supports the assumption that the IGFBP-1 variants found after 2D-WLB represent different phosphorylation variants of IGFBP-1. Due to the low abundance of IGFBP-1 in human serum, only a faint spot, which presumably represents IGFBP-1, could be observed in the 2D-WLB of pooled sera. However, in individual serum samples with elevated IGFBP-1 levels, these spots were clearly visible (data not shown). When human cerebrospinal fluid was analyzed by 2D-WLB, the same pI isoforms of IGFBP-2, IGFBP-4, and IGFBP-6 were found as those in adult serum, indicating that similar IGFBP variants are present in different biological fluids.
The fact that all six human IGFBPs can be separately identified on a single blot after 2D-WLB makes this method a potent tool for the analysis of IGFBPs in complex biological fluids. Only a few other studies report on the characterization of IGFBPs in complex biological fluids by conventional 2D-WLB in rat and baboon sera (15, 16) and in conditioned medium of a rat neuronal cell line (17). Although in baboon serum only IGFBP-3 was detected as a faint doublet by 2D-WLB (16), rat serum and conditioned medium of rat B104 neuronal cells exhibited a similar pattern for IGFBP-2, IGFBP-3, and IGFBP-4 as that in human serum (15, 17). The fact that the different isoforms of serum IGFBPs are conserved among different species supports the hypothesis that the posttranslational modifications of IGFBPs are tightly regulated and play an important role in the modulation of IGF action by IGFBPs. Furthermore, our results add further proof to the assumption of Chan and Nicoll, who tentatively identified an acidic set of 29-kDa spots in the 2D-WLB of rat serum as IGFBP-1, although they could not prove this by immunoblotting (15).
Much interest has recently been focused on the characterization of posttranslational modifications of IGFBPs. All six forms of IGFBPs undergo some form of posttranslational modification, such as phosphorylation, glycosylation, or proteolytic changes (18). The increasing interest in studying posttranslational modified IGFBP variants stems from the findings that some of these IGFBP isoforms show different biological effects, as has been shown for different phosphorylation variants of IGFBP-1 as well as for proteolytical modifications of IGFBPs in pregnancy serum and serum from children with chronic renal failure (2, 3, 4, 19). Therefore, IGFBP isoforms might represent another level and mechanism of regulating the biological actions of IGFBPs, and two-dimensional WLB analysis will be a valuable tool in detecting and evaluating the biological significance of IGFBP variants (18, 20). In conclusion, 2D-WLB is a potent method to identify different isoforms of all six human high affinity IGFBPs in complex biological fluids.
|
Footnotes |
|---|
Received November 11, 1998.
Revised December 30, 1998.
Accepted February 9, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Einspanier, D. Muller, J. Lubberstedt, O. Bartsch, A. Jurdzinski, K. Fuhrmann, and R. Ivell Characterization of relaxin binding in the uterus of the marmoset monkey Mol. Hum. Reprod., October 1, 2001; 7(10): 963 - 970. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Smith, S. K. Bains, J. C. Betts, E. H. S. Choy, and E. D. Zanders Use of Two-Dimensional Gel Electrophoresis To Measure Changes in Synovial Fluid Proteins from Patients with Rheumatoid Arthritis Treated with Antibody to CD4 Clin. Vaccine Immunol., January 1, 2001; 8(1): 105 - 111. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
-lactalbumin (21 kDa)
are indicated on the left. The blot is representative of
at least 20 individual and 5 pooled serum samples from adult healthy
individuals.
