Endocrinology Vol. 138, No. 4 1649-1657
Copyright © 1997 by The Endocrine Society
Insulin-Like Growth Factor-I (IGF-I) Concentration in 150-Kilodalton Complexes Containing Human IGF-Binding Protein-3 (hIGFBP-3) after Intravenous Injection of Adult Rats with hIGFBP-3
C. Young Lee1,
Hai-Bin Wu,
Dae-Shik Suh and
Matthew M. Rechler
Growth and Development Section, Molecular and Cellular
Endocrinology Branch, National Institute of Diabetes and Digestive and
Kidney Diseases, National Institutes of Health, Bethesda, Maryland
20892
Address all correspondence and requests for reprints to: Dr. M. M. Rechler, National Institutes of Health, Building 10, Room 8D-14, Bethesda, Maryland 20892. E-mail: mrechler{at}helix.nih.gov
 |
Abstract
|
|---|
After iv injection into adult rats, human insulin-like growth
factor-binding protein-3 (hIGFBP-3) forms 150-kDa complexes with excess
endogenous rat acid-labile subunit (ALS) within 2 min (Lewitt et
al., 1993, Endocrinology 133:1797). Because their
previous in vitro studies indicated that hIGFBP-3 only
bound to ALS in the presence of IGF-I, and because little free IGF-I is
present in plasma, the authors postulated that IGF-I had been mobilized
to the circulation to saturate the 150-kDa hIGFBP-3 complexes. We
examined this hypothesis by determining whether the hIGFBP-3 that
appears in 150-kDa complexes 2 min after iv injection is accompanied by
an increase in IGF-I. Within 2 min, some of the injected hIGFBP-3
(
30% as much as endogenous intact rat IGFBP-3) is present in
complexes that are cleared slowly from the circulation and presumed to
be 150-kDa complexes. Gel filtration and immunoprecipitation studies
performed on blood collected 2 min after injection confirmed that the
injected hIGFBP-3 (46–82% as much as rat IGFBP-3) was associated with
ALS in 150-kDa complexes. The formation of 150-kDa complexes containing
hIGFBP-3 was not accompanied by a corresponding change in the IGF-I
content (determined by RIA) of whole serum or 150-kDa serum fractions,
suggesting that the hIGFBP-3 had rapidly associated with rALS in
vivo without mobilizing IGF-I. Surprisingly, however, hIGFBP-3
was cleared much more rapidly from 150-kDa complexes formed after
injection of hIGFBP-3 than after injection of hIGFBP-3:IGF-I complexes,
suggesting that the complexes observed after hIGFBP-3 injection might
not have formed in vivo. In fact, 150-kDa complexes
formed to a similar extent when hIGFBP-3 was added ex
vivo to blood collected from rats that had not received
hIGFBP-3. We conclude that hIGFBP-3 can rapidly associate with rALS to
form 150-kDa complexes in vivo without the mobilization
of IGF-I. Because 150-kDa complexes also are formed ex
vivo, however, we are unable to resolve whether the complexes
that formed in vivo formed as binary or ternary
complexes.
|
Introduction
|
|---|
INSULIN-LIKE growth factors (IGFs) occur in plasma
and tissue fluids complexed to IGF-binding proteins
(IGFBPs) (1, 2, 3). The IGF:IGFBP complexes in tissues are about 40
kDa in size, whereas the predominant complexes in plasma are 150 kDa,
and are composed of 40- to 44-kDa IGFBP-3 and an approximately 85-kDa
acid-labile subunit (ALS). The 150-kDa complexes sequester the IGFs in
plasma (4), providing a reservoir from which they may be released by
proteolysis of IGFBP-3 (5) while protecting the host from the potential
hypoglycemic effects of free IGF and IGF present in 40-kDa IGFBP
complexes that can gain access to tissues (6).
Based on in vitro studies, Baxter et al.
(7, 8) proposed that the 150-kDa complex is formed in two steps:
binding of IGF to IGFBP-3, followed by association of this binary
complex with the ALS. To study the formation of the 150-kDa complex
in vivo, Lewitt et al. (9) injected
purified human (h) IGFBP-3 iv into adult rats and observed that over
50% of the injected hIGFBP-3 was found in the 150-kDa fraction when
serum collected 2 min after injection was examined by size-exclusion
chromatography followed by RIA. Interpreting these results according to
their in vitro model, they postulated that a large pool
of free IGF-I (700 ng/ml, comparable to the concentration of IGF-I
in plasma before injection) must have been mobilized to saturate the
150-kDa hIGFBP-3 complexes. It was unclear, however, why the host did
not become hypoglycemic in the face of such a large pool of free IGF-I
(6).
Recently, we (10) and Barreca et al. (11) reported that
hIGFBP-3 and rat ALS (rALS) can form binary complexes in
vitro in the absence of IGF. As Lewitt et al.
(9) had not demonstrated in their in vivo study that the
150-kDa hIGFBP-3 complexes that they observed actually were saturated
with IGF-I, we considered an alternative explanation for their results,
namely that the 150-kDa complexes had formed in the absence of
IGF-I by direct association of injected hIGFBP-3 with rALS. The
present study was undertaken to determine whether hIGFBP-3 can form
binary complexes with rALS in vivo or whether the
150-kDa hIGFBP-3 complexes that formed in vivo contain
equimolar amounts of IGF as required by the hypothesis that free IGFs
synthesized in the tissues are rapidly mobilized into the circulation
(9).
|
Materials and Methods
|
|---|
Materials
Recombinant nonglycosylated hIGFBP-3 expressed in
Escherichia coli and rabbit antiserum to hIGFBP-3 were
generous gifts from A. Sommer (Celtrix, Santa Clara, CA). Rabbit
antiserum 5025 to rat ALS was raised in our laboratory (5). Recombinant
human IGF-I used for iv injection was a gift from A. Skottner
(Pharmacia-Kabi Peptide Hormones, Stockholm, Sweden). Recombinant human
IGF-I used as a RIA standard was purchased from Biosource International
(Camarillo, CA). Rabbit antiserum UB3–189 to human somatomedin-C/IGF-I
was obtained through the National Hormone and Pituitary Program
(NIDDK). Sodium pentobarbital was purchased from Richmond Veterinary
Supply Co. (Richmond, VA), and Jelico catheters (24 gauge; 19 mm) were
obtained from Kritikon (Tampa, FL). [125I]IGF-I,
[125I]IGF-II, [125I]protein-A, and
14C-methylated protein markers were purchased from Amersham
(Arlington Heights, IL). Immunoprecipitin, a 10% suspension of
protein-A-coated Staphylococcus aureus, was purchased from
Life Technologies (Gaithersburg, MD).
Intravenous injection of hIGFBP-3 into adult rats
Adult rats were injected iv with hIGFBP-3 or hIGFBP-3:IGF-I
complexes according to protocols that had been approved by the NIDDK
animal use committee. On the day of injection, 6-week-old male
Sprague-Dawley rats (164 ± 8 g; Taconic Farms, Germantown,
NY) were anesthetized by ip injection of sodium pentobarbital [65 mg
(1 ml of a 6.5% solution)/kg]. Animals that were killed 240 min after
injection received a second injection of sodium pentobarbital at 120
min. The femoral vein was surgically exposed on one side of the groin,
and a 24-gauge catheter was inserted (12, 13). After collecting 0.3 ml
blood for a baseline (time zero) sample, hIGFBP-3 (0.5 mg/kg BW),
hIGFBP-3:IGF-I, or vehicle (50 mM sodium phosphate, pH 6.5;
150 mM NaCl; and 0.1% BSA) that had been prewarmed to room
temperature was injected through the catheter followed by 0.1 ml
heparin solution (20 U/ml saline). [hIGFBP-3:IGF-I complexes were
prepared by incubating hIGFBP-3 (0.5 mg/ml) and an equimolar amount
of IGF-I in vehicle overnight at 4 C.] Four rats were included in each
treatment group; animals 1–4 were injected with vehicle, animals 5–8
were injected with hIGFBP-3, and animals 9–12 were injected with
hIGFBP-3:IGF-I complexes. Blood (0.3 ml) was collected 2, 8, 30, and
120 min after injection, followed by injection of 0.1 ml heparin
solution. An additional sample was collected at 240 min in the
hIGFBP-3:IGF-I group. All animals were killed at 120 or 240 min by
iv injection of 65 mg sodium pentobarbital. Blood samples were clotted
on ice (4–5 h) and centrifuged (16,000 x g, 10 min, 4
C). The serum was harvested and stored at -70 C. In some experiments,
as noted in the text, blood was clotted on ice for 20 min and
centrifuged, and the serum was immediately gel-filtered.
Fractionation of serum by Sephadex G-100 gel filtration at neutral
pH
To determine the distribution of IGFBPs in serum, 50 or 100 µl
serum were fractionated on a Sephadex G-100 column (0.9 x 57 cm)
at 4 C under neutral conditions (50 mM sodium phosphate, pH
7.4; 0.15 M NaCl; and 0.02% sodium azide) as previously
described (14); 0.5-ml fractions were collected. Aliquots (20 or 40
µl) of odd-numbered fractions in the 150- and 40-kDa regions
(30–61% bed volume) were examined by ligand blotting (see below).
Immunoprecipitation with antiserum to hIGFBP-3 or rALS
Complexes containing hIGFBP-3 or rALS were immunoprecipitated
from rat serum as previously described (15). In brief, 10 µl serum
(from which endogenous IgG had been removed using Immunoprecipitin)
were incubated with 5 µl rabbit antiserum to hIGFBP-3 overnight at 4
C in 0.2 ml immunoprecipitation buffer (0.1 M Tris-Cl, pH
7.4; 0.15 M NaCl; 5 mM EDTA; 1% sodium
deoxycholate; and 1% Triton X-100). The incubated sample was mixed
with 0.2 ml 10% Immunoprecipitin for 1 h at 4 C on a rotating
shaker, and centrifuged for 5 min at 16,000 x g. The
pellet was washed twice with 0.6 ml buffer, resuspended in 40 µl
1.5 x sample buffer, boiled for 5 min, and centrifuged, after
which the supernatant was subjected to 8.5% SDS-PAGE under reducing
conditions (5% 2-mercaptoethanol) before immunoblotting with antiserum
to rALS (see below).
In the reciprocal experiments, serum was immunoprecipitated with 5 µl
rabbit antiserum to rALS (5). The resuspended immunoprecipitates were
examined by 12.5% SDS-PAGE under nonreducing conditions before ligand
blotting (see below).
Ligand blotting of IGFBPs and immunoblotting of rALS
The IGFBPs present in different samples were determined by
ligand blotting as previously described (15). Samples were fractionated
by SDS-PAGE under nonreducing conditions, the proteins were transferred
electrophoretically to nitrocellulose membranes, and IGFBPs were
identified by incubation with [125I]IGF-II followed by
autoradiography.
Endogenous rALS bound to injected hIGFBP-3 was determined by
immunoprecipitation of the serum with antibodies to hIGFBP-3 as
described above, followed by immunoblotting using rabbit antiserum 5025
to rALS and [125I]protein A. The general immunoblotting
procedure has been described previously (15). After electrotransfer to
the nitrocellulose membrane, blocking with TS-T buffer [10
mM Tris-Cl, 150 mM NaCl, and 0.1% (vol/vol)
Tween-20, pH 7.4] containing 1% BSA, and washing, the membrane was
incubated (1 h, 24 C) with rabbit antiserum to rALS (1:500). After
extensive washing, the membrane was incubated with
[125I]protein A (Amersham, Arlington Heights, IL; 1
x 106 cpm/ml) in TS-T buffer containing 1% BSA for 1
h at 24 C, washed six times (15 min each), and autoradiographed.
Autoradiographs were analyzed quantitatively using the NIH Image
Program for densitometric analysis of one-dimensional gels (Division of
Computer Research and Technology, NIH, Bethesda, MD). The abundance of
injected hIGFBP-3 (32 kDa) was determined directly in autoradiographs
in which it was well resolved from 30-kDa IGFBPs and normalized to the
abundance of intact, endogenous, 40- to 44-kDa rIGFBP-3. In
autoradiographs in which the 32- and 30-kDa bands were not well
resolved, the abundance of 32-kDa hIGFBP-3 was calculated by
subtracting the intensity of the 30/32-kDa band in the same animal at
zero time (before injection of hIGFBP-3) or in animals injected with
vehicle from the intensity of the 30/32-kDa band at different times
after injection.
IGF-I RIA
The IGF-I content of whole serum or fractions in the 150- and
40-kDa region after Sephadex G-100 gel filtration of serum samples (see
above) was determined by RIA (16) after acid-ethanol extraction (17).
The total amount of IGF-I contained in the 150-kDa (30–44% bed
volume) and 40-kDa (44–61% bed volume) fractions was calculated as
the sum of the IGF-I content in individual fractions. Fractions
corresponding to the free IGF pool (61–96% bed volume) were collected
in tubes containing 0.5 ml 2 x RIA buffer (1 x = 0.03
M sodium phosphate, pH 7.5; 0.02% protamine sulfate;
0.02% sodium azide; 0.01 M EDTA; and 0.05% Tween-20) to
prevent any loss of IGF. The IGF-I content of the free IGF pool was
determined without acid-ethanol extraction.
|
Results
|
|---|
Injected hIGFBP-3 associates with rALS to form 150-kDa
complexes
Rats were injected with hIGFBP-3, hIGFBP-3:IGF-I complexes, or
vehicle; the blood was collected at various times from 2–240 min after
injection, and IGFBPs in the unfractionated serum were analyzed by
ligand blotting (Fig. 1
). In animals injected with
vehicle alone (upper panel), a 40/44-kDa doublet
(corresponding to intact endogenous rIGFBP-3) and a 30-kDa band
[containing a proteolytic fragment of rIGFBP-3 (see below) and
possibly other lower molecular mass rIGFBPs] were observed.
View larger version (55K):
[in this window]
[in a new window]
|
Figure 1. Ligand blotting of sera from rats at different
times after the injection of hIGFBP-3 (middle),
hIGFBP-3:IGF-I complexes (bottom), or vehicle
(top). Adult rats were injected iv with nonglycosylated
hIGFBP-3 (animals 5 and 6), hIGFBP-3:IGF-I complexes (animals 9 and
10), or vehicle (animals 1 and 2), and blood samples were taken at the
indicated times. Serum was analyzed by ligand blotting using
[125I]IGF-II. Zero minutes refers to samples taken just
before injection. The two lanes shown at each time point are taken from
two representative animals in each group (animals 1, 5, and 9 in the
left lane; animals 2, 6, and 10 in the right
lane). The 32-kDa band corresponding to injected hIGFBP-3 is
designated by a solid arrowhead; the 30-kDa band
representing endogenous rIGFBPs is indicated by an open
arrowhead. The fifth lane in the bottom panel is
blank because the sample was lost.
|
|
In blood collected 2 min after the injection of hIGFBP-3 or
hIGFBP-3:IGF-I complexes, a prominent new band appeared at 32 kDa, the
size of the injected nonglycosylated recombinant hIGFBP-3. As estimated
by densitometry (Fig. 2), the hIGFBP-3 contents of
samples collected 2 min after the injection of hIGFBP-3 or
hIGFBP-3:IGF-I complexes were 85% and 127% that of intact rIGFBP-3,
respectively.
The level of 32-kDa hIGFBP-3 declined significantly between 2 and 8 min
after injection. The decrease (65%) was more marked for animals that
had been injected with hIGFBP-3 than for animals that had been
injected with complex (27%). This rapid clearance is consistent with
the half-life of IGF-I in 40-kDa IGFBP complexes and shorter than the
3- to 4-h half-life of IGF-I in 150-kDa complexes (18, 19), suggesting
that a substantial fraction of the injected hIGFBP-3 or
hIGFBP-3:IGF-I complexes had been cleared from the circulation before
it could associate with the excess endogenous ALS that is present in
rat plasma (20). The more rapid clearance of hIGFBP-3 after the
injection of hIGFBP-3 than after the injection of hIGFBP-3:IGF-I
complexes suggested that free hIGFBP-3 associated less readily with
rALS in vivo than did hIGFBP-3:IGF-I complexes.
After the initial rapid decline, the level of 32-kDa hIGFBP-3 decreased
at a similar slow rate for both samples; 52–57% of the hIGFBP-3
present 8 min after the injection remained 120 min after the injection
of either hIGFBP-3 or hIGFBP-3:IGF-I complexes. The appearance of this
long-lived component after the injection of hIGFBP-3 supports the
conclusion of Lewitt et al. (9) that injected hIGFBP-3
formed 150-kDa complexes with rALS. By extrapolation to the time of
injection (Fig. 2
), the abundance of 150-kDa complexes containing
hIGFBP-3 (relative to the level of intact endogenous 40–44 kDa
rIGFBP-3) was estimated to be 80% for injected hIGFBP-3:IGF-I
complexes and 30% for injected hIGFBP-3 (Table 1).
View this table:
[in this window]
[in a new window]
|
Table 1. Relative abundance of hIGFBP-3 in 150-kDa complexes
in blood collected 2 min after the injection of hIGFBP-3 or
hIGFBP-3:IGF-I complexes
|
|
Gel filtration and immunological assays were performed to demonstrate
directly that the hIGFBP-3 observed in serum from blood collected 2 min
after injection was present in 150-kDa complexes and associated with
rALS. First, the same sera were fractionated using Sephadex G-100, the
fractions were examined by ligand blotting, and the autoradiographs
were analyzed by densitometry (Fig. 3). Although the
32-kDa hIGFBP-3 band was poorly resolved from endogenous 30-kDa IGFBPs
in this experiment, the appearance of 32-kDa hIGFBP-3 in samples
collected 2 min after injection was indicated by a 2.7- to 5-fold
increase in the intensity of the 30/32-kDa autoradiographic signal
relative to the 30/32-kDa band in samples obtained before injection
(zero time). The hIGFBP-3 content of the 150-kDa region was 1.6 times
greater in serum from blood collected 2 min after the injection of
hIGFBP-3:IGF-I complexes than in that from blood collected after the
injection of hIGFBP-3 (Table 1).
View larger version (86K):
[in this window]
[in a new window]
|
Figure 3. Ligand blotting of Sephadex G-100 fractions of
sera collected at different times from rats injected with hIGFBP-3
(left) or hIGFBP-3:IGF-I complexes
(right). Sera obtained 0, 2, 30, and 240 min after the
injection of hIGFBP-3 (left panels) or from a rat
injected with hIGFBP-3:IGF-I complexes (right panels)
were fractionated on a Sephadex G-100 column at neutral pH. Aliquots
from odd-numbered fractions (no. 23–41) were analyzed by ligand
blotting using [125I]IGF-II. Fractions 23–29 (indicated
by a bracket) correspond to the 150-kDa peak. On this
column, 40-kDa [125I]IGF:IGFBP complexes elute at
fraction 34, and free [125I]IGF appears at fraction 52.
The 32-kDa hIGFBP-3 band is indicated by an arrowhead.
The sera examined were from the animals shown in the left
lanes of each group in Fig. 1. Results from the other two
animals (not shown) were virtually identical.
|
|
The hIGFBP-3 in serum from blood collected 2 min after injection was
associated with endogenous rALS. This was demonstrated by
immunoprecipitating the serum with antibody to hIGFBP-3, fractionating
the immunoprecipitates by SDS-PAGE, and immunoblotting with antiserum
to rALS (Fig. 4). Immunoreactive rALS was detected in
the anti-hIGFBP-3 immunoprecipitates 2 min after the injection of
hIGFBP-3 or complex. Although the relative amounts of rALS associated
with hIGFBP-3 (immunoprecipitated) and 40- to 44-kDa intact rIGFBP-3
(not immunoprecipitated by antiserum to hIGFBP-3) could not be compared
in this experiment, the autoradiographic signal 2 min after the
injection of hIGFBP-3:IGF-I complexes was 1.5-fold greater than that
observed after the injection of hIGFBP-3 alone.
View larger version (42K):
[in this window]
[in a new window]
|
Figure 4. Immunoblotting of serum with antiserum to rALS
after immunoprecipitation with hIGFBP-3 antibodies. Sera from rats
injected with hIGFBP-3 (upper) or hIGFBP-3:IGF-I
complexes (lower), collected at the indicated times
after injection, were immunoprecipitated with rabbit antiserum to
hIGFBP-3, and the immunoprecipitates were analyzed by immunoblotting
using rabbit antiserum to rALS and [125I]protein A. The
85-kDa rALS band is shown by an arrow. The sera examined
were from the same two animals shown in Fig. 1.
|
|
Association of injected hIGFBP-3 with rALS in samples from blood
obtained 2 min after injection also was demonstrated by ligand blotting
following immunoprecipitation with rALS antiserum (Fig. 5). The 32-kDa hIGFBP-3 and 30-kDa rIGFBP bands were
clearly resolved in this experiment, so that it was possible to compare
the 32-kDa hIGFBP-3 signal directly with that of 40- to 44-kDa intact
rIGFBP-3. In samples collected 2 min after injection, hIGFBP-3 was 46%
as abundant as intact rIGFBP-3 when hIGFBP-3 was injected alone, and
188% as abundant when hIGFBP-3:IGF-I complexes were injected.
View larger version (82K):
[in this window]
[in a new window]
|
Figure 5. Ligand blotting of serum after immunoprecipitation
with rALS antibodies. The same sera examined in Fig. 1 were
immunoprecipitated with rALS antibodies, and the immunoprecipitates
were analyzed by ligand blotting using [125I]IGF-II. In
lane 13, the same serum used in lane 4 was immunoprecipitated with
nonimmune rabbit serum. The 32-kDa hIGFBP-3 band is indicated by an
arrowhead.
|
|
In summary, we have confirmed the observation of Lewitt et
al. (9) that in serum prepared from blood collected 2 min after
the injection of hIGFBP-3, hIGFBP-3 is found in 150-kDa complexes and
is associated with rALS. As summarized in Table 1, we estimate from the
various experimental approaches that 150-kDa complexes containing
hIGFBP-3 in these samples are 30–82% as abundant as endogenous
150-kDa complexes that contain intact rIGFBP-3. In the comparable
samples from blood collected 2 min after injection of hIGFBP-3:IGF-I
complexes, 150-kDa complexes containing hIGFBP-3 were 80–188% as
abundant as those containing intact rIGFBP-3. We estimate that the
abundance of 150-kDa complexes containing hIGFBP-3 was 1.5–4.1 times
greater in samples prepared from blood collected 2 min after injection
of hIGFBP-3:IGF-I complexes than from blood collected 2 min after
injection of hIGFBP-3 alone, suggesting that although hIGFBP-3 can
complex with rALS without IGF-I, 150-kDa complex formation is more
efficient when hIGFBP-3:IGF-I complexes are injected.
Formation of 150-kDa complexes 2 min after injection of hIGFBP-3 is
not accompanied by a corresponding increase in IGF-I
To determine whether the increase in hIGFBP-3 associated with
150-kDa complexes in blood collected 2 min after the injection of
hIGFBP-3 was accompanied by a comparable increase in IGF-I, IGF-I RIA
was performed on whole serum after acid-ethanol extraction to remove
interfering IGFBPs (Fig. 6). The IGF-I content of whole
serum obtained from animals 2 min after the injection of hIGFBP-3 was
not significantly increased from preinjection (zero time) values in the
same animal.
View larger version (15K):
[in this window]
[in a new window]
|
Figure 6. IGF-I concentrations in unfractionated sera. The
IGF-I concentrations in sera from rats injected with hIGFBP-3
(open circles), hIGFBP-3:IGF-I complexes (solid
circles), or vehicle (PBS; open squares) were
measured by RIA after acid-ethanol extraction. Data represent the
mean ± SE of the four animals in each group.
|
|
Next, the samples were fractionated by gel filtration on Sephadex G-100
at neutral pH, and the IGF-I content in the 150-kDa region was analyzed
by RIA (Fig. 7). In the two animals shown that were
injected with hIGFBP-3, IGF-I in the 150-kDa fractions of serum from
blood collected 2 min after injection was slightly (14%) greater than
the IGF-I content of the 150-kDa fractions in serum from blood
collected before injection. The magnitude of this increase is smaller
than the abundance of 150-kDa complexes containing hIGFBP-3, estimated
as 46–82% compared to 150-kDa complexes containing endogenous intact
rIGFBP-3 (Table 1).2 By contrast, in rats
injected with hIGFBP-3:IGF-I complexes, a larger increase in the IGF-I
content of the 150-kDa peak was observed (124% and 263%) similar to
the abundance of 150-kDa complexes containing hIGFBP-3 (80–188% that
of complexes containing endogenous intact rIGFBP-3; Table 1).
Kinetics of disappearance of 150-kDa hIGFBP-3 complexes from
plasma
The preceding results showing that hIGFBP-3 formed 150-kDa
complexes with endogenous rALS in blood collected 2 min after injection
and that the formation of these complexes was not accompanied by a
corresponding increase in IGF-I in either whole serum or the 150-kDa
fraction appeared to indicate that hIGFBP-3 could form binary complexes
with rALS in vivo. To confirm that the 150-kDa complexes
containing hIGFBP-3 had formed in vivo, we examined the
clearance of these complexes from plasma, anticipating that they would
have the same long half-life in the circulation as 150-kDa ternary
complexes that contained IGF-I regardless of whether they were formed
after the injection of hIGFBP-3 or hIGFBP-3:IGF-I
complexes.3
The kinetics of disappearance of hIGFBP-3 from 150-kDa complexes formed
after the injection of hIGFBP-3 or hIGFBP-3:IGF-I complexes are
compared in the experiment shown in Fig. 5 and analyzed quantitatively
in Fig. 8. (This experiment was chosen for analysis
because it gave the best resolution of 32-kDa hIGFBP-3 and 30-kDa
rIGFBPs.) In this experiment, IGFBPs complexed to rALS and
immunoprecipitated by antiserum to rALS were examined by ligand
blotting at different times after injection. Two minutes after the
injection of hIGFBP-3:IGF-I complexes, immunoprecipitated complexes
containing hIGFBP-3 were 188% as abundant as complexes containing
intact rIGFBP-3 (Fig. 8). As expected for 150-kDa complexes, the
relative abundance of complexes containing hIGFBP-3 was virtually
unchanged in samples obtained up to 4 h after the injection of
hIGFBP-3:IGF-I complexes.
View larger version (16K):
[in this window]
[in a new window]
|
Figure 8. Kinetics of disappearance of hIGFBP-3 from
complexes immunoprecipitated by antiserum to rALS after the injection
of hIGFBP-3 (solid circles, left scale) or
hIGFBP-3:IGF-I complexes (open circles, right scale).
The autoradiographs of the ligand blot shown in Fig. 5 were analyzed by
quantitative densitometry. The amount of 32-kDa hIGFBP-3 was determined
and is expressed as a percentage of intact rIGFBP-3 and plotted against
the time after injection when the samples were collected. The mean of
the two samples is plotted. Results for the other two animals in each
group were similar (not shown).
|
|
In contrast to the persistence of hIGFBP-3 complexes formed after the
injection of hIGFBP-3:IGF-I, the abundance of hIGFBP-3 complexes formed
after the injection of hIGFBP-3 decreased progressively. Two
minutes after the injection of hIGFBP-3, immunoprecipitated complexes
containing hIGFBP-3 were 46% as abundant as complexes containing
intact rIGFBP-3, but became only 17% as abundant as complexes
containing intact rIGFBP-3 by 120 min after injection. This result was
unexpected. The hIGFBP-3 that was complexed to rALS in blood collected
after the injection of hIGFBP-3 disappeared from these complexes
considerably more rapidly than would have been expected from the slow
in vivo clearance of 150-kDa complexes (18, 19).
One possible explanation for this result is that the complexes observed
in blood collected 2 min after the injection of hIGFBP-3 might not have
formed in vivo. This possibility was examined by adding
hIGFBP-3 ex vivo to blood collected before the injection of
hIGFBP-3 in an amount comparable to the amount of hIGFBP-3 that would
be present 2 min after the injection of 0.5 mg/kg (Fig. 9).4 This sample together
with blood collected 2 min after in vivo injection of
hIGFBP-3 were processed rapidly and fractionated on Sephadex G-100.
Densitometric analysis of the ligand blot indicated that 32-kDa
hIGFBP-3 in fractions in the 150-kDa region of the ex vivo
sample was 23% as abundant as intact rIGFBP-3 in the same fractions. A
nearly identical amount of 32-kDa hIGFBP-3 (24%) was present in the
150-kDa region 2 min after injecting the same rat with hIGFBP-3. We
conclude that serum prepared from blood collected 2 min after in
vivo injection of hIGFBP-3 contains sufficient free hIGFBP-3 to
allow the formation ex vivo of 150-kDa complexes in
substantial amounts despite rapid processing of the samples.
View larger version (46K):
[in this window]
[in a new window]
|
Figure 9. Comparison of complex formation between hIGFBP-3
and rALS ex vivo and in vivo. Upper panels, ex
vivo, Blood was collected before the injection of hIGFBP-3, and
3 µg hIGFBP-3/ml blood were added to the sample ex
vivo. The blood was clotted on ice for 20 min and centrifuged
(10 min, 4 C, 16,000 x g), and 0.1 ml serum was
immediately loaded onto the Sephadex G-100 column. Aliquots of
odd-numbered fractions (no. 21–41) were analyzed by ligand blotting.
Fractions 23–29 (brackets) represent the 150-kDa peak.
Lower panels, in vivo, The same animal was injected with
0.5 mg/kg hIGFBP-3. Blood was collected 2 min after the injection and
processed rapidly for Sephadex G-100 gel filtration as described above,
except that no hIGFBP-3 was added ex vivo.
Autoradiographs of the ligand blots are presented. Lane S represents
the same serum before gel filtration. The position of hIGFBP-3 is
indicated by an arrowhead.
|
|
|
Discussion
|
|---|
Lewitt et al. (9) made the intriguing observation that
2 min after hIGFBP-3 was injected iv into adult rats, it appeared in
150-kDa complexes that were presumed to contain equimolar amounts of
endogenous rALS and IGF-I. Although rat plasma contains excess ALS
(20), it contains negligible amounts of free IGF-I. This led the
authors to postulate that IGF-I, in an amount approximately equal to
the total concentration of IGF-I in plasma before hIGFBP-3 was
injected, had been mobilized from the tissues to the circulation. They
did not, however, demonstrate that the increase in the IGF-I content of
the 150-kDa complexes after injection of hIGFBP-3 required by their
proposed mechanism had actually occurred. In fact, Zapf et
al. (21), performing similar hIGFBP-3 injection experiments, were
unable to detect a significant change in IGF-I levels in the serum of
rats injected with hIGFBP-3 during a 4-h period of observation. These
investigators did not, however, examine the hIGFBP-3 and IGF-I contents
of the 150-kDa peak. The present studies were undertaken to resolve
whether IGF-I had been mobilized from tissues to 150-kDa complexes
containing iv injected hIGFBP-3 by determining whether injected
hIGFBP-3 rapidly appeared in 150-kDa complexes, and whether complex
formation was accompanied by a corresponding increase in the IGF-I
content of 150-kDa fractions.
First, we confirmed the observation of Lewitt et al. (9)
that after iv injection of hIGFBP-3 into adult rats, the hIGFBP-3
became associated with rALS in 150-kDa complexes in blood collected 2
min after injection. This was indicated by the presence of a slowly
disappearing component of hIGFBP-3 (which we presumed to be 150-kDa
complexes) when whole serum was examined by ligand blotting. This long
half-life component was approximately 30% as abundant as endogenous
rIGFBP-3 in unfractionated serum. Association of the injected hIGFBP-3
with rALS in 150-kDa complexes in blood collected 2 min after injection
was demonstrated directly by gel filtration and coimmunoprecipitation
of hIGFBP-3 and rALS. The relative abundance of 150-kDa complexes
containing hIGFBP-3 in fractionated or immunoprecipitated samples was
estimated to be 46–82% that of 150-kDa complexes containing intact
endogenous rIGFBP-3, somewhat lower than the approximately equal
abundance of hIGFBP-3 and rIGFBP-3 in 150-kDa complexes 2 min after
injection reported by Lewitt et al. (9).
Formation of 150-kDa complexes also was evaluated in the same
experiments after the injection of preformed binary complexes of
hIGFBP-3:IGF-I. Several significant differences were observed. First,
in unfractionated samples, hIGFBP-3 was present at a higher
concentration in serum obtained 2 min after the injection of
hIGFBP-3:IGF-I complexes than in serum obtained 2 min after the
injection of hIGFBP-3 (127% vs. 85% relative to intact
rIGFBP-3) and decreased more slowly between 2 and 8 min (27% decrease
vs. 65%). These results are similar to those reported by
Zapf et al. (21), who observed that only 5–10% as much
hIGFBP-3 was retained in serum 5 min after the injection of hIGFBP-3
than after the injection of hIGFBP-3:IGF-I complexes. Secondly, in
fractionated or immunoprecipitated samples, 150-kDa complexes
containing hIGFBP-3 were 1.5–4 times more abundant when hIGFBP-3:IGF-I
complexes were injected than after the injection of hIGFBP-3. Together,
these results suggested that hIGFBP-3:IGF-I complexes associated more
efficiently with rALS than did free hIGFBP-3. The more rapid initial
decline in total hIGFBP-3 after the injection of hIGFBP-3 was thought
to represent clearance of free hIGFBP-3 that had not yet associated
with rALS to form longer half-life 150-kDa complexes.
Having confirmed that hIGFBP-3 formed 150-kDa complexes with rALS
within 2 min after iv injection, albeit less efficiently than injected
hIGFBP-3:IGF-I complexes, we next determined the effect of hIGFBP-3
injection on the IGF-I content of unfractionated serum and 150-kDa
serum fractions. In unfractionated serum, similar to the results of
Zapf et al. (21), we did not observe any increase in the
IGF-I content in contrast with a 30% increase in the slowly cleared
hIGFBP-3 in 150-kDa complexes. In fractionated serum, a small
(14%) increase was observed in the IGF-I content of 150-kDa fractions
2 min after the injection of hIGFBP-3, significantly lower than the
46–82% increase in the hIGFBP-3 content of these fractions. These
results suggested that hIGFBP-3 rapidly associated with endogenous
rALS to form 150-kDa complexes, but that this association did not
require the mobilization of IGF-I as proposed by Lewitt et
al. (9). They are consistent, however, with our in
vitro studies (10) showing that hIGFBP-3 can form binary 150-kDa
complexes with rALS in the absence of IGF-I.
If the foregoing interpretation were correct, one would predict that
the putative 150-kDa binary complexes formed after the injection of
hIGFBP-3 would be cleared as slowly from the circulation as the 150-kDa
ternary complexes formed after the injection of hIGFBP-3:IGF-I because
they would be similar in size. To our surprise, this prediction was not
confirmed. Complexes containing hIGFBP-3 that were immunoprecipitated
by antiserum to rALS persisted in the circulation for 240 min after the
injection of hIGFBP-3:IGF-I complexes, but were rapidly cleared from
the circulation after the injection of hIGFBP-3 alone (decreasing from
46% to 17% between 2–120 min). Lewitt et al. (9) also
observed a more rapid disappearance of hIGFBP-3 from the 150-kDa peak
in serum from rats injected with hIGFBP-3 than in that from rats
injected with hIGFBP-3:IGF-I complexes, but did not comment on these
results.
The shorter half-life of hIGFBP-3 in the serum of rats that had been
injected with hIGFBP-3 was difficult to reconcile with the
interpretation that the injected hIGFBP-3 had associated with rALS
in vivo to form 150-kDa complexes. An alternative
explanation was that the rapid clearance reflected the fact that
appreciable hIGFBP-3 remained as free hIGFBP-3 in these samples and
that the observed 150-kDa complexes had formed ex vivo after
the collection of blood. To examine this possibility, blood was
collected from rats before they were injected with hIGFBP-3, and
hIGFBP-3 was added to it ex vivo at a concentration
comparable to that which would be present 2 min after the injection of
hIGFBP-3. For comparison, a second sample was collected 2 min after the
same rat was injected with hIGFBP-3. Both samples were processed
rapidly and fractionated by gel filtration, and the distribution of
IGFBP-3 in the fractions was examined by ligand blotting. Nearly
identical amounts of hIGFBP-3 were present in the 150-kDa peak in both
samples. We conclude that free hIGFBP-3 remaining in blood collected 2
min after the injection of hIGFBP-3 can form 150-kDa complexes ex
vivo, undoubtedly contributing to our estimates of the abundance
of these complexes in size-fractionated or immunoprecipitated samples.
Although the experimental procedures used by Lewitt et al.
(9) were not identical to ours (e.g. using purified
glycosylated hIGFBP-3 rather than nonglycosylated recombinant hIGFBP-3
and separating plasma immediately), we suggest that ex vivo
association of hIGFBP-3 with rALS probably occurred in their
experiments as well, as they also observed that hIGFBP-3 disappeared
more rapidly from 150-kDa complexes formed after the injection of
hIGFBP-3 than when hIGFBP-3:IGF-I complexes had been injected.
The slowly disappearing component observed in our studies as well as
those of Lewitt et al. (9) indicates that iv injected
hIGFBP-3 does rapidly associate with rALS in vivo. We
estimate from unfractionated samples that the 150-kDa complexes
containing hIGFBP-3 are at least 30% as abundant 150-kDa complexes
containing endogenous rIGFBP-3.5 Formation
of these complexes is not accompanied by an increase in the IGF-I
content of unfractionated serum. Our attempts to refine this analysis
using 150-kDa fractions indicate a small (14%) increase in the IGF-I
content of 150-kDa fractions 2 min after the injection of hIGFBP-3 in
contrast with an apparent 46–82% increase in hIGFBP-3. However, the
abundance of 150-kDa complexes containing hIGFBP-3 in these samples is
almost certainly an overestimate because it includes the contributions
of both 150-kDa hIGFBP-3 complexes that truly formed in vivo
and complexes that formed ex vivo after blood
collection.
In summary, our results confirm that after iv injection, hIGFBP-3
rapidly forms 150-kDa complexes, but the magnitude of this association
is lower than previously estimated. Although we found no evidence that
IGF-I has been mobilized to the circulation after hIGFBP-3 injection as
previously proposed, because of the uncertainties introduced by the
additional formation of 150-kDa complexes ex vivo, we were
unable to conclusively resolve the question of whether the 150-kDa
complexes that formed in vivo 2 min after the injection of
hIGFBP-3 are binary complexes, ternary complexes, or a mixture of
both.
|
Acknowledgments
|
|---|
We thank Peter Nissley for critical reading of the
manuscript.
|
Footnotes
|
|---|
1 Present address: Department of International Livestock Industry,
Chinju National University, Chinju 660-758, Korea. 
2 We have quantitated 150-kDa complexes containing
hIGFBP-3 relative to intact endogenous rIGFBP-3 because this is the
IGFBP species that is responsible for binding most of the IGF-I in rat
plasma. Similarly, we have expressed the increase in the IGF-I content
of the 150-kDa peak after injection (representing IGF-I bound to
complexes containing either hIGFBP-3 or intact endogenous rIGFBP-3)
relative to the level before injection (representing IGF-I bound to
intact endogenous rIGFBP-3). Assuming that intact endogenous rIGFBP-3
was saturated with IGF-I before injection, if IGF-I binding were a
prerequisite for association with ALS in vivo, one would
anticipate that hIGFBP-3 appearing in 150-kDa complexes would be
accompanied by an equal increase in IGF-I content.
3 This argument assumes that the half-lives of
putative 150-kDa binary complexes of hIGFBP-3 and rALS would not differ
appreciably from those of 150-kDa ternary complexes because the
difference in size of the two complexes would be minimal.
4 The amount of hIGFBP-3 added ex
vivo in the experiment shown in Fig. 9 (3 µg/ml blood) is
less than the hIGFBP-3 concentration in rats during the first 2 min
after injection, as estimated from the kinetics of disappearance
presented in Fig. 2. Specifically, the concentration of injected
hIGFBP-3 was about 8 µg/ml (0.5 mg/kg; 60 ml blood/kg). If the
half-life of free hIGFBP-3 were 2 min, approximately 4 µg/ml
hIGFBP-3 would remain in the blood 2 min after injection. In fact,
as the half-life was estimated to be longer than 3 min (Fig. 2), the
concentration of hIGFBP-3 remaining in the blood after 2 min would be
greater than 4 µg/ml.
5 The 30% figure is probably an underestimate, as
the hIGFBP-3 content is determined from the slowly cleared 150-kDa
component, whereas intact rIGFBP-3 is present in 50-kDa as well as
150-kDa fractions.
Received October 30, 1996.
|
References
|
|---|
-
Rechler MM 1993 Insulin-like growth factor
binding proteins. Vitam Horm 47:1–114[Medline]
-
Rechler MM 1994 Non-receptor binding proteins for
insulin-like growth factors and other cytokines: modulators of peptide
action. In: Weintraub BD (ed) Molecular Endocrinology: Basic Concepts
and Clinical Correlations. Raven Press, New York, pp 155–180
-
Jones JI, Clemmons DR 1995 Insulin-like growth
factors and their binding proteins: biological actions. Endocr Rev 16:3–34[CrossRef][Medline]
-
Binoux M, Hossenlopp P 1988 Insulin-like growth
factor (IGF) and IGF-binding proteins: comparison of human serum and
lymph. J Clin Endocrinol Metab 67:509–514[Abstract]
-
Lee CY, Rechler MM 1996 Proteolysis of
insulin-like growth factor (IGF)-binding protein-3 (IGFBP-3) in
150-kilodalton IGFBP complexes by a cation-dependent protease activity
in adult rat serum promotes the release of bound IGF-I. Endocrinology 137:2051–2058[Abstract]
-
Zapf J 1993 Role of insulin-like growth factor
(IGF) II and IGF binding proteins in extrapancreatic tumour
hypoglycaemia. J Intern Med 234:543–552[Medline]
-
Baxter RC, Martin JL, Beniac VA 1989 High
molecular weight insulin-like growth factor binding protein complex.
J Biol Chem 264:11843–11848[Abstract/Free Full Text]
-
Baxter RC, Martin JL 1989 Structure of the Mr
140,000 growth hormone-dependent insulin-like growth factor binding
protein complex: determination by reconstitution and affinity-labeling.
Proc Natl Acad Sci USA 86:6898–6902[Abstract/Free Full Text]
-
Lewitt MS, Saunders H, Baxter RC 1993 Bioavailability of insulin-like growth factors (IGFs) in rats
determined by the molecular distribution of human IGF-binding
protein-3. Endocrinology 133:1797–1802[Abstract]
-
Lee CY, Rechler MM 1995 Purified rat acid-labile
subunit (ALS) and recombinant human insulin-like growth factor binding
protein-3 (IGFBP-3) can form a 150-kilodalton binary complex in
vitro in the absence of IGFs. Endocrinology 136:4982–4989[Abstract]
-
Barreca A, Ponzani P, Arvigo M, Giordano G, Minuto
F 1995 Effect of the acid-labile subunit on the binding of
insulin-like growth factor (IGF)-binding protein-3 to
[125I]IGF-I. J Clin Endocrinol Metab 80:1318–1324[Abstract]
-
Constant RB, Weintraub BD 1986 Differences in the
metabolic clearance of pituitary and serum thyrotropin (TSH) derived
from euthyroid and hypothyroid rats: effects of chemical
deglycosylation of pituitary TSH. Endocrinology 119:2720–2727[Abstract]
-
Szkudlinski MW, Thotakura NR, Tropea JE, Grossmann M,
Weintraub BD 1995 Asparagine-linked oligosaccharide structures
determine clearance and organ distribution of pituitary and recombinant
thyrotropin. Endocrinology 136:3325–3330[Abstract]
-
Lee CY, Rechler MM 1995 A major portion of the
150-kilodalton insulin-like growth factor binding-protein (IGFBP)
complex in adult rat serum contains unoccupied, proteolytically nicked
IGFBP-3 that binds IGF-II preferentially. Endocrinology 136:668–678[Abstract]
-
Yang YW-H, Rechler MM 1993 Identification and
characterization of insulin-like growth factor-binding proteins. In:
dePablo F, Scanes CG, Weintraub BD (eds) Techniques in Endocrine
Research. Academic Press, New York, pp 181–204
-
Lee CY, Bazer FW, Etherton TD, Simmen FA 1991 Ontogeny of insulin-like growth factors (IGF-I and IGF-II) and
IGF-binding proteins in porcine serum during fetal and postnatal
development. Endocrinology 128:2336–2344[Abstract]
-
Daughaday WH, Mariz IK, Blethen SL 1980 Inhibition
of access of bound somatomedin to membrane receptor and immunobinding
sites: a comparison of radioreceptor and radioimmunoassay of
somatomedin in native and acid-ethanol-extracted serum. J Clin
Endocrinol Metab 51:781–788[Medline]
-
Schoenle EJ, Haselbacher GK, Briner J, Janzer RC,
Gammeltoft S, Humbel RE, Prader A 1986 Elevated concentration of
IGF II in brain tissue from an infant with macrencephaly. J
Pediatr 108:737–740[CrossRef][Medline]
-
Zapf J, Hauri C, Waldvogel M, Froesch ER 1986 Acute
metabolic effects and half-lives of intravenously administered
insulin-like growth factors I and II in normal and hypophysectomized
rats. J Clin Invest 77:1768–1775
-
Baxter RC 1988 Characterization of the acid-labile
subunit of the growth hormone-dependent insulin-like growth factor
binding protein complex. J Clin Endocrinol Metab 67:265–272[Abstract]
-
Zapf J, Hauri C, Futo E, Hussain M, Rutishauser J, Maack
CA, Froesch ER 1995 Intravenously injected insulin-like growth
factor (IGF) I/IGF binding protein-3 complex exerts insulin-like
effects in hypophysectomized, but not in normal rats. J Clin
Invest 95:179–186
This article has been cited by other articles:
|
|
|
|
 
S. M. Twigg and R. C. Baxter
Insulin-like Growth Factor (IGF)-binding Protein 5 Forms an Alternative Ternary Complex with IGFs and the Acid-labile Subunit
J. Biol. Chem.,
March 13, 1998;
273(11):
6074 - 6079.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. A. Hill and J. M. Pell
Regulation of Insulin-Like Growth Factor I (IGF-I) Bioactivity in Vivo: Further Characterization of an IGF-I-Enhancing Antibody
Endocrinology,
March 1, 1998;
139(3):
1278 - 1287.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
