| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Diagnostics Systems Laboratories, Inc. (Canada) (A.D., J.K.), Toronto, Ontario, Canada M5G 1X5; Diagnostics Systems Laboratories, Inc. (J.M., R.G.K.), Webster, Texas 77598; and Department of Laboratory Medicine and Pathobiology, Faculty of Medicine, University of Toronto (J.K.), Toronto, Ontario, Canada M5G 1L5
Address all correspondence and requests for reprints to: J. Khosravi, Ph.D., Diagnostics Systems Laboratories, Inc. (Canada), Mount Sinai Hospital, Room 653, 600 University Avenue, Toronto, Ontario, Canada M5G 1X5. E-mail: jkhosravi{at}mtsinai.on.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Experimental evidence supporting the potential involvement of the regulatory IGFBP proteases stems from the initial observations of pregnancy-associated IGFBP-3 proteolysis (14, 15, 16, 17) and biochemical identification of prostate-specific antigen as an IGFBP-3 protease (18). Numerous reports have since emerged, describing the production of IGFBP proteases by a variety of cell types, and significant enhancement of IGFBP-3 proteolytic activity in response to several different pathophysiological conditions, including diabetes, various catabolic states, and malignant and benign proliferative disorders (3, 4, 8, 13, 14). These processes appear to involve enzymes belonging to at least three different classifications, including kallikreins, cathepsins, and metalloproteinases (13, 19, 20, 21). Intriguing new evidence has established a positive link between IGF/IGFBP dynamics and cancer development risk (22, 23, 24, 25, 26, 27) and has identified the pregnancy-associated plasma protein A as an IGF-dependent IGFBP-4 protease (28).
Our knowledge of IGFBP-3 proteolysis has been aided by methodologies that rely on separation of intact and proteolyzed IGFBP-3 by SDS-PAGE followed by application of detection systems based on autoradiography of 125I-labeled IGFBP-3 degradation or by either ligand or immunoblot analysis using radiolabeled IGF or anti-IGFBP-3 antibodies, respectively (13, 29, 30). Although highly useful for comparative evaluation of IGFBP-3 proteolytic profiles, the methodology is at best semiquantitative and is of limited use for large scale research and clinical evaluations. As proteolytically altered IGF/IGFBP-3 balance may represent an underlying defect or a normal response to altered physiology, methods for direct quantification and monitoring of changes in IGFBP-3 proteolytic activity may be of significant value. We recently reported development of an enzyme-linked immunosorbent assay (ELISA) for direct quantification of IGFBP-3 proteolytic activity in seminal plasma (31). To further expedite investigations of pathophysiological relevance and potential clinical applications of IGFBP-3 proteolysis, we here report the development and preliminary evaluation of three novel ELISAs capable of differential determination of intact and fragmented IGFBP-3 variants. We present data on the ability of the assays to quantify total (intact and fragments), intact, and proteolyzed (fragmented) IGFBP-3 and demonstrate their potential for research or clinical investigations by comparative measurements of IGFBP-3 in various biological fluids (serum, seminal plasma, amniotic fluids), pregnancy serum, and breast tumor cytosol.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Nonpregnancy serum samples were obtained from Diagnostics Systems Laboratories, Inc. (Webster, TX). The specimens were collected from a consenting population of apparently normal laboratory personnel (14 women and 14 men, aged 20–60 yr; median age, 35 yr). Upon collection, blood samples were allowed to clot, then separated, and the serum portion was stored at -70 C for less than 4 weeks before analysis. Pregnancy serum (PS; n = 30), amniotic fluid (AF; n = 30), seminal plasma (SP; n = 14), and human breast tumor cytosolic extracts (n = 18) were provided by Dr. E. P. Diamandis (Mount Sinai Hospital, Toronto, Canada). The PS and AF were from different group of individuals, collected at 14–17 weeks gestation (aged 20–45 yr). The cytosolic cell extracts were prepared from tumor tissues surgically removed from patients undergoing treatment for primary breast carcinoma. All specimens were residuals from routine or research tests samples and were stored frozen at -70 C until analyzed.
Recombinant human IGF-I, IGF-II, IGFBP-3, and IGFBP-2 to -6 were obtained as previously described (36). Purified IGFBP-1 was obtained from Diagnostic Systems Laboratories, Inc. Glycosylated recombinant IGFBP-3 (rIGFBP-3) was a product of Austral Biologicals (San Roman, CA). Human thrombin was obtained from Sigma (St. Louis, MO), and plasmin was obtained from Fluka (Ronkonkoma, NY). Other materials and chemicals were obtained as previously described (32).
Polyclonal and monoclonal antibodies (mAb 1–10) to human IGFBP-3 were raised by Diagnostics Systems Laboratories, Inc., against nonglycosylated rIGFBP-3 and purified by affinity chromatography. The specificity of the antibodies for synthetic fragments of IGFBP-3 (33) and their epitope recognition of the high molecular weight IGFBP complex (34) were recently described. The ELISA assay buffers, stopping solution, and coating and blocking buffers were described previously (32). The standard matrix was heat-inactivated normal goat serum.
Procedures
Procedures for antibody coating of microwells and antibody conjugation to horseradish peroxidase have been previously described (32, 35). Lyophilized nonglycosylated rIGFBP-3 was reconstituted with deionized water and appropriately diluted with the standard matrix to give reference standard values of 2–100 µg/L. The standards were stable for up to 7 days at 4 C and for more than 6 months at -20 C or lower.
Intact, fragment, and total IGFBP-3 ELISAs
Development of intact (ELISA-1), fragment (ELISA-2), and total (ELISA-3) IGFBP-3 ELISAs was based on our knowledge of antibody binding specificity and IGFBP-3 complex epitope recognition derived by systematic evaluation of the antibodies in four different binding experiments (34), including further performance assessment using a polyclonal detection antibody (data not shown). The ELISA methods incorporate identical components and protocols and involve a common monoclonal capture antibody in combination with a polyclonal (ELISA-3) or two different monoclonal (ELISA-1 and -2) detection antibodies in a manner previously reported for the development of total and nonphosphorylated IGFBP-1 (36). In the assay, standards or samples (0.025 mL after appropriate dilution in the standard matrix) are added in duplicate to precoated wells, followed by addition of the assay buffer (0.05 mL) and 2-h room temperature incubation with continuous shaking. The wells are washed four times and incubated with 0.1 mL/well of the appropriate anti-IGFBP-3 antibody-horseradish peroxidase conjugate (diluted in the assay buffer to approximately 0.1–0.25 mg/L) for 30 min as described above. The wells are washed, and the reaction is developed colorimetrically as previously described (32).
IGFBP-3 ELISA validation procedure
The analytical performance characteristics of the assays were determined as previously described (35, 37). Cross-reactivity was analyzed by assaying IGF-I and IGF-II (up to 1000 µg/L), and IGFBP-1, -2, -4, -5, and -6 (up to 500 µg/L) added to the assay standard matrix.
Proteolysis of IGFBP3
Proteolysis of IGFBP-3 was accomplished by incubation of 50 µL of serum samples or rIGFBP-3 (15 µg/mL) with 100 µL of an appropriate buffer containing increasing amounts of seminal plasma (0–50 µL), plasmin (0–230 ng/mL), or thrombin (0–10 U/mL) in a 150-µL total reaction volume for 2 h at 37 C. The untreated (control) and protease-treated samples were then diluted (10- to 100-fold) with the zero standard buffer and assayed for IGFBP-3 in duplicate by ELISA-1–3. The means of IGFBP-3 levels measured in the treated samples are presented as a percentage of the mean of the expected concentrations detected in the untreated controls.
Other assays and data analysis
A commercial IGFBP-3 ELISA available from Diagnostic Systems Laboratories, Inc., was used for comparative evaluation
of the present methods. The performance characteristics of the
Diagnostic Systems Laboratories, Inc., ELISA have been
recently described (38). In principle, the assay is similar to the
IGFBP-3 ELISA-1–3 described in the present report in terms of assay
configuration and protocol, except that it involves polyclonal
antibodies for both capture and detection. As summarized below, the
Diagnostic Systems Laboratories, Inc., ELISA
(x) demonstrated the best correlation vs. the
currently described total IGFBP-3 ELISA (y) when
IGFBP-3 levels in normal serum, PS, AF, and SP were compared (r =
0.83–0.98). However, the absolute levels of IGFBP-3 measured in the
above samples by the two methods appeared sample dependent, as judged
by the slopes of the comparative measurements, which were 0.96, 0.70,
1.29, and 2.2, respectively. Whether the observed differences are due
to the relatively lower reactivity of ELISA-3 (as well as those of
ELISA-1 and -2) for the glycosylated form of IGFBP-3 (Table 1
) and/or to the differential
immunoreactivity of proteolyzed IGFBP-3 as recognized by the
Diagnostic Systems Laboratories, Inc., ELISA remains to be
investigated.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Assessment of a panel of IGFBP-3 mAbs provided information on
their binding characteristics, particularly identifying recognition
epitopes corresponding to their reported specificity for N-terminal,
C-terminal, and intermediate sequences of IGFBP-3 (34). Further
evaluation of the candidate antibodies in pairwise combinations as well
as in combinations with a polyclonal detection antibody led to the
construction of three new IGFBP-3 ELISAs. The ELISA development was
centered on identification of a capture mAb that paired with all of the
remaining Abs (34) and in sandwich ELISAs generated acceptable
analytical performance characteristics, particularly in combination
with two of the mAbs (ELISA-1 and -2) or the polyclonal Ab (ELISA-3)
selected for detection. The employment of a common capture antibody
allowed establishment of a common two-step protocol in which IGFBP-3
captured in the first step could be differentially detected in the
second step by each of the three different detection antibodies. The
IGFBP-3 binding characteristics of the individual antibodies as
determined by Western immunoblotting are summarized in Table 1
.
Analytical performance characteristics of the assays are summarized in
Table 2. The protocol optimization was
performed as previously described (32, 36).
|
The IGFBP-3 ELISAs measured considerably different concentrations
in various biological fluids (Table 3).
In randomly selected nonpregnant adults serum samples (n = 28),
the median IGFBP-3 levels determined by ELISA-1 and -3 were identical
(
2.93 mg/L), but were higher than the ELISA-2 level (0.86 mg/L) by
about 3.4-fold. Comparatively better correlation (r = 0.75–0.89)
was also observed between values by ELISA-1 and -3 vs. those
detected by the Diagnostic Systems Laboratories, Inc.,
IGFBP-3 ELISA than in comparisons involving the ELISA-2 levels (r
= 0.51; Fig. 1). In general, the results
of ELISA-3 paralleled more tightly those of the Diagnostic Systems Laboratories, Inc., method, whereas measurements by
ELISA-1 were, on the average, lower by about 20% and showed
significant scattering of the data points around the regression line
(Fig. 1).
|
|
). Regression analysis of IGFBP-3 levels detected in PS, AF,
and SP by ELISA-1–3 (y) vs those measured by the
Diagnostic Systems Laboratories, Inc., method
(x) yielded r values of 0.58–0.98 and slopes ranging from
approximately 0.4x to 19x (data not shown). Effect of IGFBP-3 proteolysis
As the observed variability in IGFBP-3 levels by ELISA-1–3 could
not be explained on the basis of antibody cross-reactivity or the state
of IGFBP-3 glycosylation (Table 2) or phosphorylation (data not shown),
the potential effects of IGFBP-3 proteolysis on the assay response
were examined.
Proteolysis by SP. Serum samples (50 µL) were incubated as
described with increasing volumes of SP (10–50 µL). The SP-treated
and untreated samples were then analyzed by ELISA-1–3. As represented
in Fig. 2, ELISA-1 detected
progressively lower concentrations in response to increasing SP
digestion, with resulting loss of up to 85% of the expected IGFBP-3
levels at 50-µL additions. Similar results were obtained when
rIGFBP-3 was similarly digested and analyzed (data not shown). In
contrast, there were no significant variations in IGFBP-3 measured by
ELISA-3 before and after SP treatment. The mean (±SD)
IGFBP-3 detected in the treated samples as a percentage of the mean of
concentrations measured in the untreated samples was 101.3 ±
6.25, which is well within the overall assay precision of less than
10%. In several repeat experiments, the results of SP digestion
collectively suggested recognition of intact IGFBP-3 by ELISA-1,
whereas ELISA-3 appeared to measure intact and proteolyzed IGFBP-3 with
similar potencies. Because of the detection of high IGFBP-3
immunoreactivity in SP by ELISA-2, addition of increasing volumes of SP
as the source of proteolytic enzymes largely obscured the assay
response to proteolysis of the exogenously added IGFBP-3 substrate.
However, the assay appeared to detect higher levels of IGFBP-3 than
could be accounted for, suggesting possible recognition of fragmented
IGFBP-3 (data not shown).
|
). Again, ELISA-3
appeared to be relatively insensitive to the effect of IGFBP-3
proteolysis, as the assay detected 111.8 ± 5.3% and 102.3
± 10.3% of the expected mean of IGFBP-3 concentrations in the
plasmin- and thrombin-digested samples, respectively. The suggestion of
preferential detection of fragmented IGFBP-3 by ELISA-2 was also
substantiated by the observation of increased IGFBP-3 immunoreactivity
in response to both plasmin and thrombin digestions. In both cases,
ELISA-2 measured up to 3-fold higher IGFBP-3 immunoreactivity in the
proteolyzed samples (Fig. 3).
|
To further substantiate the differential specificity of
ELISA-1–3, IGFBP-3 in breast tumor cytosol (n = 18) was measured.
As shown in Fig. 4, the highest level of
IGFBP-3 immunoreactivity (mean ± SD) was detected by
ELISA-2 (17.5 ± 10.3 µg/L), whereas intact and total IGFBP-3
levels measured by ELISA-1 and ELISA-3 were 4.8 ± 3.1 and
8.4 ± 4.0, respectively, indicating the fragmented nature of
IGFBP-3 in breast tumor cytosol.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We here reported the development of three novel IGFBP-3 ELISAs (ELISA-1 to -3), involving a common capture antibody which in combination with three different detection antibodies demonstrated preferential specificity for intact, proteolyzed (fragmented), and total IGFBP-3 levels, respectively. The analytical specificity of ELISA-1–3 was initially based on the finding that the assay-dependent differences in IGFBP-3 immunoreactivity detected in pregnancy sera and in various biological fluids could not be explained on the basis of antibody cross-reactivity or posttranslational modification of IGFBP-3 by glycosylation or phosphorylation. The comparative variability in IGFBP-3 levels detected appears to, therefore, reflect the reported changes in the state of IGFBP-3 proteolysis in various biological fluids (43, 44, 45) and suggests differential recognition of intact and/or proteolytically modified IGFBP-3 subfractions by ELISA-1–3. Accordingly, intact IGFBP-3 levels determined by ELISA-1 were greater in nonpregnant adult samples than in PS, which were greater than in AF and were undetectable in SP, which primarily contains fragmented IGFBP-3 (43, 44, 45). The readily detectable levels of intact IGFBP-3 in AF (0.88 ± 0.12 µg/mL) by ELISA-1 appears inconsistent with the reported failure of Western ligand blotting to detect intact IGFBP-3 in AF (4). However, intact IGFBP-3 in AF could also be detected by Western ligand blotting, provided that higher amounts of AF (5–10 µL) are loaded onto the gel (43). Although a number of variables, including differences in sensitivity and use of denaturing conditions by Western blotting, may be contributing, recognition of moderately truncated IGFBP-3 by ELISA-1 that would not efficiently bind to IGFs in ligand blotting could not be ruled out. Because demonstration of immunoblot reactivity may not completely correspond to simultaneous two-site ELISA recognition, the question of specificity could be only resolved by systematic ELISA testing of incrementally truncated (C- and/or N-terminally) native IGFBP-3, which are not currently available. In contrast to ELISA-1, ELISA-2 detected comparatively higher levels in samples expected to contain proteolyzed IGFBP-3 (PS, AF, and SP) (43, 44, 45) and measured severalfold higher levels of immunoreactivity in AF and SP than the detectable levels by ELISA-1 or even by the total IGFBP-3 ELISA-3. The latter may be explained by the ability of ELISA-2 to better bind to fragmented IGFBP-3 than the intact molecule and by the fact that intact IGFBP-3 was used for calibration. Obviously, with fragmented IGFBP-3 as standard, ELISA-2 would detect comparatively lower levels, but the approach would have been problematic, particularly with respect to consistent preparation and standardization of the fragmented IGFBP-3 preparations. As indicated, the response of ELISA-3 was relatively independent of the state of IGFBP-3 proteolysis, detecting similar levels in AF, nonpregnant, and pregnancy samples and significant levels in SP.
The differential specificity of ELISA-1–3 was further confirmed by comparative determination of the assay response to IGFBP-3 before and after digestion with proteases that reportedly cleave IGFBP-3 (13, 18, 19, 20, 31, 39, 40, 41, 46). In accordance with the above findings, ELISA-1 detected little or no immunoreactivity after IGFBP-3 digestion with SP, plasmin, or thrombin, whereas ELISA-2 generated the reverse response, measuring up to 3-fold higher postdigestion IGFBP-3 immunoreactivity. In fact, the decreasing response of ELISA-1 to IGFBP-3 proteolysis was sufficiently dose dependent to allow recent application of ELISA-1 to quantitative determination of IGFBP-3 protease activity in seminal plasma (31). In the same samples, determinations by ELISA-3 appeared independent of the extent of IGFBP-3 digestion, as the assay measured similar levels in protease-treated and untreated samples.
We propose that the demonstrated preferential specificity of ELISA-1 and -2 for intact and fragmented IGFBP-3 subfractions, respectively, should allow for highly sensitive and convenient monitoring of changes in the state of IGFBP-3 proteolysis as well as for the measurement of IGFBP-3 where proteolysis might be suspected. As exemplified in this report, the median IGFBP-3 levels in pregnancy samples by ELISA-1 and -2 were 63% and 311% of the levels found in the nonpregnant subjects, respectively. This would amount to a significant change of up to 5-fold in the median levels if the ELISA-2/ELISA-1 concentration ratio is considered. Even in situations where comparison of the test and control samples may not be possible, measurement of levels by the most appropriate method or direct comparison of the concentrations measured may be highly informative. This is readily evident by comparison of IGFBP-3 ELISA-1–3 levels measured in a given sample group and the use of concentration ratios to further improve the sensitivity of the measurements. In this context, we demonstrated a 6-fold increase in AF median IGFBP-3 levels determined by ELISA-2 relative to the levels detected by ELISA-1. Similarly, ELISA-2 detected the highest levels of IGFBP-3 immunoreactivity in SP, whereas by ELISA-1, levels were undetectable. We observed similar assay-dependent differences in IGFBP-3 measured in breast tumor cytosols, with ELISA-2, on the average, detecting about 3.6- and 2.1-fold higher levels of IGFBP-3 immunoreactivity than ELISA-1 or ELISA-3, respectively. As tumor cells invariably overproduce a variety of proteinases (13, 42), the use of well defined assay capable of measuring the relevant marker is extremely important, particularly in attempts to explore the potential relation between expression of IGFBP-3 and other diagnostic, prognostic, or therapeutic indicators of tumor growth and development. This recommendation is in line with our recent collaborative finding that determination of fragmented IGFBP-3 in breast nipple aspirate fluid may be a useful marker for the assessment of breast cancer risk (Sauter, E. R., S. Litwin, P. F. Engstrom, A. Diamandis, J. Khosravi, and E. P. Diamandis, submitted for publication).
Although immunoassays for IGFBP-3 are now widely available and accepted in routine or research investigations, there are little or no data indicating the relative impact of IGFBP-3 proteolysis on the assays’ performance and whether the assays are capable of measuring total IGFBP-3 levels. As IGFBP-3 is susceptible to significant proteolysis in response to a host of pathophysiological changes (3, 4, 8, 13, 14, 42), immunoassays capable equivalent determination of intact and fragmented IGFBP-3 would be of great value. This requirement may even be relevant to IGFBP-3 analysis under apparently normal conditions when physiological or nonphysiological proteolysis could not be expected. The present total IGFBP-3 ELISA is, therefore, ideal for general laboratory use as well as for applications where determinations of potential changes in the levels of intact and/or fragmented IGFBP-3 relative to the total levels might be important.
In summary, we describe the first report on the development and validation of highly specific and simple ELISAs for intact, fragmented, and total IGFBP-3 in serum, biological fluids, or tumor cell extracts. ELISA-1–3 are based on the selection of a common capture antibody and ELISA protocol, thus allowing convenient applications to monitoring changes in the state of IGFBP-3 proteolysis. The demonstrated specificity of the assays for IGFBP-3 subfractions should facilitate informed investigations of the pathophysiological relevance and potential clinical value of IGFBP-3 proteolysis.
|
Acknowledgments |
|---|
Received October 1, 1999.
Revised February 9, 2000.
Accepted February 27, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  D. K. Tisi, X.-J. Liu, L. J. Wykes, C. D. Skinner, and K. G. Koski Insulin-Like Growth Factor II and Binding Proteins 1 and 3 from Second Trimester Human Amniotic Fluid Are Associated with Infant Birth Weight J. Nutr., July 1, 2005; 135(7): 1667 - 1672. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Rinaldi, R. Kaaks, A. Zeleniuch-Jacquotte, A. A. Arslan, R. E. Shore, K. L. Koenig, L. Dossus, E. Riboli, P. Stattin, A. Lukanova, et al. Insulin-Like Growth Factor-I, IGF Binding Protein-3, and Breast Cancer in Young Women: A Comparison of Risk Estimates Using Different Peptide Assays Cancer Epidemiol. Biomarkers Prev., January 1, 2005; 14(1): 48 - 52. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Coutant, F. Boux de Casson, O. Douay, E. Mathieu, S. Rouleau, F. Beringue, P. Gillard, J. M. Limal, and P. Descamps Relationships between Placental GH Concentration and Maternal Smoking, Newborn Gender, and Maternal Leptin: Possible Implications for Birth Weight J. Clin. Endocrinol. Metab., October 1, 2001; 86(10): 4854 - 4859. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Binoux Tools Now Exist to Measure Circulating IGF-Binding Protein-3 Proteolysis, to Study Its Regulation and Its Effects on IGF-I Bioavailability J. Clin. Endocrinol. Metab., October 1, 2001; 86(10): 5088 - 5089. [Full Text] [PDF]
|
|
|
|
|
|
J. Khosravi, A. Diamandi, J. Mistry, and A. Scorilas Insulin-Like Growth Factor I (IGF-I) and IGF-Binding Protein-3 in Benign Prostatic Hyperplasia and Prostate Cancer J. Clin. Endocrinol. Metab., February 1, 2001; 86(2): 694 - 699. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
