0021-972X/91/7306-l216$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1991 by The Endocrine Society
Vol. 73, No. 6 Printed in U.S.A.
Ligand-Mediated Immunofunctional Assay for Quantitation of Growth Hormone-Binding Protein in Human Blood LENA M. S. CARLSSON*, ANN M. ROWLAND, ROSS G. CLARK, NEIL GESUNDHEIT, AND WAI LEE T. WONG Departments of Immunology Research and Assay Technologies (A.M.R., W.L.T. W.), Developmental Biology (L.M.S.C, R.G.C.), and Clinical Research (N.G.), Genentech, Inc., South San Francisco, California 94080
ABSTRACT. Human serum contains a high affinity GHbinding protein (GHBP) whose amino-terminal sequence is identical to the extracellular domain of the GH receptor. Current methods that measure GHBP are laborious, require size or charcoal separation of the GH/GHBP complex, and may be influenced by ambient GH concentrations. We have developed a novel assay method that allows quantitation of the total amount of functional GHBP in serum or plasma. The assay can also be used to measure the concentration of the circulating GH/ GHBP complex. An anti-GHBP monoclonal antibody, which recognizes both free GHBP and GH-bound GHBP, is used to capture the GHBP on a microtiter plate. Recombinant human GH is added to saturate all binding sites, and an anti-GH antibody conjugated with horseradish peroxidase is used to detect the amount of GH (endogenous and exogenous) bound to
T
HE PRESENCE of a protein in blood that binds to GH was first suggested in the 1960s (reviewed in Ref. 1). Reports on the conversion of GH to macromolecular form upon incubation with plasma were explained by the aggregation of radiolabeled GH, and the prevailing opinion was that GH circulated in the free form. Even though the serum GH-binding protein (GHBP) was characterized in the mouse in 1977 (2), it was not until 1986 that two groups reported the existence of a human GHBP (3, 4). There is now good evidence that the principal GHBP is identical to the extracellular domain of the hepatic GH receptor (5, 6) and that GHBP may be produced by proteolytic cleavage of the full-length GH receptor (7) or derive from a separate mRNA (8, 9). To understand hormonal mediators of growth, it is essential to be able to measure all parts of the axis (e.g. the hormone, its circulating binding protein, and the hormone-binding protein complex). Such measurements Received March 5,1991. Address requests for reprints to: Dr. Wai Lee T. Wong, Genentech, Inc., South San Francisco, California 94080. * Genentech Postdoctoral Fellow, supported in part by the Royal Swedish Academy of Sciences and the Swedish Medical Research Council.
the GHBP. The same procedure, but without incubation with GH, allows measurement of the endogenous GH/GHBP complex. The assay is sensitive (detection range, 31-2000 pmol/L), with average inter- and intraassay precisions of 11.3% and 7.3%, respectively. Measurements in random blood samples from 16 healthy adults showed that all subjects had clearly detectable GHBP concentrations (range, 65.8-305.6 pmol/L). In contrast, GHBP levels were undetectable in samples from 2 patients with Laron-type dwarfism. We believe that this ligand-mediated immunofunctional assay, which combines the simplicity and specificity of an enzyme-linked immunosorbent assay with the ability to detect only biochemically active binding protein, will be useful for studies of the role of the GHBP in health and disease. (J Clin Endocrinol Metab 73: 1216-1223, 1991)
can be difficult because of the interference in the assays by the different components and the fact that the hormone and the BP can be present in either free or bound (complexed) form. Currently, the standard method for quantitation of circulating proteins that bind to GH is incubation of serum or plasma with radiolabeled GH, followed by chromatography or charcoal separation of the bound from the free hormone (10-12). The procedures are cumbersome, and the results difficult to interpret because of the possible interference by endogenous GH in the sample (3). In addition, the proportions of free GHBP and GHBP bound with GH (complexed) cannot be determined with precision. The methods give an estimate of the binding capacity of serum proteins of a certain size, and the results are expressed in relation to reference serum pools, making it difficult to compare results from different studies. In our efforts to develop an assay for the GHBP, we chose a new approach. This report describes a very simple specific assay for quantitation of biochemically active GHBP in human serum and plasma. The ligand-mediated immunofunctional assay (LIFA) is based on the binding of GHBP to its ligand GH and the 1216
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
LIFA FOR GHBP
specific reactions between both the immobilized monoclonal antibody (MAb) to the GHBP and the enzymelabeled MAb to the GH bound to GHBP. The LIFA can measure both total and complexed GHBP and does not require prior removal of endogenous GH or physicochemical methods to separate free GH from the GH/GHBP complex. The assay, which does not involve the use of unstable radioactive chemicals, is simple, fast, and suitable for large scale clinical studies.
Materials and Methods Reagents Recombinant 22-kDa human GH (rhGH), 20-kDa rhGH, human placental lactogen (rhPL), human PRL (rhPRL), rhCD4, HIV envelope protein (rgpl20), HER2 extracellular domain (rHER2 ECD), human PRL receptor (rhPRL receptor), and human GHBP (rhGHBP) were produced at Genentech (South San Francisco, CA). The epidermal growth factor (EGF) receptor was purified from human epidermoid carcinoma A431 cell by T. Lipari (Genentech, Inc.), using previously described methods (13). Native GHBP was purified from human serum by S. Spencer (Genentech), essentially as described previously (6). rhGH was labeled with 125I using a modified chloramine-T procedure (14) with sequential addition of chloramine-T reagent to give [12BI]rhGH with a specific activity of 94 ^Ci//ig. Selection of coat MAb
Four MAbs to the rabbit liver GH receptor, MAbs 1, 2, 5, and 7, were provided by Dr. M. Waters (University of Queensland, Queensland, Australia), and an additional two MAbs (43 and 263, Agen, Australia), were purified from mouse ascites fluid using protein-A-Sepharose (Repligen Corp., Cambridge, MA) following established procedures (15, 16). Using radiolabeled hGH for detection, MAbs were evaluated to determine their binding sites on rhGHBP relative to rhGH (rhGHBP and [125I]rhGH added sequentially to the plate), their ability to bind to rhGHBP in the presence of [125I]rhGH (simultaneous incubation format), and their ability to bind to the rhGHBP/[125I] rhGH-complex (rhGHBP and [125I]rhGH preincubated and then added to the plate). MAb 263 was the most suitable MAb as coat, since it gave the highest percentage of bound [125I] rhGH in all three conditions and was able to bind the free rhGHBP as well as the rhGHBP bound to rhGH. The optimal coat concentration for the LIFA (MAb 263 at 10 Mg/mL) was determined as the lowest coat antibody concentration that resulted in maximal signal. Enzyme-conjugated anti-hGH antibodies The anti-GH MAb (MCB), which was provided by B. Fendly (Genentech), was selected for conjugation because it contains no overlapping determinants with the GHBP-binding site (17). The antibodies were purified from ascites fluid using proteinA-Sepharose (Repligen Corp.) following established procedures (15,16) and stored sterile in PBS at 4 C. A polyclonal anti-GH was purified from goat serum using recombinant protein-G-
1217
Sepharose (Zymed, South San Francisco, CA) (18). Purified antibodies were conjugated to horseradish peroxidase (HRPO) (19) and stored at -20 C in 50% glycerol. LIFA assay standards and samples The rhGHBP was purified from a mammalian cell line by P. McKay (Genentech) following the procedure outlined by Spencer et al. (6). The purified rhGHBP amino acid composition, as determined by quantitative amino acid analysis, matched what was theoretically expected for the cloned gene product. The purified rhGHBP was homogeneous based on analysis of sodium dodecyl sulfate-gel. The concentrations of biochemically active rhGHBP and native GHBP in the purified preparations were established by Scatchard analysis (20). The association constants (Ka) were 2.9 X 109 and 3.3 X 109 M"1 for the native and recombinant GHBP, respectively. When appropriate, the concentration of rhGHBP was converted to micrograms per L using 26 kDa as the mol wt of the protein part of the molecule. Dilutions of rhGHBP in PBS containing per liter, 5 g BSA, 5.0 mM EDTA, 0.5 mL Tween-20, and 0.1 g Thimerosal (assay buffer) were used as standards in the LIFA. Serum and plasma (EDTA, citrate, and heparin) samples were obtained from healthy adult volunteers taking no medications, except for subject 3, who was receiving fluoxetine. All subjects, except no. 1 and 14, were nonfasting, and the samples were taken between 0900-1100 h. The samples were centrifuged and stored at -70 or -20 C until assayed. Sera from two unrelated patients with Laron-type dwarfism were supplied by Dr. L. Underwood (University of North Carolina, Chapel Hill, NC) and Dr. R. Rosenfeld (Stanford University, Stanford, CA). LIFA assay procedure The assay procedure is outlined in Fig. 1. Ninety-six-well microtiter plates (Corning Glass Works, Corning, NY) were coated with MAb 263 by incubating overnight at 4 C with 100 /uL/well antibody at 10 Mg/mL in 50 mmol/L carbonate buffer, pH 9.6. After removal of the coating solution, nonspecific binding sites on the coated wells were blocked with 150 nL phosphate-buffered isotonic saline pH 7.2 (PBS), containing 5 g BSA (98-99% pure; Sigma Chemical Co., St. Louis, MO) per L for 1 h at room temperature, followed by six washes at room temperature with wash buffer (0.5 mL Tween-20 and 0.1 g Thimerosal/L PBS). Standards or samples (50 jtL/well) were dispensed into the coated wells containing 50 juL/well 0 ixg/L (for measurement of GH/GHBP complex concentration) or 200 jug/L rhGH (for measurement of total GHBP concentration) in sample buffer (assay buffer with 10 g/L mouse immunoglobulin G (IgG) from Fitzgerald Industries, Chelmsford, MA). Plates were sealed, incubated at room temperature for 2 h with gentle agitation, then washed before the addition of HRPO-labeled MAb MCB (100 juL/well). In the experiment in which the interference by heterophilic antibodies was determined, a HRPO-labeled polyclonal goat anti-rhGH antibody was used. After further incubation for 2 h at room temperature, the plates were washed six times with wash buffer. Freshly prepared substrate solution (0.4 g o-phenylenediamine dihydrochloride in 1 L PBS plus 0.4
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
CARLSSON ET AL.
1218
Assay Procedure
MAb263
Coat plates with MAb 263, 4°C overnight
JCE & M • 1991 Vol73-No6
unlabeled rhGH (to give final concentrations of 0, 6.7, 20, and 60 /ng/L) were incubated with a mixture of sample and [125I] rhGH. The radioactivity in the eluates was measured with an on-line 7-detection system. The radioactivity elution profile of the reference showed two main peaks: peak I, representing the [125I] hGH-GHBP complex, and peak II, representing free [125I] hGH. GHPB activity was determined as the area under peak I, expressed as a percentage of the sum of the areas of peak I and II.
Block at room temperature for 1 h rhGH
rhGH-GHBP complex
hGH-GHBP complex
Measurement of serum hGH concentrations Wash
Add sample and rhGH (0or200ng/ml) Incubate at room temperature for 2 h
Serum hGH concentrations were determined in duplicate by a polyclonal antibody-based enzyme-linked immunosorbent assay (ELISA) developed by A. Celniker, Genentech. Results
I Wash
Assay range and sensitivity anti-hGH-HRPO antibody
Add anti-hGH-HRPO Incubate at room temperature for 2 h Wash
Add substrate and develop at room temp for 15 min
FIG. 1. Assay procedure for the GHBP LIFA. Microtiter plates were coated with MAb 263 directed against the GHBP, blocked for 2 h, and washed, then assay sample was added with either 0 or 200 Mg/L rhGH and incubated for 2 h (• and ® represent endogenous hGH and rhGH, respectively). After a second wash step, HRPO-conjugated MAb MCB was added to detect the bound GH. After 2 h of incubation, the plates were washed, and substrate was added. The reaction was stopped after 15 min, and optical density at 490 nm was measured. mL 30% hydrogen peroxide) was added to the plates (100 well), and the incubation was carried out in the dark for 15 min at room temperature. The reaction was stopped by the addition of 100 ML 2.25 mol/L sulfuric acid and absorbance at 490 nm determined on a Vmax plate reader (Molecular Devices, Menlo Park, CA). A standard curve was generated by plotting absorbance vs. log of rhGHBP concentration, using a four-parameter nonlinear regression curve-fitting program (developed at Genentech). Sample concentrations were obtained by interpolation of their absorbance on the standard curve. Superose-12 fast protein liquid chromatography (FPLC) column assay Following a modification of the methods used by others (3, 4, 10, 11, 28, 30, 36), 200 nL reference (384 pmol/L rhGHBP diluted in PBS) or sample were incubated with 200 nL [125I] rhGH (0.6 ng) overnight at 4 C. The mixtures were subjected to gel chromatography in PBS on a 4.7 X 60-cm Superose-12 column (Pharmacia, Piscataway, NJ). As a control [125I]rhGH was incubated with PBS alone. To investigate the interference of hGH in this assay, 20 nL each of varying amounts of
Figure 2 shows two LIFA standard curves obtained by serial dilutions of rhGHBP and GHBP purified from human serum. As shown in the figure, the curves are parallel. At the concentration determined by Scatchard analysis, native GHBP gives a slightly higher signal than rhGHBP in the LIFA. Using the rhGHBP as standard, the assay range for the LIFA was determined to be 31-2000 pmol/L. The lower limit of detection (sensitivity) was 7.8 pmol/L, as defined by Rodbard et al. (21) to be equivalent to the concentration corresponding to the mean absorbance of zero plus twice the SD. Saturation of GHBP with rhGH and detection of bound hGH Multiple rhGHBP standard curves were assayed in the LIFA using increasing concentrations of rhGH to determine the optimal rhGH concentration for saturation of the rhGHBP-binding sites. This experiment showed that
CO
•e o
A
100
1000
10000
GHBP Concentration (pmol/L)
FIG. 2. Comparison of two LIFA standard curves generated by serial dilutions of GHBP purified from human serum (O—O) and rhGHBP
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
LIFA FOR GHBP
for 50 /xL sample, an equal volume of rhGH at different concentrations up to 200 Mg/L resulted in increased signal in the LIFA. However, increasing the rhGH concentration from 200 to 400 ixg/L did not further increase the absorbance obtained with the highest standard, and 200 Mgl/L is, therefore, used to saturate the GHBPbinding sites in the assay. The possible interference in the assay by endogenous GH complexed to the GHBP was then evaluated by comparing the ability of the LIFA to detect free GHBP vs. the GH/GHBP complex. Two standard curves were generated, of which one was incubated with rhGH (final concentration, 200 /ug/L) overnight at 4 C, and the other was incubated with assay buffer. The samples so generated were then assayed in the LIFA according to the standard protocol (i.e. all samples were incubated with additional rhGH on the microtiter plates.) As now shown in Fig. 3, similar results were obtained regardless of whether the rhGHBP was preincubated with rhGH, indicating that the GHBP measured by this assay should not be affected by the concentration of endogenous GH. The MCB, which was used for conjugation with HRPO, binds to the 22-kDa GH with high affinity, but has very low affinity for the 20-kDa GH (17). The possible interference by 20-kDa GH in the assay was tested. Incubation of the samples with 20-kDa rhGH (200 Mg/L) resulted in values indistinguishable from the blanks (Fig. 4). The addition of 200 /ig/L 20-kDa rhGH to the 200 Mg/L 22-kDa rhGH solution resulted in a standard curve similar to that obtained by incubation with 200 Mg/L 22-kDa rhGH alone. The 20-kDa rhGH does not interfere in this GHBP assay. Assay specificity The specificity of the LIFA was tested by substituting rhGHBP with four other soluble receptors (rhCD4, 2.51
2.01.51.0-
1219
2-
1 •
100
10
1000
10000
GHBP Concentration (pmol/L)
FIG. 4. Lack of interference of 20-kDa GH in the GHBP LIFA. Shown are standard curves of rhGHBP using 22-kDa rhGH (O—O), 20-kDa rhGH ( • — • ) , and an equimolar combination of 22- and 20-kDa rhGH (A—A) as ligand(s). TABLE 1. Cross-reactivity Cone.
Protein tested
Tested (mg/L)
Cross-reactivity Measured 4
(Mg/L)
(%Y
rHER2 ECD
Vol. 73, No. 6 Printed in U.S.A.
Ligand-Mediated Immunofunctional Assay for Quantitation of Growth Hormone-Binding Protein in Human Blood LENA M. S. CARLSSON*, ANN M. ROWLAND, ROSS G. CLARK, NEIL GESUNDHEIT, AND WAI LEE T. WONG Departments of Immunology Research and Assay Technologies (A.M.R., W.L.T. W.), Developmental Biology (L.M.S.C, R.G.C.), and Clinical Research (N.G.), Genentech, Inc., South San Francisco, California 94080
ABSTRACT. Human serum contains a high affinity GHbinding protein (GHBP) whose amino-terminal sequence is identical to the extracellular domain of the GH receptor. Current methods that measure GHBP are laborious, require size or charcoal separation of the GH/GHBP complex, and may be influenced by ambient GH concentrations. We have developed a novel assay method that allows quantitation of the total amount of functional GHBP in serum or plasma. The assay can also be used to measure the concentration of the circulating GH/ GHBP complex. An anti-GHBP monoclonal antibody, which recognizes both free GHBP and GH-bound GHBP, is used to capture the GHBP on a microtiter plate. Recombinant human GH is added to saturate all binding sites, and an anti-GH antibody conjugated with horseradish peroxidase is used to detect the amount of GH (endogenous and exogenous) bound to
T
HE PRESENCE of a protein in blood that binds to GH was first suggested in the 1960s (reviewed in Ref. 1). Reports on the conversion of GH to macromolecular form upon incubation with plasma were explained by the aggregation of radiolabeled GH, and the prevailing opinion was that GH circulated in the free form. Even though the serum GH-binding protein (GHBP) was characterized in the mouse in 1977 (2), it was not until 1986 that two groups reported the existence of a human GHBP (3, 4). There is now good evidence that the principal GHBP is identical to the extracellular domain of the hepatic GH receptor (5, 6) and that GHBP may be produced by proteolytic cleavage of the full-length GH receptor (7) or derive from a separate mRNA (8, 9). To understand hormonal mediators of growth, it is essential to be able to measure all parts of the axis (e.g. the hormone, its circulating binding protein, and the hormone-binding protein complex). Such measurements Received March 5,1991. Address requests for reprints to: Dr. Wai Lee T. Wong, Genentech, Inc., South San Francisco, California 94080. * Genentech Postdoctoral Fellow, supported in part by the Royal Swedish Academy of Sciences and the Swedish Medical Research Council.
the GHBP. The same procedure, but without incubation with GH, allows measurement of the endogenous GH/GHBP complex. The assay is sensitive (detection range, 31-2000 pmol/L), with average inter- and intraassay precisions of 11.3% and 7.3%, respectively. Measurements in random blood samples from 16 healthy adults showed that all subjects had clearly detectable GHBP concentrations (range, 65.8-305.6 pmol/L). In contrast, GHBP levels were undetectable in samples from 2 patients with Laron-type dwarfism. We believe that this ligand-mediated immunofunctional assay, which combines the simplicity and specificity of an enzyme-linked immunosorbent assay with the ability to detect only biochemically active binding protein, will be useful for studies of the role of the GHBP in health and disease. (J Clin Endocrinol Metab 73: 1216-1223, 1991)
can be difficult because of the interference in the assays by the different components and the fact that the hormone and the BP can be present in either free or bound (complexed) form. Currently, the standard method for quantitation of circulating proteins that bind to GH is incubation of serum or plasma with radiolabeled GH, followed by chromatography or charcoal separation of the bound from the free hormone (10-12). The procedures are cumbersome, and the results difficult to interpret because of the possible interference by endogenous GH in the sample (3). In addition, the proportions of free GHBP and GHBP bound with GH (complexed) cannot be determined with precision. The methods give an estimate of the binding capacity of serum proteins of a certain size, and the results are expressed in relation to reference serum pools, making it difficult to compare results from different studies. In our efforts to develop an assay for the GHBP, we chose a new approach. This report describes a very simple specific assay for quantitation of biochemically active GHBP in human serum and plasma. The ligand-mediated immunofunctional assay (LIFA) is based on the binding of GHBP to its ligand GH and the 1216
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
LIFA FOR GHBP
specific reactions between both the immobilized monoclonal antibody (MAb) to the GHBP and the enzymelabeled MAb to the GH bound to GHBP. The LIFA can measure both total and complexed GHBP and does not require prior removal of endogenous GH or physicochemical methods to separate free GH from the GH/GHBP complex. The assay, which does not involve the use of unstable radioactive chemicals, is simple, fast, and suitable for large scale clinical studies.
Materials and Methods Reagents Recombinant 22-kDa human GH (rhGH), 20-kDa rhGH, human placental lactogen (rhPL), human PRL (rhPRL), rhCD4, HIV envelope protein (rgpl20), HER2 extracellular domain (rHER2 ECD), human PRL receptor (rhPRL receptor), and human GHBP (rhGHBP) were produced at Genentech (South San Francisco, CA). The epidermal growth factor (EGF) receptor was purified from human epidermoid carcinoma A431 cell by T. Lipari (Genentech, Inc.), using previously described methods (13). Native GHBP was purified from human serum by S. Spencer (Genentech), essentially as described previously (6). rhGH was labeled with 125I using a modified chloramine-T procedure (14) with sequential addition of chloramine-T reagent to give [12BI]rhGH with a specific activity of 94 ^Ci//ig. Selection of coat MAb
Four MAbs to the rabbit liver GH receptor, MAbs 1, 2, 5, and 7, were provided by Dr. M. Waters (University of Queensland, Queensland, Australia), and an additional two MAbs (43 and 263, Agen, Australia), were purified from mouse ascites fluid using protein-A-Sepharose (Repligen Corp., Cambridge, MA) following established procedures (15, 16). Using radiolabeled hGH for detection, MAbs were evaluated to determine their binding sites on rhGHBP relative to rhGH (rhGHBP and [125I]rhGH added sequentially to the plate), their ability to bind to rhGHBP in the presence of [125I]rhGH (simultaneous incubation format), and their ability to bind to the rhGHBP/[125I] rhGH-complex (rhGHBP and [125I]rhGH preincubated and then added to the plate). MAb 263 was the most suitable MAb as coat, since it gave the highest percentage of bound [125I] rhGH in all three conditions and was able to bind the free rhGHBP as well as the rhGHBP bound to rhGH. The optimal coat concentration for the LIFA (MAb 263 at 10 Mg/mL) was determined as the lowest coat antibody concentration that resulted in maximal signal. Enzyme-conjugated anti-hGH antibodies The anti-GH MAb (MCB), which was provided by B. Fendly (Genentech), was selected for conjugation because it contains no overlapping determinants with the GHBP-binding site (17). The antibodies were purified from ascites fluid using proteinA-Sepharose (Repligen Corp.) following established procedures (15,16) and stored sterile in PBS at 4 C. A polyclonal anti-GH was purified from goat serum using recombinant protein-G-
1217
Sepharose (Zymed, South San Francisco, CA) (18). Purified antibodies were conjugated to horseradish peroxidase (HRPO) (19) and stored at -20 C in 50% glycerol. LIFA assay standards and samples The rhGHBP was purified from a mammalian cell line by P. McKay (Genentech) following the procedure outlined by Spencer et al. (6). The purified rhGHBP amino acid composition, as determined by quantitative amino acid analysis, matched what was theoretically expected for the cloned gene product. The purified rhGHBP was homogeneous based on analysis of sodium dodecyl sulfate-gel. The concentrations of biochemically active rhGHBP and native GHBP in the purified preparations were established by Scatchard analysis (20). The association constants (Ka) were 2.9 X 109 and 3.3 X 109 M"1 for the native and recombinant GHBP, respectively. When appropriate, the concentration of rhGHBP was converted to micrograms per L using 26 kDa as the mol wt of the protein part of the molecule. Dilutions of rhGHBP in PBS containing per liter, 5 g BSA, 5.0 mM EDTA, 0.5 mL Tween-20, and 0.1 g Thimerosal (assay buffer) were used as standards in the LIFA. Serum and plasma (EDTA, citrate, and heparin) samples were obtained from healthy adult volunteers taking no medications, except for subject 3, who was receiving fluoxetine. All subjects, except no. 1 and 14, were nonfasting, and the samples were taken between 0900-1100 h. The samples were centrifuged and stored at -70 or -20 C until assayed. Sera from two unrelated patients with Laron-type dwarfism were supplied by Dr. L. Underwood (University of North Carolina, Chapel Hill, NC) and Dr. R. Rosenfeld (Stanford University, Stanford, CA). LIFA assay procedure The assay procedure is outlined in Fig. 1. Ninety-six-well microtiter plates (Corning Glass Works, Corning, NY) were coated with MAb 263 by incubating overnight at 4 C with 100 /uL/well antibody at 10 Mg/mL in 50 mmol/L carbonate buffer, pH 9.6. After removal of the coating solution, nonspecific binding sites on the coated wells were blocked with 150 nL phosphate-buffered isotonic saline pH 7.2 (PBS), containing 5 g BSA (98-99% pure; Sigma Chemical Co., St. Louis, MO) per L for 1 h at room temperature, followed by six washes at room temperature with wash buffer (0.5 mL Tween-20 and 0.1 g Thimerosal/L PBS). Standards or samples (50 jtL/well) were dispensed into the coated wells containing 50 juL/well 0 ixg/L (for measurement of GH/GHBP complex concentration) or 200 jug/L rhGH (for measurement of total GHBP concentration) in sample buffer (assay buffer with 10 g/L mouse immunoglobulin G (IgG) from Fitzgerald Industries, Chelmsford, MA). Plates were sealed, incubated at room temperature for 2 h with gentle agitation, then washed before the addition of HRPO-labeled MAb MCB (100 juL/well). In the experiment in which the interference by heterophilic antibodies was determined, a HRPO-labeled polyclonal goat anti-rhGH antibody was used. After further incubation for 2 h at room temperature, the plates were washed six times with wash buffer. Freshly prepared substrate solution (0.4 g o-phenylenediamine dihydrochloride in 1 L PBS plus 0.4
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
CARLSSON ET AL.
1218
Assay Procedure
MAb263
Coat plates with MAb 263, 4°C overnight
JCE & M • 1991 Vol73-No6
unlabeled rhGH (to give final concentrations of 0, 6.7, 20, and 60 /ng/L) were incubated with a mixture of sample and [125I] rhGH. The radioactivity in the eluates was measured with an on-line 7-detection system. The radioactivity elution profile of the reference showed two main peaks: peak I, representing the [125I] hGH-GHBP complex, and peak II, representing free [125I] hGH. GHPB activity was determined as the area under peak I, expressed as a percentage of the sum of the areas of peak I and II.
Block at room temperature for 1 h rhGH
rhGH-GHBP complex
hGH-GHBP complex
Measurement of serum hGH concentrations Wash
Add sample and rhGH (0or200ng/ml) Incubate at room temperature for 2 h
Serum hGH concentrations were determined in duplicate by a polyclonal antibody-based enzyme-linked immunosorbent assay (ELISA) developed by A. Celniker, Genentech. Results
I Wash
Assay range and sensitivity anti-hGH-HRPO antibody
Add anti-hGH-HRPO Incubate at room temperature for 2 h Wash
Add substrate and develop at room temp for 15 min
FIG. 1. Assay procedure for the GHBP LIFA. Microtiter plates were coated with MAb 263 directed against the GHBP, blocked for 2 h, and washed, then assay sample was added with either 0 or 200 Mg/L rhGH and incubated for 2 h (• and ® represent endogenous hGH and rhGH, respectively). After a second wash step, HRPO-conjugated MAb MCB was added to detect the bound GH. After 2 h of incubation, the plates were washed, and substrate was added. The reaction was stopped after 15 min, and optical density at 490 nm was measured. mL 30% hydrogen peroxide) was added to the plates (100 well), and the incubation was carried out in the dark for 15 min at room temperature. The reaction was stopped by the addition of 100 ML 2.25 mol/L sulfuric acid and absorbance at 490 nm determined on a Vmax plate reader (Molecular Devices, Menlo Park, CA). A standard curve was generated by plotting absorbance vs. log of rhGHBP concentration, using a four-parameter nonlinear regression curve-fitting program (developed at Genentech). Sample concentrations were obtained by interpolation of their absorbance on the standard curve. Superose-12 fast protein liquid chromatography (FPLC) column assay Following a modification of the methods used by others (3, 4, 10, 11, 28, 30, 36), 200 nL reference (384 pmol/L rhGHBP diluted in PBS) or sample were incubated with 200 nL [125I] rhGH (0.6 ng) overnight at 4 C. The mixtures were subjected to gel chromatography in PBS on a 4.7 X 60-cm Superose-12 column (Pharmacia, Piscataway, NJ). As a control [125I]rhGH was incubated with PBS alone. To investigate the interference of hGH in this assay, 20 nL each of varying amounts of
Figure 2 shows two LIFA standard curves obtained by serial dilutions of rhGHBP and GHBP purified from human serum. As shown in the figure, the curves are parallel. At the concentration determined by Scatchard analysis, native GHBP gives a slightly higher signal than rhGHBP in the LIFA. Using the rhGHBP as standard, the assay range for the LIFA was determined to be 31-2000 pmol/L. The lower limit of detection (sensitivity) was 7.8 pmol/L, as defined by Rodbard et al. (21) to be equivalent to the concentration corresponding to the mean absorbance of zero plus twice the SD. Saturation of GHBP with rhGH and detection of bound hGH Multiple rhGHBP standard curves were assayed in the LIFA using increasing concentrations of rhGH to determine the optimal rhGH concentration for saturation of the rhGHBP-binding sites. This experiment showed that
CO
•e o
A
100
1000
10000
GHBP Concentration (pmol/L)
FIG. 2. Comparison of two LIFA standard curves generated by serial dilutions of GHBP purified from human serum (O—O) and rhGHBP
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
LIFA FOR GHBP
for 50 /xL sample, an equal volume of rhGH at different concentrations up to 200 Mg/L resulted in increased signal in the LIFA. However, increasing the rhGH concentration from 200 to 400 ixg/L did not further increase the absorbance obtained with the highest standard, and 200 Mgl/L is, therefore, used to saturate the GHBPbinding sites in the assay. The possible interference in the assay by endogenous GH complexed to the GHBP was then evaluated by comparing the ability of the LIFA to detect free GHBP vs. the GH/GHBP complex. Two standard curves were generated, of which one was incubated with rhGH (final concentration, 200 /ug/L) overnight at 4 C, and the other was incubated with assay buffer. The samples so generated were then assayed in the LIFA according to the standard protocol (i.e. all samples were incubated with additional rhGH on the microtiter plates.) As now shown in Fig. 3, similar results were obtained regardless of whether the rhGHBP was preincubated with rhGH, indicating that the GHBP measured by this assay should not be affected by the concentration of endogenous GH. The MCB, which was used for conjugation with HRPO, binds to the 22-kDa GH with high affinity, but has very low affinity for the 20-kDa GH (17). The possible interference by 20-kDa GH in the assay was tested. Incubation of the samples with 20-kDa rhGH (200 Mg/L) resulted in values indistinguishable from the blanks (Fig. 4). The addition of 200 /ig/L 20-kDa rhGH to the 200 Mg/L 22-kDa rhGH solution resulted in a standard curve similar to that obtained by incubation with 200 Mg/L 22-kDa rhGH alone. The 20-kDa rhGH does not interfere in this GHBP assay. Assay specificity The specificity of the LIFA was tested by substituting rhGHBP with four other soluble receptors (rhCD4, 2.51
2.01.51.0-
1219
2-
1 •
100
10
1000
10000
GHBP Concentration (pmol/L)
FIG. 4. Lack of interference of 20-kDa GH in the GHBP LIFA. Shown are standard curves of rhGHBP using 22-kDa rhGH (O—O), 20-kDa rhGH ( • — • ) , and an equimolar combination of 22- and 20-kDa rhGH (A—A) as ligand(s). TABLE 1. Cross-reactivity Cone.
Protein tested
Tested (mg/L)
Cross-reactivity Measured 4
(Mg/L)
(%Y
rHER2 ECD
