002l-972X/91/7202-0387$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright© 1991 by The Endocrine Society
Vol. 72, No. 2 Printed in U.S.A.
Effects of Growth Hormone-Binding Proteins on Serum Growth Hormone Measurements* TAUHID JAN, MELISSA A. SHAW, AND GERHARD BAUMANN Center for Endocrinology, Metabolism, and Nutrition, Department of Medicine, Northwestern University Medical School, Chicago, Illinois 60611
ation of the tracer with the solid phase reagent. However, these effects were minor at the BP concentrations realistically encountered in practice. Furthermore, the impact of BPs on GH standard curves and final results was negligible because BP effects fall into the category of nonspecific or zero dose counts, which are subtracted during data reduction. We conclude that GH-BPs are only a minor disturbance in GH immunoassays, which is completely compensated for by conventional assay design. Disparities among results yielded by different assays are probably not due to BP interference. (J Clin Endocrinol Metab 72:387-391, 1991)
ABSTRACT. GH-binding proteins (GH-BPs) in human blood theoretically may interfere with measurements of immunoreactive GH by forming complexes with GH and competing with antibody reagents for ligand. Indeed, results of serum GH obtained by immunoassays are known to differ markedly depending on the assay employed. To assess the potential role of circulating GH-BPs in this phenomenon, we systematically examined their effect on GH measurement in four RIAs and two immunoradiometric assays. In all except one RIA, the effect of the BPs on tracer binding to antibody was mildly inhibitory. In both immunoradiometric assays, BPs increased the nonspecific associ-
T
systematically examined this possibility using a variety of GH immunoassays.
WO CIRCULATING GH-binding proteins (GHBPs) have recently been discovered in human plasma/serum (1-3). The main, high affinity GH-BP corresponds to the extracellular portion of the hepatic GH receptor (4-6). The existence of these GH-BPs raises new questions about their possible influence on GH determination in serum. A substantial part of serum GH is complexed with the BPs (7, 8), and little is known about the degree of recognition of complexed GH by different anti-GH antibodies. In addition, GH-BPs can be expected to compete with antibodies for GH tracer during the incubation. Variability of serum GH results obtained in different immunoassays is well recognized (9-14). This phenomenon is of considerable clinical concern because the diagnosis, and consequent therapy, of GH deficiency rests in large part on the attainment of a certain level of serum GH either after provocative stimuli or during spontaneous pulses (15). The reasons for these disparities in GH measurements are not well understood. Varying degrees of interference by BPs in different assays is one possible explanation. In the present study we
Materials and Methods Materials A partially purified GH-BP preparation was prepared by affinity chromatography from human plasma, as previously described (1). It contained both BPs in a proportion comparable to that of whole plasma (3), represented an approximately 8000fold purification from plasma [SA, 110 U/mg protein, where 1 U is the activity contained in 1 mL of an adult plasma pool (16)] and was only minimally contaminated with GH (0.29 ng GH/U BP). Six different immunoassay systems for GH were examined [four polyclonal RIAs and two monoclonal immunoradiometric assays (IRMAs)]. Reagents for our in-house RIA have been previously described (17). Reagents for GH RIA were kindly provided by the National Hormone and Pituitary Program, NIH, and the University of Maryland (NIH RIA). Commercially available reagents for GH RIA were obtained from Cambridge Medical Diagnostics (Billerica, MA) and Incstar (Stillwater, MN); those for GH IRMA from Hybritech (Tandem-R-HGH IRMA, San Diego, CA) and Nichols Institute (Allegro HGH IRMA, San Juan Capistrano, CA).1 The RIAs employ an anti-GH antiserum, radiolabeled GH, and unlabeled GH in standard competitive binding RIA design, with a secondary immunoprecipitation step. The IRMAs employ two monoclonal anti-GH antibodies, one of which is radiolabeled
Received July 25, 1990. Address all correspondence and requests for reprints to: G. Baumann, M.D., 303 East Chicago Avenue, Chicago, Illinois 60611. * This work was presented in part at the 72nd Annual Meeting of The Endocrine Society, Atlanta, GA, June 1990, and reported in abstract form (Abstract 898). It was supported by NIH Grants DK38128 and RR-05370 and a grant from the Northwestern Memorial Foundation.
1 The Nichols IRMA employed human serum as a matrix for the preparation of the standard curve, rather than the previously used equine serum.
387
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
JAN, SHAW, AND BAUMANN
388
and the other linked to a solid phase (bead) by either direct coating or a biotin-avidin linkage. The two antibodies are directed to different epitopes on GH, thus permitting a sandwich-type radiolabeled complex to form in the presence of GH. In contrast to the RIA, where radioactivity in the immunoprecipitate is inversely related to GH concentration, in the IRMA there is a direct relationship between solid phase-associated radioactivity and GH concentration. Assays were performed according to the instructions of the suppliers; the conditions of the in-house RIA have been previously reported (17). GH was preincubated with BP for 45 min at 37 C to assure formation of a complex before adding the immunoassay reagents. Final volumes for the primary incubation step were 1 mL for the in-house and NIH RIAs, 0.5 mL for Incstar and Cambridge RIAs, and 0.2 mL for the IRMAs. The maximal serum volume that can be assayed without significant nonspecific interference is 0.2 mL in the in-house and NIH RIA and 0.1 mL in the other RIAs. The serum volume assayed in the IRMAs is 0.1 mL.
150n
JCE & M • 1991 Vol 72 • No 2
IN-HOUSE RIA
1OO-
50-
150-1
100
50-
BP ( U/tube )
Experimental design The effect of GH-BP on immunoassays was tested in two different ways. In a first investigation (part A), varying concentrations of BP were incubated with immunoassay reagents in the absence of unlabeled GH. In RIAs, this procedure directly tested the effect of BP on GH tracer binding to antibody [i.e. Bmax or Bo], whereas in IRMAs, it tested the effect of BP on zero dose binding or nonspecific binding (i.e. the association of the labeled antibody with the solid phase in the absence of GH). In a second type of investigation (part B), the effect of BP on GH standard curves in these assays was determined. It should be noted that part B differed from part A in that standard curves were typically corrected for nonspecific binding. In addition, in RIAs the standard curve was expressed in terms of binding at zero dose (tracer only; Bo). Within each individual standard curve, the BP concentration was held constant at all GH dose levels, but several BP concentrations were tested in parallel. From the data in part B, the impact of BP on GH measurements in unknown samples can be directly derived.
FIG. 1. Effect of GH-BPs on GH tracer binding to anti-GH antibody in four RIAs. Points represent the means of two to four experiments. B/Bo denotes the fraction of bound tracer relative to that in the absence of BP. The maximal volume of serum that would be assayed is 0.2 mL in the in-house and NIH RIAs and 0.1 mL in the other RIAs (corresponding to 0.1 or 0.2 U BP, respectively). 2000-1 HYBRITECH IRMA
1000-
2OOOO-. NICHOLS IRMA
10000" Data analysis Results from two to four experiments were pooled and are expressed as average values.
Results Figure 1 shows the effect of GH-BP on GH tracer binding to antibody in RIAs (part A). An effect was seen in all four RIAs. In all but one RIA (the in-house assay), the effect was inhibitory. However, at the BP doses realistically encountered in a RIA (0.1-0.2 U), these effects were relatively minor. Figure 2 shows the corresponding data for the IRMAs. Here, GH-BP increased the association of the tracer (labeled anti-GH antibody) with the solid phase reagent
0
0
.2
.4
.6
.8
1.0
BP(U/tube)
FIG. 2. Effect of GH-BPs on tracer binding to solid phase reagent in two IRMAs. The serum volume that would be assayed is 0.1 mL (corresponding to 0.1 U BP). Presentation and abbreviations are explained in Fig. 1.
in a dose-dependent manner. Again, at the dose present in a real serum assay (0.1 U), the effect was relatively minor, particularly in the Hybritech assay. The effect of GH-BP on GH standard curves in RIAs (part B) is shown in Fig. 3. BP in the range up to 0.2 U
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
GH-BP EFFECTS IN GH IMMUNOASSAYS 1 0 O i IN-HOUSE RIA
389 -i
NIH RIA D OU/tubeBP • .05U/tubeBP o ,10U/tubeBP .20U/tubeBP
75-
50-
25-
FIG. 3. Effects of GH-BPs on GH standard curves in four RIAs. Presentation and abbreviations are explained in Fig. 1.
1001
75-
50-
25-
.01
.1
100
.01
1O
1OO
hGH(ng/ml)
(corresponding to 0.2 mL serum) had no discernible effect on standard curves. Thus, the small effect on GH binding to antibody (Fig. 1) was of little consequence in the final results of the assay. The reason for this lies in the fact that correction of raw counts for nonspecific binding and expression relative to zero dose counts (B/ Bo) eliminates BP effects from the final calculation of results. Corresponding standard curves for the IRMAs are shown in Fig. 4. Here again, no significant effect of BP on the curves was seen over the range of GH and BP tested. The reason is analogous to that stated above, i.e. raw data are corrected for zero dose data or nonspecific counts, which essentially eliminates BP effects.
Discussion This study demonstrates that GH-BPs in human plasma can affect the raw data obtained in immunoassays for GH. It also shows that despite that, the final results rendered by such assays are not noticeably influenced by
the presence of BPs, at least for the six immunoassays tested. Although these findings cannot necessarily be extrapolated to other immunoassay systems, we believe that we tested a representative spectrum of assays, some of them in common use. It, therefore, appears unlikely that BP interference is the cause of different GH results obtained by different assays. A more likely explanation for that phenomenon is the varying recognition of different molecular forms of GH by different antibodies (see Ref. 18 for review). The relative freedom of immunoassays from interference by BPs contrasts with receptor assays, where the high affinity BP has a major effect on results (19). The reasons for the lack of substantial interference by BPs in immunoassays are several. In the case of RIAs, the affinity of anti-GH antibodies typically used is at least 2 orders of magnitude higher than that of the high affinity BP and 4-5 orders of magnitude higher than the low affinity BP (1-3). Thus, any GH associated with BPs will tend to dissociate and combine with the antibody during incubation. BPs, therefore, have only a small
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
JAN, SHAW, AND BAUMANN
390 100
i HYBRITECH IRMA
10 i
D OU/tubeBP • O5U/tubeBP O.1OU/tube BP • .20 U/tube BP
1 --.
o x 2
10
100
100 i NICHOLS IRMA
10-=
1
•:
.1
1
10
100
hGH(ng/ml) FIG. 4. Effects of GH-BPs on GH standard curves in two IRMAs. Presentation and abbreviations are explained in Fig. 1.
inhibitory effect on antibody binding. In addition, GH may retain full immunoreactivity even when bound to the BPs, as is the case with our in-house antiserum (20). This property may explain the lack of inhibition with this antiserum (although it does not explain the phenomenon of enhanced binding). Another reason for noninterference is the small volume of serum used in immunoassays, which corresponds to 0.1 or 0.2 U BP at most. At these BP concentrations, the effect on GH binding to antibody is minimized. The principal reason for noninterference, however, is the manner in which RIA data are expressed. Expression of bound counts as a percentage or fraction of counts bound at zero dose (B/Bo) as well as subtracting nonspecific counts automatically corrects for factors extraneous to GH itself, including BPs. In the case of IRMAs, the reasons for an increase in bound counts effected by BPs are not entirely clear. Both IRMAs tested employ a two-site anti-GH monoclonal antibody design. The radiolabeled antibody associates with the solid phase antibody by forming a ternary complex, with GH as the intermediate. The BP preparation promoted association of the labeled antibody with the solid phase (bead) through an unknown mechanism. Although the BP we used contained a small amount of GH (0.29 ng GH/U BP), which is derived from the affinity column used for purification, that amount is far
JCE&M«1991 Vol 72 • No 2
too low to explain the effect seen. Since the BP preparation was only partially pure, it contained a variety of other serum proteins that may have promoted nonspecific association of the labeled antibody with the polystyrene bead through a matrix effect. This effect was again relatively minor at the BP dose realistically encountered (0.1 U). Furthermore, it was corrected for in the final data calculation. Thus, in IRMAs, as in RIAs, the presence of GH-BPs appears to have little practical significance. In summary, we have shown BPs do not substantially affect GH immunoassay results. Although BPs have minor effects on antibody binding in RIAs or other nonspecific matrix effects in IRMAs, conventional assay designs minimize or eliminate their impact on final results. This can be further minimized by preparing immunoassay standard curves in GH-free or -poor serum/ plasma at a concentration ::epresentative of that used for the unknown samples. Many commercial assays already incorporate this design. Based on the present study, BP interference is not a likely cause of the discrepancies among GH results yielded by various immunoassays.
References 1. Baumann G, Stolar MW, Aniburn K, Barsano CP, DeVries BC. A specific growth hormone-binding protein in human plasma: initial characterization. J Clin Endocrinol Metab. 1986;62:134-41. 2. Herington AC, Ymer S, Stevsnson J. Identification and characterization of specific binding proteins for growth hormone in normal human sera. J Clin Invest. D86;77:1817-23. 3. Baumann G, Shaw MA. A second, lower affinity growth hormonebinding protein in human plasma. J Clin Endocrinol Metab. 1990;70:680-6. 4. Leung DW, Spencer SA, Cachianes G, et al. Growth hormone receptor and serum binding protein: purification, cloning and expression. Nature. 1987;330:537-43. 5. Baumann G, Shaw MA. Immunochemical similarity of the human plasma growth hormone-binding protein and the rabbit liver growth hormone receptor. Biochem Biophys Res Commun. 1988;152:573-8. 6. Barnard R, Quirk P, Waters MJ. Characterization of the growth hormone-binding protein of human serum using a panel of monoclonal antibodies. J Endocrinol. 1989;123:327-32. 7. Baumann G, Amburn K, Shaw MA. The circulating growth hormone (GH)-binding protein complex: a major constituent of plasma GH in man. Endocrinology. 1988;122:976-84. 8. Baumann G, Vance ML, Shaw MA, Thorner M. Plasma transport of human growth hormono in vivo. J Clin Endocrinol Metab. 1990;71:470-3. 9. Blethen SL, Chasalow FI. Use of a two-site immunoradiometric assay for growth hormone (GH) in identifying children with GHdependent growth failure. J Clin Endocrinol Metab. 1983;57:10315. 10. Rudman D, Chawla R, Ku;ner M. Heterogeneity of growth hormone in the nocturnal serum of children. Pediatr Res. 1985,19:9815. 11. Levin PA, Chalew SA, Martin L, Kowarski AA. Comparison of assays for growth hormone using monoclonal or polyclonal antibodies for diagnosis of growth failure. J Lab Clin Med. 1987; 109:858. 12. Reiter EO, Morris AH, MacGillivray MH, Weber D. Variable estimates of serum growth hormone concentrations by different radioassay systems. J Clin Endocrinol Metab. 1988;66:68-71.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
GH-BP EFFECTS IN GH IMMUNOASSAYS 13. Celniker AC, Chen AB, Wert Jr RM, Sherman BM. Variability in the quantitation of circulating growth hormone using commercial immunoassays. J Clin Endocrinol Metab. 1989;68:469-76. 14. Felder RA, Holl RW, Martha Jr P, et al. Influence of matrix on concentrations of somatotropin measured in serum with commercial immunoradiometric assays. Clin Chem. l989;35:1423-6. 15. Bierich JR, Briigmann G, Kiessling E. Use of provocative tests and measurement of spontaneous GH-secretion to diagnose GH deficiency. In: Miiller EE, Cocchi D, Locatelli V, eds. Advances in growth hormone and growth factor research. Rome/Berlin: Pythagora Press/Springer-Verlag; 1989;299-320. 16. Baumann G, Shaw MA, Amburn K. Regulation of plasma growth hormone-binding proteins in health and disease. Metabolism.
391
1989;38:683-9. 17. Drobny EC, Amburn K, Baumann G. Circadian variation of basal plasma growth hormone in man. J Clin Endocrinol Metab. 1983;57:524-8. 18. Baumann G. Growth hormone binding proteins and various forms of growth hormone: implications for measurement. Acta Paediatr Scand. In Press. 19. Mannor DA, Shaw MA, Winer LM, Baumann G. Circulating growth hormone-binding proteins inhibit growth hormone (GH) binding to GH receptors but not in vivo GH action [Abstract]. Clin Res. 1988;36:870A. 20. Baumann G, Shaw MA, Buchanan TA. In vivo kinetics of a covalent growth hormone-binding protein complex. Metabolism. l988;38:330-3.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
Vol. 72, No. 2 Printed in U.S.A.
Effects of Growth Hormone-Binding Proteins on Serum Growth Hormone Measurements* TAUHID JAN, MELISSA A. SHAW, AND GERHARD BAUMANN Center for Endocrinology, Metabolism, and Nutrition, Department of Medicine, Northwestern University Medical School, Chicago, Illinois 60611
ation of the tracer with the solid phase reagent. However, these effects were minor at the BP concentrations realistically encountered in practice. Furthermore, the impact of BPs on GH standard curves and final results was negligible because BP effects fall into the category of nonspecific or zero dose counts, which are subtracted during data reduction. We conclude that GH-BPs are only a minor disturbance in GH immunoassays, which is completely compensated for by conventional assay design. Disparities among results yielded by different assays are probably not due to BP interference. (J Clin Endocrinol Metab 72:387-391, 1991)
ABSTRACT. GH-binding proteins (GH-BPs) in human blood theoretically may interfere with measurements of immunoreactive GH by forming complexes with GH and competing with antibody reagents for ligand. Indeed, results of serum GH obtained by immunoassays are known to differ markedly depending on the assay employed. To assess the potential role of circulating GH-BPs in this phenomenon, we systematically examined their effect on GH measurement in four RIAs and two immunoradiometric assays. In all except one RIA, the effect of the BPs on tracer binding to antibody was mildly inhibitory. In both immunoradiometric assays, BPs increased the nonspecific associ-
T
systematically examined this possibility using a variety of GH immunoassays.
WO CIRCULATING GH-binding proteins (GHBPs) have recently been discovered in human plasma/serum (1-3). The main, high affinity GH-BP corresponds to the extracellular portion of the hepatic GH receptor (4-6). The existence of these GH-BPs raises new questions about their possible influence on GH determination in serum. A substantial part of serum GH is complexed with the BPs (7, 8), and little is known about the degree of recognition of complexed GH by different anti-GH antibodies. In addition, GH-BPs can be expected to compete with antibodies for GH tracer during the incubation. Variability of serum GH results obtained in different immunoassays is well recognized (9-14). This phenomenon is of considerable clinical concern because the diagnosis, and consequent therapy, of GH deficiency rests in large part on the attainment of a certain level of serum GH either after provocative stimuli or during spontaneous pulses (15). The reasons for these disparities in GH measurements are not well understood. Varying degrees of interference by BPs in different assays is one possible explanation. In the present study we
Materials and Methods Materials A partially purified GH-BP preparation was prepared by affinity chromatography from human plasma, as previously described (1). It contained both BPs in a proportion comparable to that of whole plasma (3), represented an approximately 8000fold purification from plasma [SA, 110 U/mg protein, where 1 U is the activity contained in 1 mL of an adult plasma pool (16)] and was only minimally contaminated with GH (0.29 ng GH/U BP). Six different immunoassay systems for GH were examined [four polyclonal RIAs and two monoclonal immunoradiometric assays (IRMAs)]. Reagents for our in-house RIA have been previously described (17). Reagents for GH RIA were kindly provided by the National Hormone and Pituitary Program, NIH, and the University of Maryland (NIH RIA). Commercially available reagents for GH RIA were obtained from Cambridge Medical Diagnostics (Billerica, MA) and Incstar (Stillwater, MN); those for GH IRMA from Hybritech (Tandem-R-HGH IRMA, San Diego, CA) and Nichols Institute (Allegro HGH IRMA, San Juan Capistrano, CA).1 The RIAs employ an anti-GH antiserum, radiolabeled GH, and unlabeled GH in standard competitive binding RIA design, with a secondary immunoprecipitation step. The IRMAs employ two monoclonal anti-GH antibodies, one of which is radiolabeled
Received July 25, 1990. Address all correspondence and requests for reprints to: G. Baumann, M.D., 303 East Chicago Avenue, Chicago, Illinois 60611. * This work was presented in part at the 72nd Annual Meeting of The Endocrine Society, Atlanta, GA, June 1990, and reported in abstract form (Abstract 898). It was supported by NIH Grants DK38128 and RR-05370 and a grant from the Northwestern Memorial Foundation.
1 The Nichols IRMA employed human serum as a matrix for the preparation of the standard curve, rather than the previously used equine serum.
387
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
JAN, SHAW, AND BAUMANN
388
and the other linked to a solid phase (bead) by either direct coating or a biotin-avidin linkage. The two antibodies are directed to different epitopes on GH, thus permitting a sandwich-type radiolabeled complex to form in the presence of GH. In contrast to the RIA, where radioactivity in the immunoprecipitate is inversely related to GH concentration, in the IRMA there is a direct relationship between solid phase-associated radioactivity and GH concentration. Assays were performed according to the instructions of the suppliers; the conditions of the in-house RIA have been previously reported (17). GH was preincubated with BP for 45 min at 37 C to assure formation of a complex before adding the immunoassay reagents. Final volumes for the primary incubation step were 1 mL for the in-house and NIH RIAs, 0.5 mL for Incstar and Cambridge RIAs, and 0.2 mL for the IRMAs. The maximal serum volume that can be assayed without significant nonspecific interference is 0.2 mL in the in-house and NIH RIA and 0.1 mL in the other RIAs. The serum volume assayed in the IRMAs is 0.1 mL.
150n
JCE & M • 1991 Vol 72 • No 2
IN-HOUSE RIA
1OO-
50-
150-1
100
50-
BP ( U/tube )
Experimental design The effect of GH-BP on immunoassays was tested in two different ways. In a first investigation (part A), varying concentrations of BP were incubated with immunoassay reagents in the absence of unlabeled GH. In RIAs, this procedure directly tested the effect of BP on GH tracer binding to antibody [i.e. Bmax or Bo], whereas in IRMAs, it tested the effect of BP on zero dose binding or nonspecific binding (i.e. the association of the labeled antibody with the solid phase in the absence of GH). In a second type of investigation (part B), the effect of BP on GH standard curves in these assays was determined. It should be noted that part B differed from part A in that standard curves were typically corrected for nonspecific binding. In addition, in RIAs the standard curve was expressed in terms of binding at zero dose (tracer only; Bo). Within each individual standard curve, the BP concentration was held constant at all GH dose levels, but several BP concentrations were tested in parallel. From the data in part B, the impact of BP on GH measurements in unknown samples can be directly derived.
FIG. 1. Effect of GH-BPs on GH tracer binding to anti-GH antibody in four RIAs. Points represent the means of two to four experiments. B/Bo denotes the fraction of bound tracer relative to that in the absence of BP. The maximal volume of serum that would be assayed is 0.2 mL in the in-house and NIH RIAs and 0.1 mL in the other RIAs (corresponding to 0.1 or 0.2 U BP, respectively). 2000-1 HYBRITECH IRMA
1000-
2OOOO-. NICHOLS IRMA
10000" Data analysis Results from two to four experiments were pooled and are expressed as average values.
Results Figure 1 shows the effect of GH-BP on GH tracer binding to antibody in RIAs (part A). An effect was seen in all four RIAs. In all but one RIA (the in-house assay), the effect was inhibitory. However, at the BP doses realistically encountered in a RIA (0.1-0.2 U), these effects were relatively minor. Figure 2 shows the corresponding data for the IRMAs. Here, GH-BP increased the association of the tracer (labeled anti-GH antibody) with the solid phase reagent
0
0
.2
.4
.6
.8
1.0
BP(U/tube)
FIG. 2. Effect of GH-BPs on tracer binding to solid phase reagent in two IRMAs. The serum volume that would be assayed is 0.1 mL (corresponding to 0.1 U BP). Presentation and abbreviations are explained in Fig. 1.
in a dose-dependent manner. Again, at the dose present in a real serum assay (0.1 U), the effect was relatively minor, particularly in the Hybritech assay. The effect of GH-BP on GH standard curves in RIAs (part B) is shown in Fig. 3. BP in the range up to 0.2 U
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
GH-BP EFFECTS IN GH IMMUNOASSAYS 1 0 O i IN-HOUSE RIA
389 -i
NIH RIA D OU/tubeBP • .05U/tubeBP o ,10U/tubeBP .20U/tubeBP
75-
50-
25-
FIG. 3. Effects of GH-BPs on GH standard curves in four RIAs. Presentation and abbreviations are explained in Fig. 1.
1001
75-
50-
25-
.01
.1
100
.01
1O
1OO
hGH(ng/ml)
(corresponding to 0.2 mL serum) had no discernible effect on standard curves. Thus, the small effect on GH binding to antibody (Fig. 1) was of little consequence in the final results of the assay. The reason for this lies in the fact that correction of raw counts for nonspecific binding and expression relative to zero dose counts (B/ Bo) eliminates BP effects from the final calculation of results. Corresponding standard curves for the IRMAs are shown in Fig. 4. Here again, no significant effect of BP on the curves was seen over the range of GH and BP tested. The reason is analogous to that stated above, i.e. raw data are corrected for zero dose data or nonspecific counts, which essentially eliminates BP effects.
Discussion This study demonstrates that GH-BPs in human plasma can affect the raw data obtained in immunoassays for GH. It also shows that despite that, the final results rendered by such assays are not noticeably influenced by
the presence of BPs, at least for the six immunoassays tested. Although these findings cannot necessarily be extrapolated to other immunoassay systems, we believe that we tested a representative spectrum of assays, some of them in common use. It, therefore, appears unlikely that BP interference is the cause of different GH results obtained by different assays. A more likely explanation for that phenomenon is the varying recognition of different molecular forms of GH by different antibodies (see Ref. 18 for review). The relative freedom of immunoassays from interference by BPs contrasts with receptor assays, where the high affinity BP has a major effect on results (19). The reasons for the lack of substantial interference by BPs in immunoassays are several. In the case of RIAs, the affinity of anti-GH antibodies typically used is at least 2 orders of magnitude higher than that of the high affinity BP and 4-5 orders of magnitude higher than the low affinity BP (1-3). Thus, any GH associated with BPs will tend to dissociate and combine with the antibody during incubation. BPs, therefore, have only a small
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
JAN, SHAW, AND BAUMANN
390 100
i HYBRITECH IRMA
10 i
D OU/tubeBP • O5U/tubeBP O.1OU/tube BP • .20 U/tube BP
1 --.
o x 2
10
100
100 i NICHOLS IRMA
10-=
1
•:
.1
1
10
100
hGH(ng/ml) FIG. 4. Effects of GH-BPs on GH standard curves in two IRMAs. Presentation and abbreviations are explained in Fig. 1.
inhibitory effect on antibody binding. In addition, GH may retain full immunoreactivity even when bound to the BPs, as is the case with our in-house antiserum (20). This property may explain the lack of inhibition with this antiserum (although it does not explain the phenomenon of enhanced binding). Another reason for noninterference is the small volume of serum used in immunoassays, which corresponds to 0.1 or 0.2 U BP at most. At these BP concentrations, the effect on GH binding to antibody is minimized. The principal reason for noninterference, however, is the manner in which RIA data are expressed. Expression of bound counts as a percentage or fraction of counts bound at zero dose (B/Bo) as well as subtracting nonspecific counts automatically corrects for factors extraneous to GH itself, including BPs. In the case of IRMAs, the reasons for an increase in bound counts effected by BPs are not entirely clear. Both IRMAs tested employ a two-site anti-GH monoclonal antibody design. The radiolabeled antibody associates with the solid phase antibody by forming a ternary complex, with GH as the intermediate. The BP preparation promoted association of the labeled antibody with the solid phase (bead) through an unknown mechanism. Although the BP we used contained a small amount of GH (0.29 ng GH/U BP), which is derived from the affinity column used for purification, that amount is far
JCE&M«1991 Vol 72 • No 2
too low to explain the effect seen. Since the BP preparation was only partially pure, it contained a variety of other serum proteins that may have promoted nonspecific association of the labeled antibody with the polystyrene bead through a matrix effect. This effect was again relatively minor at the BP dose realistically encountered (0.1 U). Furthermore, it was corrected for in the final data calculation. Thus, in IRMAs, as in RIAs, the presence of GH-BPs appears to have little practical significance. In summary, we have shown BPs do not substantially affect GH immunoassay results. Although BPs have minor effects on antibody binding in RIAs or other nonspecific matrix effects in IRMAs, conventional assay designs minimize or eliminate their impact on final results. This can be further minimized by preparing immunoassay standard curves in GH-free or -poor serum/ plasma at a concentration ::epresentative of that used for the unknown samples. Many commercial assays already incorporate this design. Based on the present study, BP interference is not a likely cause of the discrepancies among GH results yielded by various immunoassays.
References 1. Baumann G, Stolar MW, Aniburn K, Barsano CP, DeVries BC. A specific growth hormone-binding protein in human plasma: initial characterization. J Clin Endocrinol Metab. 1986;62:134-41. 2. Herington AC, Ymer S, Stevsnson J. Identification and characterization of specific binding proteins for growth hormone in normal human sera. J Clin Invest. D86;77:1817-23. 3. Baumann G, Shaw MA. A second, lower affinity growth hormonebinding protein in human plasma. J Clin Endocrinol Metab. 1990;70:680-6. 4. Leung DW, Spencer SA, Cachianes G, et al. Growth hormone receptor and serum binding protein: purification, cloning and expression. Nature. 1987;330:537-43. 5. Baumann G, Shaw MA. Immunochemical similarity of the human plasma growth hormone-binding protein and the rabbit liver growth hormone receptor. Biochem Biophys Res Commun. 1988;152:573-8. 6. Barnard R, Quirk P, Waters MJ. Characterization of the growth hormone-binding protein of human serum using a panel of monoclonal antibodies. J Endocrinol. 1989;123:327-32. 7. Baumann G, Amburn K, Shaw MA. The circulating growth hormone (GH)-binding protein complex: a major constituent of plasma GH in man. Endocrinology. 1988;122:976-84. 8. Baumann G, Vance ML, Shaw MA, Thorner M. Plasma transport of human growth hormono in vivo. J Clin Endocrinol Metab. 1990;71:470-3. 9. Blethen SL, Chasalow FI. Use of a two-site immunoradiometric assay for growth hormone (GH) in identifying children with GHdependent growth failure. J Clin Endocrinol Metab. 1983;57:10315. 10. Rudman D, Chawla R, Ku;ner M. Heterogeneity of growth hormone in the nocturnal serum of children. Pediatr Res. 1985,19:9815. 11. Levin PA, Chalew SA, Martin L, Kowarski AA. Comparison of assays for growth hormone using monoclonal or polyclonal antibodies for diagnosis of growth failure. J Lab Clin Med. 1987; 109:858. 12. Reiter EO, Morris AH, MacGillivray MH, Weber D. Variable estimates of serum growth hormone concentrations by different radioassay systems. J Clin Endocrinol Metab. 1988;66:68-71.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
GH-BP EFFECTS IN GH IMMUNOASSAYS 13. Celniker AC, Chen AB, Wert Jr RM, Sherman BM. Variability in the quantitation of circulating growth hormone using commercial immunoassays. J Clin Endocrinol Metab. 1989;68:469-76. 14. Felder RA, Holl RW, Martha Jr P, et al. Influence of matrix on concentrations of somatotropin measured in serum with commercial immunoradiometric assays. Clin Chem. l989;35:1423-6. 15. Bierich JR, Briigmann G, Kiessling E. Use of provocative tests and measurement of spontaneous GH-secretion to diagnose GH deficiency. In: Miiller EE, Cocchi D, Locatelli V, eds. Advances in growth hormone and growth factor research. Rome/Berlin: Pythagora Press/Springer-Verlag; 1989;299-320. 16. Baumann G, Shaw MA, Amburn K. Regulation of plasma growth hormone-binding proteins in health and disease. Metabolism.
391
1989;38:683-9. 17. Drobny EC, Amburn K, Baumann G. Circadian variation of basal plasma growth hormone in man. J Clin Endocrinol Metab. 1983;57:524-8. 18. Baumann G. Growth hormone binding proteins and various forms of growth hormone: implications for measurement. Acta Paediatr Scand. In Press. 19. Mannor DA, Shaw MA, Winer LM, Baumann G. Circulating growth hormone-binding proteins inhibit growth hormone (GH) binding to GH receptors but not in vivo GH action [Abstract]. Clin Res. 1988;36:870A. 20. Baumann G, Shaw MA, Buchanan TA. In vivo kinetics of a covalent growth hormone-binding protein complex. Metabolism. l988;38:330-3.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.
