Assay of Somatomedin C by Cartridge Extraction Prior to Radioimmunoassay With Antiserum Developed Against Synthetic Somatomedin C

Transcription

1 ANNALS OF CLNCAL AND LABORATORY SCENCE, Vol. 8, No. 2 Copyright 988, nstitute for Clinical Science, nc. Assay of Somatomedin C by Cartridge Extraction Prior to Radioimmunoassay With Antiserum Developed Against Synthetic Somatomedin C PA CHH KAO, Ph.D., KAYOKO TATESH, Ph.D., CHARLES F. ABBOUD,.D., DONALD ZMMERMAN, M.D., RAYMOND V. RANDALL, M.D.t, and CHOH HAO L, Ph.D. Section of Clinical Chemistry, Division of Endocrinology, Metabolism, and nternal Medicine, Section o f General Pediatrics and Pediatric Endocrinology and Metabolism, Mayo Clinic and Mayo Foundation, Rochester, MN and Department of Biochemistry, School of Medicine, Fukuoka University, Fukuoka, Japan and Laboratory of Molecular Endocrinology, University of California, San Francisco,CA 9402 ABSTRACT The whole molecule of human somatomedin C (SM-C) prepared by the total synthesis m ethod was used as an antigen to produce an antiserum for a radioimmunoassay. Since plasma proteins that bind SM-C interfere with the assay, a m ethod was developed that uses acid dissociation followed by C-2 cartridge extraction to strip SM-C from its binding proteins before assay. This assay has no cross-reactivity with human proinsulin or insulinlike growth factor (GF-). The SM-C values in 339 normal subjects showed age-dependence, increasing from childhood to a peak at age 4 to 6 years and decreasing sharply before adulthood. n adults, the SM-C values decreased gradually with age. All 3 patients with acromegaly who were tested had an increased SM-C value, with no overlap with the normal range. The 2 patients with prolactinoma but non-growth-hormoneproducing pituitary tum or had no increase in SM-C. Two children with pituitary deficiency had low SM-C values; one of these children received growth hormone therapy, and his SM-C value increased from undetectable to normal. By three weeks after discontinuation of the therapy, his SM-C value was again undetectable. Of 20 children with short stature and * Presented at the Spring Meeting of the Association of Clinical Scientists, C harleston, SC, May 200 First Street, S.W., Rochester, MN , 987. $ Emeritus member, Division of Endocrinology, t Address rep rin t requests to Pai C hih Kao, Metabolism and nternal Medicine, Mayo Clinic and Ph.D., Section of Clinical Chemistry, Mayo Clinic, Mayo Foundation, Rochester, MN /88/ $0.20 nstitute for Clinical Science, nc.

2 SOMATOMEDN C ASSAY 2 constitutional delay of growth and development, SM-C was below normal in 70 percent and normal in 30 percent. Two patients with malnutrition had below-normal SM-C values. ntroduction Somatomedin C (SM-C), also called insulin-like growth factor (GF-), is a polypeptide chain of 70 amino acid residues with three disulfide bonds. Structurally, it is similar to proinsulin but has a shorter connecting peptide (C-peptide), 2 amino acid residues instead of 33 as in proinsulin. One major difference is that the biologically active SM-C molecule includes the 2-amino acid C-peptide, in contrast to proinsulin from which the 33- amino acid C -peptide is enzymatically rem oved to form the active hormone, insulin. A radioimmunoassay for SM-C would be useful in the management of growth h o rm o n e-rela te d p itu ita ry disord ers because the concentration of SM-C circ u la tin g in th e b lo o d is c o n sta n t, whereas that of growth hormone fluctuates6,9,0 during the day. Because the biosynthesis of SM-C is under the control of grow th horm one, circulating levels of SM-C reflect growth hormone secretion. T he first problem in developing a SM-C assay is the lack of a source of SM-C antigen. There is no depot organ for this hormone in the body, and it has to be isolated either from whole plasma or from a b y -pro duct of com m ercial blood protein fractionation, Cohn fraction V;,2 however, the yield is very low. Only about 20 mg of crude SM-C was isolated from 6,000 kg of Cohn fraction V.8 An alternative m ethod is to use the 2-amino acid C-peptide as antigen to produce antiserum in the hope that there will be cross-reactivity against the whole molecule of SM-C. However, few animals produced antisera having such cross-reactivity.4 Binding of SM-C by plasma proteins interferes or competes with binding of SM-C to antibody and thus poses a second problem in a SM-C assay. Sample treatm ents required prior to assay range from the simple addition of protamine to the assay buffer to acid-ethanol extraction to separation with acid buffer on a Sephadex column followed by lyophilization.,8,2,4 An antiserum was produced to synthetic SM-C, which solves the problem of finding readily available antigen. Acid dissociation and reverse-phase chromatography on a C-2 cartridge to extract SM-C prior to assay solve the second problem of interference by plasma binding protein in the interaction betw een SM-C and the antibodies to SM-C used as an assay reagent. The assay system was validated by analytical methods and used to quantitate SM-C levels in a large group of normal subjects. The system was then shown to give abnormal results in analyses of samples from patients with abnormalities of growth and growth hormone secretion. Materials and Methods m m u n iz a t io n o f A n im a l s Whole SM-C prepared by Li and associates6 (total chemical synthesis method) was conjugated via glutaraldehyde7 to keyhole limpet hemocyanin (KLH) at a ratio of 50 xg of SM-C to 90 jjlg of KLH. Conjugate equivalent to 50 xg of SM-C w as h o m o g e n iz e d w ith c o m p le te F reund s adjuvant and injected intradermally into two goats at m ultiple sites. Booster injections of the same preparation of antigen were given monthly at

3 22 KAO, TATESH, ABBOUD, ZMMERMAN, RANDALL, AND L the same dose. After six booster injections, one goat produced antibody with a titer of :200. ODNATON The purchased* SM-C was used to prepare radiolabeled tracer. A modified lactoperoxidase m ethod5,9 was used to iodinate 2.5 xg of SM-C with,500 fxci of 25. The labeled mixture was purified by gel filtration on two columns, first on a Bio-Rad P-6 column (0.7 X 5 cm) for desalting and then on a Sephadex G-50 column ( X 45 cm) for final purification. The elution buffer used with both columns was 0.2 M ammonium acetate at ph 4.8 containing 0. percent Tween 20 and 0. percent gelatin. The purified labeled SM-C was tested for homogeneity by high pressure liquid chrom atography (HPLC), in a Varian model 5000t with reverse-phase C8 column 300 X 4 mm$. The mobile phase was a linear gradient of 25 percent to 45 percent acetonitrile in one percent aqueous trifluoroacetic acid. The total elution tim e was 20 minutes; one-m inute fractions were collected and the radioactivity in each fraction was counted. R a d io im m u n o a s s a y The SM-C purchased from AmGen also was used to prep are a standard curve as follows. A solution of 0 ng of SM-C in 200 jl of assay buffer (0.04 M sodium phosphate containing 0. p ercent Tween 20 and 0. percent gelatin, ph 7.4) was serially diluted :2 with assay buffer to 0.08 ng p er 200 jjl. A mixture of 50 jjl of tracer containing 0,000 cpm, 50 (JL. of diluted antiserum that would bind 45 percent of the tracer, and 200 xl of standard or 200 xl of * AmGen Biologicals, Thousand Oaks, CA. t Varian, Sunnyvale, CA. t Alltech Associates, nc., Deerfield, L. extracted sample was incubated at 4 C for th re e days. T he free tra c er and bound tracer were separated with a second antiserum, burro anti-goat gg. All standards and samples were assayed in duplicate. C r o s s- R e a c t iv it ie s Human proinsulin was dissolved in assay buffer at 2,000, 500, 200, and 25 ng per ml, and 200-jxl samples of each solution (containing 400, 00, 50, and 25 ng of proinsulin, respectively) w ere assayed in duplicate to check the crossreactivity of human proinsulin with the antiserum. The cross-reactivity of insulin-like growth factor was determ ined with the same procedure at 400, 200, 00, and 50 ng per 200 xl. D is s o c ia t io n o f SM-C F r o m B in d in g P r o t e in s Elution W ith Neutral Buffer. A 0.3-ml sample of normal plasma was incubated with 25-labeled SM-C (8 X 06 cpm) at 4 C for two hours. A 0.2-ml sample of this mixture was applied to a Sephadex G-00 column ( x 50 cm) which was eluted with assay buffer at ph 7.4. The radioactivity of one-m l fractions was counted in a gamma counter. The colum n was precalibrated with blue dextran, hem oglobin, cytochrom e c, and p o tassiu m fe rro c y an id e (m olecular weights: 2 x 06, 64,500, 3,000, and 368, respectively). Acid Dissociation and Elution W ith Acidic Buffer. From the same plasma, a 0.2-ml sample was incubated with 25labeled SM-C (20 x 06 cpm) in the same m anner as that described previously. After the incubation, the sample Gift from Eli Lilly Co., ndianapolis, N. GF-, gift from Dr. Juergen Zapf, University of Zurich, Switzerland.

4 was mixed with.5 ml of M acetic acid and incubated again for two hours at 4 C. T hen 0.2 ml of the m ixture was applied to the same Sephadex G-00 column which was equilibrated and eluted with an acidic buffer (0.2 M ammonium acetate containing 0. percent gelatin and 0. percent Tween 20, ph 4.8). The one-ml fractions were assayed for radioactivity. Acid Dissociation, Cartridge Extraction, and T hen E lu tio n W ith A cidic Buffer. Another 0.2-ml sample from the same plasm a was incubated with 25labeled SM-C (3.58 X 06 cpm) and then mixed with.5 ml of M acetic acid as described in previous paragraph. This mixture was applied to a 0.5-g C-2 cartridgeh that was pretreated with three ml of methanol followed by six ml of 0.5 M acetic acid containing 0. percent trifluoroacetic acid. Then, the cartridge was washed three times with 2-ml portions of 0.5 M acetic acid containing 0. percent trifluoroacetic acid and eluted with three ml of acetonitrile per one percent aqueous trifluoroacetic acid mixture (75:25). The eluant was collected and d rie d at room te m p e ra tu re u n d e r reduced pressure in a running centrifuge.** The dried material was dissolved in 0.3 ml of the acidic buffer; radioactivity was counted; and 0.2 ml of this solution was applied to the same Sephadex column and eluted with the acidic buffer; one-m l fractions w ere collected and assayed for radioactivity. Elution Curve o f 25l-Labeled SM-C in Absence o f Plasma. 25-Labeled SM-C in 0.2 ml of acidic buffer was applied to the same Sephadex column and eluted with the acidic buffer; one-ml fractions were collected and assayed for radioactivity. f Analytichem nternational, Harbor City, CA. ** Speed Vac concentrator, Savant nstruments, Farmingdale, NY. SOMATOMEDN C ASSAY 23 S u b je c t s Blood sam ples w ere collected from 339 normal subjects (both sexes and ages less than one year to 64 years); proper informed consent was given by the subjects or by their parents. Blood also was drawn from patients with growth-hormone-related disorders or other diseases (after informed consent was obtained). Results P u r it y o f L a b e l e d SM-C The Bio-Rad P-6 column separated the iodination mixture into two peaks protein-bound iodine and free iodide. The fractions of protein peak w ere pooled and separated on the Sephadex G-50 colum n into three peaks (figure Upper). The second peak showed the greatest binding, and the first peak (void volume) and third peak (free salt) showed little binding with the antiserum. The fractions of the second peak were pooled and tested for homogeneity by HPLC on a C 8 colum n w ith a linear gradient (25 percent to 45 percent) of acetonitrile in one percent aqueous trifluoroacetic acid. This y ield ed a single peak (figure Lower) that had the same binding characteristics with the antiserum as the second peak from the Sephadex G-50 colu m n. T h is in d ic a te s th a t f u r th e r purification by HPLC is not necessary. D is s o c ia t io n o f SM-C After incubation of labeled SM-C with plasma, u n d er neutral conditions (ph 7.4), the bound labeled SM-C was eluted in th re e m ajor peaks re p re s e n tin g m olecular w eights of approxim ately 50,000, 56,000, and 8,000 (figure 2 A). The first two peaks were protein-bound material; the last one presum ably was u n b o u n d SM -C (m o lecular w eig ht, 7,469).

5 24 KAO, TATESH, ABBOUD, ZMMERMAN, RANDALL, AND L O() 05 ( > co ox c3o o F i g u r e. Upper, P urificatio n of d esalted labeled som atom edin C (SM-C) by gel filtration on Sephadex G-50 with 0.2 M am m o n iu m a c e ta te co n tain in g 0. p e rc e n t Tween 20 and 0. percent gelatin, ph 4.8. Radioactivity of -ml fractions was counted. Low er, Test of hom o gen eity of second peak from Sephadex G-50 column. The fractions of th e se c o n d p eak w e re pooled and subjected to high pressure liquid chrom a to g r a p h y on a C 8 re v e rs e -p h a s e co lu m n with a 20-ml linear grad ie n t (25 p e rc e n t to 45 p e rc en t) of aceto n itrile (ACN) in o n e p e rc e n t aqueous triflu o ro acetic acid. Fraction no. After elution under acidic condition (ph 4.8) from the same column, only the la b e le d SM -C b o u n d to m o lecu lar weight 50,000 protein disappeared (figure 2 B). W hen the acidified mixture was further extracted with a C-2 cartridge, a single peak was eluted (figure 2 C); it coincided with the peak of labeled SM-C alone in the absence of plasma protein (figure 2 D). The recoveries of radioactivity were 86 percent from the C-2 cartridge extraction and 96 percent from the Sephadex G-00 column. Thus, acidification with acetic acid and C-2 cartridge extraction became the m ethod used to extract all our plasma samples. S a m p l e C o l l e c t io n a n d S M - C E x t r a c t io n Blood (five ml from adults; two ml from children) was collected into EDTA tubes and centrifuged in a refrigerated centrifuge. The plasma was transferred to a glass tube, frozen in dry ice, and stored at 20 C in the laboratory until assayed. For assay, 0.2 ml of plasma was thawed, mixed with.5 ml of M acetic acid, and applied to a C-2 cartridge that had been pretreated with three ml of methanol and six ml of 0.5 M acetic acid containing one percent trifluoroacetic acid. The cartridge was washed three

6 FG U RE 2. D isso ciation of som atom edin C (SM-C) from plasma prot e i n. A, 25 - L a b e le d SM-C was incubated with hum an plasma, applied to Sephadex G-00 column ( x 50 cm), and elu ted w ith ph 7.4 phosphateb uffered saline; one-m l fractions w ere assayed. Molecular weight m arkers: BD, blue dextran 2 x 0s; Hb, hemoglobin; Cyt c, cytochrome c; salt, potassium ferrocyanide. B, Dissociation with M acetic acid and eluted fro m S ep h ad e x G -00 w ith ammonium acetate, ph 4.8. C, D issociated with M acetic acid and extracted w ith C-2 cartridge. The cartridge was w ashed, and th e SM -C was eluted from the cartridge, dried, dissolved, 6 2 E OQ. '> è M SOMATOMEDN C ASSAY Fraction no and applied to Sephadex G-00 column and eluted with ammonium acetate, ph 4.8. D, Labeled SM-C not bound to plasma protein eluted from Sephadex G-00 column with ammonium acetate, ph 4.8. times with two-ml portions of this acetic acid /triflu oroacetic acid m ixture and eluted with three ml of acetonitrile per one percent trifluoroacetic acid mixture (75:25). The eluant containing SM-C was dried overnight on the Speed-Vac concentrator at room tem perature, dissolved in 0.6 ml of assay buffer (as noted previously), and further diluted :5 with assay buffer; 200 jl of this diluted extract w ere assayed. Routinely, 0 sam ples (one a quality-control sample and nine unknowns) were extracted in a batch. V a l id a t io n o f A n a l y t ic a l P r o c e d u r e Recovery Study. A pooled sample was divided into two portions. To one portion was added SM-C standard at 62.5 ng per ml; the o th er portion w ithout added SM-C standard was the baseline. Each portion was further divided into 20 aliquots, and then all 40 aliquots were put through the extraction and assay procedures. The results (mean ± SD) were 84 ± 9 ng per ml for the baseline aliquots and 29 ± 6 ng per ml for the aliquots with added standard, for a recovery of 72 percent of the added SM-C. Linearity. Plasma samples from three normal subjects were extracted, diluted :5, :0, and :20 with assay buffer, and then assayed for SM-C. The results were proportional to the dilutions (table ). S e n s it iv it y a n d C r o s s -R e a c t iv it y The sensitivity of this assay is at least ng per tube because ng of SM-C gave a 0 percent suppression on the standard curve (figure 3). There was no cross-reactivity with human proinsulin up to 2,000 ng per ml of plasma. The cross-reactivity with GF- was. percent. Because it was suspected that the cross-reactivity w ith G F - would be lower after extraction, 00 ng of G F- w ere mixed w ith.5 ml of M acetic acid, extracted by C-2 cartridge, and radioim m unoassayed. T here was no

7 26 KAO, TATES H, ABBOUD, ZMMERMAN, RANDALL, AND L T A BLE Linearity of Assay Somatomedin C, ng/ml Subject Dilution Observed Theoretical Subject :5 : : Subject 2 : : : Subject 3 : : : m easurable im m unoreactive activity of the added G F- after extraction. That means that, after extraction, G F- up to 500 ng per ml will not show any measurable cross-reactivity. N o r m a l S u b je c t s The concentration of SM-C in plasma samples from 339 normal subjects plotted against the age (figure 4) indicates age-dependence. The SM-C concentration increased sharply with age to a peak at age 4 to 6 years and then began to decline (table ). Pa t ie n t s Among the adult patients, all 3 acromegalic patients with high growth horm one values had high SM-C plasm a values with no overlapping with normal (figure 5). Of the 2 patients with prolactin-producing pituitary adenoma but not growth-hormone-producing tumor (their growth hormone values were within the normal range), nine had normal SM-C and three had lower than normal levels. Both patients with nonfunctioning pituitary tum or had below -norm al SM-C values and undetectable plasma growth hormone levels. Among pediatric patients, of the 20 patients with short stature owing to constitutional delay of growth and developm ent, SM-C concentration was lower than normal in 4 and normal in six. Two patients with growth hormone deficiency had lower than normal values; neither showed an increase in growth hormone co n cen tratio n after in su lin-in d u ced hypoglycemia and arginine stimulation. O ne p atient (both w eight and height below the 5th percentile for his chronologic age of nine years) received human growth hormone (produced by recombi- Proinsulin b /b 0 F ig u r e 3. S t a n d a r d c u rv e fo r so m a to m e d in C (SM-C) a n d c ro ss-reactiv - ity w ith p r o in s u lin a n d in su lin -lik e g ro w th factor- (GF-) Hormone, ng/tube

8 250 H t* SOMATOMEDN C ASSAY 27 en c *** * * * * * * * ** * j * V * * * *** * lai * ****** * * * * ** * H * ** * * * * * * * * î V *** 4 \ m * ** *îé* î **** ki** *àv.% % **$ * * ** $ J* è* *** * * * * * * ^ * * * * * * * * * r F ig u r e 4. age. Age, year Somatomedin C (SM-C) concentration in plasma of normal subjects, plotted as a function of nant deoxyribonucleic acid [DNA] technology) subcutaneously, at two U three times a week. His SM-C value increased from undetectable to 42 ng per ml, a level within the normal range for his age. Three weeks after treatm ent with growth horm one was discontinued, his SM-C value was undetectable. Two patients w ith m alnutrition or insufficient calorie intake had lower than normal SM-C values. One patient s body weight was well below the 5th percentile for her age, and the other was at the 0th percentile for his age. The boy improved his food in tak e to som e ex ten t; six months later his body weight was at the 25th p e rc e n tile, and his SM -C was in creased about 2-fold b u t was still slightly below the normal range for his age (5 years). Two c h ild re n w ho h ad excessiv e growth, but otherwise were healthy, had normal SM-C levels. Discussion The usefulness of a reliable SM-C assay in the evaluation of patients with acromegaly or with disorders of growth has b e e n am p ly d e m o n s tr a te d by T A B L E Plasma Somatomedin C in 339 Normal Subjects by Age Somatomedin C, ng/ml 90 Percent Age Confidence Year n Mean SD Range Limits P* < < < < < NS < 0.05 > < 0.0 *For difference from next younger age group

9 28 KAO, TATESH, ABBOUD, ZMMERMAN, RANDALL, AND L others. Nonetheless, the availability of reliable SM-C assays has been restricted to only a few laboratories because of difficulties in making antibodies specific for SM-C in titers and amounts adequate to support a reliable assay for clinical use over long periods of time. Because of the num ber of patients seen by the present authors for whom SM-C assays would be applicable, the development of a dependable assay for SM-C was undertaken. However, the facilities were lacking to isolate SM-C from large am ounts of human plasma and it was not possible to obtain an antiserum with useful crossreactivity despite repeatedly using the 2-amino acid C -peptide in different species of animals. Hence, the synthetic whole molecule of SM-C as antigen was tried when it became available and an alternative m ethod was developed to make a reliable, clinically useful assay for SM-C. The interference of binding proteins in the radioim m unoassay for SM-C is well known. A num ber of methods have been used to m inim ize this problem, such as exposure of the plasma sample to h eparin or p ro tam in e, acid -ethanol extraction, acid Sephadex column separation, or treatm ent with acid and then lyophilization. 4 5,8 2 4 W hen plasma interference is high or when hormone concentration is low assay strategies, including an extraction procedure, have become increasingly popular. n the case of SM-C assay, a concentration proced u re is unnecessary because plasm a SM-C concentrations are high and the assay is adequately sensitive. However, the interference of binding proteins is a major problem. Extraction was used by the present authors to remove the interfering SM-C binding proteins prior to assay. Various cartridges becam e com m ercially available for extraction, including silica, amino, and C 8 cartridges. These have been applied in the extraction of a num ber of hormones prior to assay o> c d C O * n = 3 n = 2 n = 20 - A r Ä - A A A J ' / A A b o v e n o r m a l O N o r m a l B e l o w n o r m a l U n L n J! D i i i # i o - i «i *» * i «% Adults / / / & Pediatrics F i g u r e 5. Somatomedin C (SM-C) concentrations in patients with various d iseases. B ecause normal SM-C concentratio n is age d e p en d en t, symbols are used to indic a te w h e th e r o r n o t a value is normal. Particularly in pediatric patients, the normal value of SM-C is lower in a young child th an in an o lder child; thus, the far right column shows b o th norm al and below-normal at the same S M -C l e v e l. A r r o w, C hange in m aln u tritio n patient at six months after improving food intake.

10 e.g., assays of adrenocorticotropin, atrial natriuretic peptide, and calcitonin.2 3,3 Although use of such cartridges changes assays from direct to indirect assays, it greatly im proves th eir reliability and sensitivity. O ne windfall advantage is that extraction assays may decrease the risk of blood-transmitted diseases, which at present is a concern of all laboratory p ersonnel. The extraction pro ced ure req u ires m ixing the p a tie n t s plasm a with a high concentration of strong acid as a first step, before applying the sample to the cartridge, and this will inactiv a te m o st of th e m ic ro o rg a n ism s present. The molecular weights of the major SM-C binding proteins w ere 50,000 and 56,000, as estimated on a Sephadex G-00 column. Studies of acid dissociation showed that the binding of SM-C with the larger protein was weaker than that with the smaller protein. The SM-C bound to the larger protein could be released with M acetic acid, but that bound to the smaller protein could not be released by acid alone (it required the additional help of a C-2 cartridge). The molecular weights of these two binding proteins are similar to those found by o th e r in v estig ato rs using S ephadex G-200 colum ns.7,20 The 50,000-dalton binding protein has been shown to be growth hormone-dependent; in patients w ith growth horm one deficiency, this binding protein disappeared and then reappeared when the patients received growth hormone therapy. By using synthetic SM-C as an antigen and acid d isso ciatio n and c a rtrid g e extraction to remove interfering binding proteins, a clinically useful radioimmunoassay for SM-C was developed by us. n addition to establishing a norm al range for various age groups, it has been shown th at patients w ith acrom egaly have high levels of SM-C. Preliminary data suggest that children with growth SOMATOMEDN C ASSAY 29 hormone deficiency and many with constitutional delay of growth and developm ent have low SM-C levels. Finally, by using this assay, it has been shown by us that patients with undernutrition have low circulating levels of SM-C. References. B a l a, R. M. and BHAUMCK, B.: Radioimmunoassay of a basic somatomedin: Com parison of various assay techniques and som atom edin levels in various sera. J. Clin. Endocrinol. Metabl. 49: , B o d y, J.-J. and H e a t h, H. ll: Nonspecific increases in plasma immunoreactive calcitonin in healthy individuals-. D iscrim ination from medullary thyroid carcinoma by a new extraction technique. Clin. Chem. 30:5-54, B u r n e t t, J. C., J r., K a o, P. C., H u, D. C., et al.: Atrial natriuretic peptide elevation in congestive h eart failure in the hum an. Science 237:45-47, C h a t e l a in, P. G., Van W yk, J. J., C o p e l a n d, K. C., et al.: Effect of in vitro action of serum proteases or exposure to acid on measurable im m unoreactive somatomedin-c in serum. J. Clin. Endocrinol. Metab. 56: , C l e m m o n s, D. R., U n d e r w o o d, L. E., C h a tela in, P. G., etal.: Liberation of immunoreactive somatomedin-c from its binding proteins by proteolytic enzymes and heparin. J. Clin. Endocrinol. Metab. 56: , C l e m m o n s, D. R., Van W yk, J. J., R id g w a y, E. C., et al.: Evaluation of acromegaly by radioimmunoassay of somatomedin-c. New Engl. J. Med. 307:38 42, C o p e l a n d, K. C., U n d e r w o o d, L. E., and Van W y k, J. J.: nduction of im m unoreactive somatomedin C in human serum by growth hormone: Dose-response relationships and effect on chromatographic profiles. J. Clin. Endocrinol. Metab. 50: , D aughaday, W. H., M a r e,. K., and Bl e t h e n, S. L.: nhibition of access of bound somatomedin to membrane receptor and immunobinding sites: A comparison of radioreceptor and radioim m unoassay of som atom edin in native and acid-ethanol-extracted serum. J. Clin. E ndocrinol. M etab.5i , D E r c o l e, A. J., U n d e r w o o d, L. E., a n d Van W yk, J. J.: S e ru m so m a to m e d in -C in h y p o p itu i ta ris m a n d in o t h e r d is o rd e rs o f g ro w th. J. P ediatr. 90: , D robny, E. C., A m b u r n, K., and Ba u m a n n, G.: Circadian variation of basal plasma growth hormone in man. J. Clin. Endocrinol. M etab. 57: , F u r l a n e t t o, R. W.: Pitfalls in the somatomedin-c radioimmunoassay. J. Clin. Endocrinol. Metab , 982.

11 3 0 KAO, TATESH, ABBOUD, ZMMERMAN, RANDALL, AND L 2. F u r l a n e t t o, R. W., U n d e r w o o d, L. E., Van W yk, J. J., et al: Estimation of somatomedin-c levels in normals and patients with pituitary disease by radioim m unoassay. J. Clin. nvest. 60: , G u p t a, M. K. and K o l a r, T.: Development and evaluation of a highly sensitive and specific radioim m unoassay for th e m easu rem en t of ACTH in hum an plasma (abstract). Clin. Chem. 3:962, H i n t z, R. L., L iu, F, and S e e g a n, G.: Characterization of an insulin-like grow th factor-/ somatomedin-c radioimmunoassay specific for the C -p eptide region. J. Clin. Endocrinol. Metab. 55: , Ka o, P. C., Jia n g, N.-S., and A b b o u d, C. F : Radioimmunoassay of human homologous prolactin in serum w ith com m ercially available reagents. Clin. Chem. 2 3 : , L i, C. H., Y a m a s h i r o, D., G o s p o d a r o w i c z, D., et al.: Total synthesis of insulin-like growth factor (somatomedin C). Proc. Natl. Acad. Sci. U.S.A. 80: , P a r k e r, C. W.: N a tu re o f im m u nological responses and antigen-antibody interactions. P rincip les o f C o m petitive P ro tein -B in din g Assays. Odell, W. D. and Daughaday, W. H., eds. Philadelphia, J. B. Lippincott, 97, pp R i n d e r k n e c h t, E. and H u m b e l, R. E.: Polypeptides with nonsuppressible insulin-like and cell-grow th prom o ting activ ities in hum an serum: solation, chemical characterization, and some biological properties of forms and. Proc. Natl. Acad. Sci. U.S.A. 73: , R o t h, J.: M ethods for assessing immunologic and biologic properties of iodinated peptide hormones. Methods Enzymol. 37: , W h i t e, R. M., N i s s l e y, S. P., M o s e s, A. C., et al.: The growth hormone dependence of a somatom edin-binding protein in hum an serum. J. Clin. Endocrinol. Metab. 53:49-57, 98.