of human growth hormone. (HGH) was made practicable by Glick et al. (2) and Hunter and Greenwood. (3) through the applica- tion of the radioimmunoassay.
Radioimmunoassay
of Human Growth
Hormone
Mechthilde Knoller, M. U. Tsao,* and George H. Lowrey
Detailedmethodsfor: (1)
iodination of human growth hormone (HGH); (2) purification of labeled HGH (HGH-139); and (3) radioimmunoassay of HGH are given. Bio-GeI filtration is introduced as a rapid and reproducible method for purifying 1311
HGH-1311which has beenobtainedby a modificationof the methodof Greenwood Highly purified HGH-’311 with a bindability of 96% or more is usually
at al. (1).
achieved.
‘T1HE ASSAY of human growth hormone (HGH) was made practicable by Glick et al. (2) and Hunter and Greenwood (3) through the application of the radioimmunoassay method. Without such a highly sensitive method, it would not be possible to measure the level of 11011 in blood in the millimicrogram range. Studies using these methods indicated that the sensitivity and reproducibility of the assay depend mainly on a high degree of bindability and purity of the HGH iodine-131 (HGH1311) that is used as a tracer. Therefore, the labeling of the UGH and the purification of the HGH-1311 are important steps of the procedure. For the labeling of HGH with 131J a modification of the Greenwood et al. method (1) proved satisfactory. To purify HGH-1311 Glick et al. (2) used dialysis followed by fractionation by starch gel electrophoresis; Hunter and Greenwood (4), Greenwood et al. (1), and TJtiger (5) used Sephadex 0-50 columns; Frantz and Rabkin (6) used Sephadex G-50 columns followed by vertical starch gel electrophoresis. In the present investigation conditions to obtain optimum results from the assay method were studied.
Materials and Methods Reagents Phosphate buffer, pH 7.4, 0.25 M Solution A: 34.50 gin, of NaH0PO4 . 1120 dissolved in 1000 ml. of water; Solution B: 67.0 gin, of From the Department of Pediatrics and Conirnuiiienble Diseases, University of Michigan Medical Center, Ann Arbor, Mich. 48104. Supported by Grant HD 01054, National Institutes of Health, U. S. Public Health Service. The authors wish to thank A. E. Wilhelnii, National Institute of Arthritis and Metabolic Diseases, for the HGH preparation, and the National Pituitary Agency for the nnti-HGH set-urn. Received for publication May 24, 1967; accepted for publication Aug. 4, 1967. ‘Present address: University of California School of Medicine, Davis, Calif. 145
146
KNOLLR
fT AL.
Clinical
Chemistry
Na0HI’04 . 71L0 dissolved in 1000 ml. of water; 95 ml. of Solution A and 405 ml. of Solution B are diluted to 1000 ml. Phosphate buffer, pH 7.4, 0.05 M Solution A: 6.9 gm. of NaHJ4. i-LU dissolved in 1000 ml. of water; Solution B: 13.4 gin. Na0HPO4 7H..O dissolved in 1000 ml. of water; 95 nil, of Solution A and 405 ml. of Solution B are diluted to 1000 ml. Barbital buffer, pH 8.6, 0.1 2W Diethyl barbituric acid, 36.8 gm. and 205.3 gm. of sodium barbiturate are dissolved in 10 L. of water and stored at 40, ILl dilnent Dilute 2 ml. of 25% (w/v) normal human serum albumille and 1 ml. of normal rabbit piasmat to 100 ml. with barbital buffer, pH 8.6, 0.1 M. ( ‘l,loran,ine-T,t 0.4 mq./ml. i)issolve 10 nig. of (‘hloramine-T in 25 ml. of phosphate buffer, p11 7.4, 0.25 M, within 2 hr. or less of each labeling piocedure. Sodium metabisuifite, 0.6 mq./nii. 1)issoive 15 mg. of sodium metabisulfite in 25 ml. of phosphate Iniffer, pH 7.4, 0.25 M, before each labeling procedure. Potassium iodide, 1% IV 16%shIcrose Dissolve 1 gm. potassium iodide and 16 gui. sucrose in 100 nil, of water ; multI a small amount of bromphenol blue. Bio-Gei P60, .50-150 mesh Willieimi HGH, HS 612 B Used both for the labeling procedure and HGH standard solutions. It was stored at 4#{176} in a desiccator until used. One-milligram portions of the pituitary extract were weighed in 0.7-mi. weighing jars using a nlicrol)alaiice and were used for the following (1) Labeling procedure: 1 mg. of if(H extract w’as dissolved in 1 ml. of barbital buffer, pH 8.6, 0.1 M, and stored in 15- to 2O-pJ. portions at -20#{176} or, preferably, at -60#{176} (the solution has been found to be stable at this temperature for at least 19 months). (2) Stock HGH standard solution: I mg. of UGH extract was dissolved in 0.2 ml. of barbital buffer in the weighing jar and was then transferred to a 100-mi. volumetric flask, the jar was quantitatively rinsed at least 10 times with barbital buffer, and the mixed washings were made to volume with the buffer. The solution was stored in 1- to 2-mI. amounts at -20#{176} or, preferably, at -60#{176}. Guinea pig anti-HGH serum As l)repai’ed ly Wick et al. (2) and .
Hyland Laboratories, Los Angeles, Calif. tColorado Serum Co., 1)enver, Cob. MaIliiickrodt Chemical Works, St. Louis, Mo. Cahbiochem, Los Angeles, Calif.
-
-
Vol. 14. No. 2, 1968
HUMAN
GROWTH
147
HORMONE
supplied by the National Pituitary Agency* at a dilution of 1 :2500 in 0.05 M barbital buffer coiltailbing 2.5 nig. human serum albumin pci’ milliliter, control (nollinimune) guinea pig plasma at a dilution of 1 :100, and 1 :10,000 merthiolate. The solution was kept at 4#{176} (storing at freezing temperatures damaged the activity of the antibody). Rabbit anti-HGH sernmt Fsed for the I)indability test of 1311.. labeled HGH. Papain (2 X crystalline suspension) lyophilized (from 0.01 cysteine solution). Activity: 12 units/mg. protein.t Sodium iodide-1.i1 (Na 1t1J) Free from thiosulfate and preservative, in 0.03-0.1 ml. NaOH. it was used within 24 hr. of arrival. Procedures Collection
of Specimen
Blood samples were withdrawn after was separated and frozen immediately. Preparation of ‘l-labeled
overnight
fasting;
the
serum
HGH
Two millicuries of Na1311 and 5 g. of HUH in buffer solution, 1 mg./ ml., were transferred into 25 l. of phosphate buffer, pH 7.4, 0.25 l, in a 2-mi. vial. Disposable 20-l. pipets (1)rummond “IMicrocaps,’’ Kensington Scientific, Berkeley, Calif.) were found convenient for transferring N&311 and HGH solutions. The concentrations of commercially obtained Na1311 varies with different batches, so tile pipets were marked prior to use at a poilit correspondutg to a volume containing 2 mc. of Na131J. (All subsequent additions of reagents were ma(le by inserting a 23-gauge needle attached to a disposable i-mi. syringe into the rubber capped vial.) Chloramnine-T reagent, 25 p1., was added and with constant swirling was allow-ed to react for 30 sec. The reaction was stopped by the addition of 50 /Ll. of sodium Inetahisulfite reagent. A total of 100 of 1% potassium iodide in 16% sucrose contaming bromphenol blue were added, and tile reaction mixture was withdrawn with the same syringe. One drop of this mixture was diluted in 5 ml. of BA diluent and the rest of the reaction mixture was transferred to a Bio-Gel column (see below) under a layer of 2 ml. of phosphate buffer, pH 7.4, 0.05 i. Separation of “I-labeled
HGH
A Bio-Gel coiumn was prepared prior to iodination by suspending approximately 1 gm. of Bio-Oel P60 in 60 ml. of phosphate buffer, pH 7.4, 0.03 M. The gei beads were washed once, and a coiume of 14 X 1 cm. National Pituitary Agency, Baltimore, Md. Mann Research Laboratory, New York, N. V. lsoserve, Inc., Cambridge, Mass.
148
KNOLLER fT AL.
Clinical
Chemistry
was prepared by pouring the washed beads into a 20 cm. long glass tubing which has been drawn to a short tip containing a small amount of glass wool. The column was presaturated with 0.5 ml. of 5% human albumin in phosphate buffer, 1)11 7.4, 0.05 M, and then washed with 10-15 ml. of the phosphate buffer. The reaction mixture was added as explained above. The gel filtration was performed at room temperature. The first 6 ml. of eluate were discarded. Four 0.3-mi. portions were collected in 0.5 ml. of BA diluent in disposable plastic tubes and stored at 4#{176} until the following day. This was as effective as the use of lower temperatures. The column was discarded after collection of the eluates. To determine the uptake of 131J by UGH (see below) and the purity of the gel-filtered eluates, 100 jJ. of bromphenol-blue-stained normal rabbit plasma was added to 1 ml. of diluted sample (see below), and 50 l. of the mixtures were applied to Whatman 3MC paper strips 1#{189} in. wide and subjected to chromatoelectrophoresis at 4#{176}. A voltage of 500 was applied for 70 mm., after which the strips were dried. The radioactivity of the separated fractions was scanned on an automatic strip scanner (Atomic Accessories, Model RS-241). The efficiency of the 131J transfer to UGH and the purity of the gel filtered fractions were calculated from the “scannograms.” The amount of the radioactivity of all samples (diluted reaction mixture and eluates from gel filtration) was determined and adjusted to the desired level prior to chromatoelectrophoretic analysis in the following manner: a SO-l. sample was pipetted into a disposable plastic tube which was then placed in a scintillation well detector (Nuclear Chicago, Model 905) that was connected to a scaler (Nuclear Chicago, Model 8725); if necessary the sample was diluted with BA diluent and the determination repeated. The optimum amount of radioactivity in the samples depended on the sensitivity range of the radioactivity scanner that was used for scanning the paper strips; hence it was established empirically. The scannograms of the reaction mixtures showed a large peak at the origin; a small peak in the and $-globulin region, which constituted damaged protein fractions (“reaction damage”); and a peak of varying size of free N&311 (Fig. iA and 2A). All protein-bound 131J is considered as 131J uptake by UGH and is expressed as per cent of total radioactivity. The scannograms of the purified fractions showed only one peak at the origin. No tailing was observed (Fig. lB and 2B). -
Bindability of “I-labeled
HGH
The purified fractions rabbit anti-UGH serum.
were tested for antigenicity For this purpose commercially
with an excess of available rabbit
Vol. 14, No. 2, 1968
HUMAN
GROWTH
149
HORMONE
anti-UGH was prepared according to a modified method of Porter (7): The protein concentration of tile rabbit anti-HUH serum was determined with the Biuret method. One milliliter of rabbit anti-UGH serum and 70 mg. protein per milliliter were mixed with 1.0 ml. of a papain solution containing 70 mg. of papain per milliliter of 0.25 M sodium phosphate buffer, and diluted with the phosphate buffer to 5.0 ml. Two drops of toiuelie were added. The mixture was incubated at 37#{176} for 16 hr., and then dialyzed against water for 48 hr. with frequent changes of the outer liquid. Some precipitate formed during dialysis and was removed by centrifugation. The clear product was stored at 4#{176} (no loss of activity has been observed in 2 years). A 1:100 dilution with BA diluent was made from the concentrated digestion product approximately every 3 months, and was stored at 4#{176}. The incubation mixture consisted of 0.4 ml. of BA diluent, 0.05 ml. of rabbit anti-UGH serum diluted 1 :100 with BA diiuent, and 0.05 ml. of the diluted gel filtrate. Tile mixture was incubated for 60 mm. at room GEL FILTRA’flON REtCTION MIXTURE
PURIFIED
HGH -
ANBOG-
SOUND
HG-’
1
Fig. 1 A. Radioactivity scannograrn of chrornatoelectrophoretogram: iodination reaction mixture of 97% efficiency in uptake of “'I by HGH. B. Same mixture as in A but after purification by gel filtration on Bio-Gel P60. C. Seannogram of incubated mixture (60 mm.) of 50 l. purified HGH solution and excess anti-human growth hormone serum (Mann), diluted with barbital buffer, pH 8.6, 0.1 M, to 0.5 ml. total volume. Arrows indicate points of application of incubation mixture. Fig. 2 A. Radioactivity scannogram of chromatoelectrophoretogram: iodination mixture of 60% efficiency in “‘I uptake by UGH. B. Same mixture as in A but after purification by gel filtration on Bio-Gel P60. C. Scannogram of 50 l. of mixture from B treated as in Fig. 10.
150
KNOLLR
temnperature results),
(incubatiolls subjected
to
overnight paper-strip
fT AL.
at 4 gave essentially chromatoelectrophoresis,
Clinical
the the
Chems+ry
same dried
strips scanned, and the scannograms calculated. Tile bindability of i3i1 labeled HGH was expressed in terms of percentage of total HGH-’311 bound by rabbit anti-UGH serum. Only purified gel filtrates of a bindability of 96% or more were used as tracers (Fig. 1C and 2C) for radioimmunoassay of HGH. Radoimmunoassay
Appropriate dilutions, depending on the sensitivity of the radioactivity scanner, were macic from one ot’ the gel filtrates that showed no trace of damaged protein and a bindability of no less than 96%. The radioactivity of 5O-jil. portions of diluted tracer was determined lim a scintillation well detector. The specific activity of the tracer usually ranged between 340-390 HUH. The guinea pig anti-HUH serum was usually used at a final dilution of 1 :1,000,000. Standard tubes were set up in duplicate at HUH colicentrations ranging from 0-2.0 mg./ml. in tile following manner: from the stock HUH standard solution a diluted standard solution, containing 0.01 m/Lg. of HUH per niicroliter of BA diluent, was freshly prepared for each experiment. No HUH standard solution was added to the tracer control tubes; 5, 10, 20, 40, 60, 80, 100, and 200 l. were added to the standard tubes; 1 l. of HGH solution, 1 mg./ml., was added to the guinea pig anti-HUH serum colitrol tubes to bind the anti-HUH completely. All standard amid standar(i control tubes contained BA diluent in amounts ranging from 0.40.2 ml. Serum samples were generally assayed in duplicate, and occasionally in triplicate. Tubes that contained all of tile reactants except anti-HUH serum were used as unknown controls. The serum was added to BA (liluent to give a final dilution of 1 :10 or greater. To all standard, standard control, and unknown tubes, 50 pl. of anti-HUH serum were added, the contents mixed, and the tubes incubated overnight at 4#{176}. After approximately 20 hr., 50 l. of tracer solution was added to all tubes. The total volume of the mixture was 0.3 nil.; the total incubation time was 5-6 days at 4#{176}. The 30 l. of bromphenol-blue-stamed normal rabbit plasma was added to the tubes just prior to chromatoelectrophoresis (normal rabbit plasma enhances the separation of the bound and free HGH-’311; bromphenol blue served as a separation indicator). Two hundred microliters from each mixture were applied to strips of Whatman paper 31IC, 1#{189} in. wide, and subjected to chromatoelectrophoresis for 70 mm. at 4#{176} at 500 v. The strips were dried, scanned n an automatic strip scanner, and the radioactivity was quantitated from the
Vol. 14, No. 2, 1968
HUMAN
GROWTH
151
HORMONE
integration counts of’ the peaks on the scannogramn. The ratio between the bound protein (B) that travels within the globulin region and the free protein (F) that stays at the origin was calculated by the following procedure: Standard: The percentage of tile ibbcul)ation damage was calculated from the control sample; tile suni of integration counts of F + B (sum of integration conllts of F + B X % incubation damage) = corrected sum of F + B. -
(‘orreeted
Tntegration counts of B sum of integration counts
2L. %
1
-
of F +
B
-
B:F
F
A standard curve was made (Fig. 3) by plotting the B :F1 ratios against the HUH concentration of the standard samples. Unknown samples: Tile percentage of the incubation damage was calculated from the unknown control samples and the B :F ratios were calculated in the same manner as for the standard samples. The UGH concentration of the unknown samples was derived from comparison of their B :F ratios with those of the standard curve (Fig. 3). Calculated from a group of 36 samples analyzed in duplicate, the standard devia-
Fig. 3. Example of standard curve obtained when the antibody -hound HGH-”I:frec HGH-”I (B:F) is plotted ;,gninst known amounts of UGH (Wilhelrni UGH, HG 612 B). HGH-”I, 61 g. UGH per milliliter; antibody final dilution, 1 :1,000,000.
I,,
1:,
I Lj
0. 0. #{163} 1.0 Human
.4
1.8
22
Growth Hormone (mjig/mI)
tion of a single determination from the same run was 3.4% of tile HUH value. From repeat analysis of 25 samples on different runs (or batches) the standam’d deviation of the mean value of duplicate analysis was 15.3% of the mean.
152
KNOLLER
fT AL.
Clinical
Chemistry
Results and Discussion Of the many factors that could influence the precision of the radioimmunoassay of HUH, the conditions of 13iJ labeling which directly affect the specific activity of labeled hormone seems critical. These conditions of labeling not only determine the efficiency of 13iJ uptake by the hormone but they also apparently influence the immunochemical property of the purified radioactive hormone. The bindability of 13iJ labeled UGH by the guinea pig anti-HUH serum appears to be critical for the Sellsitivity of the assay procedure and is affected by the quality of the Na1311 and the labeling procedure. Carrier-free, preservativefree Na1311 was procured and used within 24 hr. of arrival in order to keep the “preparation damage” of 1311-labeled HUH at a minimum and to preserve the antigenicity of the HUH molecule. The experimental conditions fom’ the preparation of ‘31T-labeled HUH were studied. For the preparation of ‘311-labeled HUH, varying amounts ranging from 4.4 to 88 g. of oxidizing agent (Chloramine-T) were used. Various reaction periods, ranging from 30 sec. to S mm. also were applied. No essential difference in the efficiency of lliJ transfer to HGH was noted under these different conditions. With 8-10 g. of Chloramine-T and a reaction period of 30 sec. an efficiency of 90-97% was generally achieved and the reaction damage of HUH was kept between 3-7% (Fig. 1A amid 1B). Occasional preparations of Na’311, labeled under identical conditions, gave efficiencies of only 50-65%. Greenwood et al. (1) obtained no 1311 transfer with amounts of Chloramine-T below 35 pg.; however, they did achieve efficiencies of up to 76% when amounts of ChloramineT ranging from 35 to 450 pg. were used with Na1311 obtained from the Radiochemical Center.* In the present investigation the specific activity ranged between 360-390 tc./g. when 131J uptake was between 90 and 97%; specific activities of only 200-280 c./ig. were obtained when 131J uptake was between 50 and 70%. Thus, with 1311-labeled HUH of high specific activity the desired radioactivity was obtained with 40-60 g. of purified HUH per milliliter, but in order to obtain the same level of radioactivity with ‘31T-labeled uGH of low specific activity, 80-100 1g. per milliliter were necessary. Higher “initial B:F ratios” (the B:F ratio obtained in the absence of unlabeled hormone) often were achieved with 131L labeled HUll of lower specific activities. This is in contrast to the findings of Berson el al. (8) who reported that higher initial B:F ratios were acilieved with higher specific activities. Comparisons of the HO-H level of a series of samples-with the use of different batches of 131T-Iabeled UGH thus determined with different standard curves of mlThe
Radiochemical
Center,
Amersham,
England.
Vol. 14, No. 2, 1968
Table
1. uGH
HUMAN
VALUES
DETERMINED
WITH
INITIAL HGII
B:F
(mpg/mI.
STANDARD
aerum)
a
inigial
1.06
0.64
32
3.0
3.6
-
-
1.5
-
-
2.3
-
-
-
-
0.3 LS
0.4 1.3
-
-
-
-
-
-
-
-
-
-
-
-
8 9 10 11 12 13 14 15 16 17 18 19 20 21
ralioa
3.6
1.4
3.5
7
B:F
1.86
L7
3.2
6
DIFFERENT
-
1.8
3.0
-
OF
1.33
1.9
1 2
5
CURVES
RATIOs
Sample
3 4
153
GROWTH HORMONE
l.2 2.1
-
-
-
-
-
-
-
-
-
-
1.5
1.5
-
-
-
2.4
-
-
-
4.2
-
-
-
-
2.4 3.9 7.5
8.0
-
-
-
-
-
-
1.4 0.4
1.4 0.2 1.1 4.2 6.0 0.3 1.1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
1.8 4.3 6.2 0.55 1.2 1.8 6.2 IA 36 L9
-
-
1.6 5.8 1.4 3.8 1.9
tial B:F ratios ranging from 2 to 1 or less-showed identical results (Table 1). initial B:F ratios most frequently obtained in this laboratory were between 1.9 and 1.4. The sensitivity of the assays appears to be dependent mainly on a steep initial slope of the standard curve, which in turn is determined by the bindability of the tracer. Therefore, the greater the bindabihity, the greater the sensitivity. Thus with a bindability of the tracer of 96% or more, the minimum detectable amount of UGH was 0.01 mpg. of purified HGH per milliliter, and 0.1 mpg. of serum HUH per milliliter regardless of the initial B :F ratio of the standard curve. The separation of 1311-labeled HUH from damaged protein and nonreacted Na1311 was performed immediately after labeling, and took only 10-15 mm. at room temperature by gel filtration with Bio-Gel P60. Separation at 4#{176} did not show any difference in results. The first fraction of eluates containing radioactive material included a damaged-protein fraction; the following fraction of 2.5-3 ml. was free from damaged protein. Only the latter fractions were routinely collected. The eluates collected subsequently contained damaged protein, and, eventually, unbound Na1311 (Fig. 4). Some radioactive material was absorbed at the origin, but complete recovery is not essential since
KNOLLER
154
El’ AL.
Clinical
Chemistry
an ample amount of pure HUH-1311 fraction, suitable for the assay, is obtained. Separation with Sephadex G-73 columns, 14 X 1 cm., under comparable conditions gave similar results. Both gel filtrations are simple, fast, and reproducible. They have considerable advantage over
‘-#{149}
TOTALRADIOACTIVITY ‘DAMAGED PROTEIN
90 Fig. 4. Typical elution tern of purified HGH-”I
80
a ration from damaged protein and Na”I on Bio-Gel P60. Vol. ume of eluate, 0.5 ml./tube. Large peak represents radioactivity counts of ‘“I.labeled HGH, sn,aller peak those of
70 60 50
a 40 .
20 10
patsep.
Na”!. Tubes 8 and 9 contain 1IGHI and damaged protein. Tubes 10-15 contain HGII-”I free of damaged protein. An increasing amount of damaged protein occurs in eluate after Tube 15. Damaged protein is expressed ‘I-labeled
as per uGH.
cent
of
total
TUBE NO.
the procedure using dialysis followed by fractionation by starch gel electrophoresis (2); aIld the method employing separation by the Sephadex G-flO column followed by vertical starch gel electrophoresis (6). The latter two methods are tinle-conSumillg and tedious, and the amount of labeled UGH recovered from starch gel is smaller and no purer than that obtained with gel filtration. SonIc breakdown of the labeled HUH occurs during incubation at 4#{176}. The released free binds with the globulin of tile human serum during incubation, and probably with time rabbit serum added prior to chromatoelectrophoresis. The ‘‘incubation damage’’ will show a tailing or small peak in the B area of the standard control samples containing excess HUH and guinea pig anti-HUH serum, and also in tile ulIkIloWil control samples to which 110 anti-HUH serum had been added. After 3-6 days of incubation at 4#{176}, the incubation damage was usually 3.5-6% with purified HGH and 8-12% with tile unknown human serum saniples tilat were assayed at a dilution of 1 :10. l)ilutions of 1 :20 or highei- i-educed this damage. It was found that incubation damage of up to 15% still gave reproducible i-esults. Because conditions may vary between assays, a serum sample of known
Vol. 14, No. 2, 1968
HUMAN
GROWTH HORMONE
155
HUH level was regularly included with each assay; the mean value of this sample was 3.17 (± 0.38) mpg./ml. Special attention should he given to the storing temperature of the anti-HUH serum. Temperatures of -20#{176} may not damage anti-HUH serum if it is undiluted or highly concentrated and if no preservative is present. Diluted anti-HUH set-mn, and in particular anti-HUH serum with preservative added, should be kept at 4#{176}, as lower temperatures gradually cause the anti-HUH serum to deteriorate, i.e., the binding power will decrease. Thus, even with 1311-labeled HUH of high bindability, the initial B :F ratio will be low, and the slope of tile standard will be reduced drastically. References 1. 2. 3. 4. 5. 6. 7. 8.
Greenwood, F. C., Hunter, W. M., amid Glover, .J. S., The preparation of “I-labelled Imumimaim growth hormone of high specific radioactivity. Biochenm. J. 89, 114 (1963). Glick, S. M., Roth, J., Yalow, IL, and Bersomm, S. A., Immunoassay of human growth hormone in plasma. Nature 199, 784 (1963). Hunter, W. M., and Greenwood, F. G., A radio-imnmunoelectrophoretic assay for humnami growth hmornione. Bioche-nm. J. 91, 43 (1964). Hunter, W. M., and Greemmwood, F. C., Preparation of iodine 131-labelled human growth hormone of high specific activity. Nature 194, 495 (1962). Utiger, R. D., Extraction and radio-iimimuiioassay of growth hormnomme in human serum. J. GUn. Endocrinot. Metab. 24, 60 (1964). Frantz, A. G., and Rabkimm, M. T., Effects of estrogen and sex difference omi secretion of human growth hormone. J. Clii,. Endocrinol. Metab. 25, 1970 (1965). Porter, R. IL, Time hydrolysis of rabbit y-globulin and antibodies with crystallimme papain. Bioeleern. J. 73, 119 (1959). Bersomi, S. A., Yalow, R. S., Ghick, S. M., and Roth, .1., I,mimumioassay for protein amid peptide hormones. M tab. Cli,,. E.i-ptl. 13, 1135 (1964).
