ALFRED E. WILHELMI*
Department of Biochemistry,
Emory University, Atlanta, Georgia30322
CANINE GROWTH HORMONEt
The recent development of methods of radioimmunoassay of a number of
protein and peptide hormones, including human growth hormone, has ex-
cited interest in applying these methods to the study of the physiology of
the hormones in the common laboratory animals. The present work was
undertaken with the objective of isolating and purifying growth hormone
from the pituitary of the dog, in order to provide material suitable for
making specifiic antisera, for radioiodination, and for use as a standard in
radioimmunoassay. This paper describes a simple procedure for the isola-
tion and purification of canine growth hormone in excellent yield. In addi-
tion, certain side fractions are obtained from which other active principles
of the dog anterior pituitary may be isolated.
MATERIALS AND METHODS
Fresh frozen dog pituitaries were obtained mainly from Pel-Freeze. A few small
lots were collected from animals used in a variety of experiments. The glands were
stored in the laboratory at -20°C. until sufficient amounts were collected to make a
Reagent grade chemicals were used throughout.
The bioassays were all carried out by testing a well-characterized reference prepara-
tion, at two doses, simultaneously with the unknown, also at two doses. The results
were calculated by standard statistical methods for parallel line bioassays.'
Growth hormone activity was measured by the 10-day weight gain test in 100 gm.
female hypophysectomized rats,2 as modified in this laboratory.8 The systemic crop
sac test in white Carneau pigeons was used for measurement of prolactin.' Follicle-
stimulating hormone was measured by the Steelman-Pohley assay.' Luteinizing hor-
mone was measured by the ovarian ascorbic acid depletion test of Parlow.' Thyroid-
stimulating hormone activity was measured by the method of Lamberg, involving 8P
uptake by the thyroids of day-old chicks.7
Polyacrylamide disc-gel electrophoresis was carried out by the method of Ornstein
and Davis.8 The analysis for amino-terminal amino acids was made by the method of
Fraenkel-Conrat, et al.9 Carboxyl-terminal amino acids were determined by digestion
with carboxypeptidase in the presence of 1% sodium dodecyl sulfate,'0 and with the
aid of a Beckman Model 120B amino acid analyzer. Amino acid analysis was done by
the method of Spackman, Stein, and Moore,"1 on 24, 48, and 72 hour hydrolysates of
the protein. Hydrolysis was carried out in 6N HCI in sealed evacuated tubes in an
*Professor of Biochemistry.
t The work reported in this paper was supported by research grants, HD 01231
and AM 03598, from the National Institutes of Health, U. S. Public Health Service.
This report is publication No. 877 of the Division of Basic Health Sciences, Emory
YALE JOURNAL OF BIOLOGY AND MEDICINE
oven at 105°C. Samples in duplicate for each period were drawn from the same solu-
tion of the protein; other samples were taken in triplicate for the determination of
Preparation of growth hormone. All operations except chromatography, which is
done at room temperature, are carried out at the bench at the temperature of melting
ice, in a cold room at 4°C., or in a refrigerated centrifuge (Sorvall RC-2).
The glands, taken from the deep-freeze, are partly thawed, weighed, and homogen-
ized in small portions in 0.1 M ammonium sulfate, pH 8.5, using a Kontes conical
glass homogenizer driven by an Eberbach high-torque motor. All glassware is kept
chilled in cracked ice. After the glands are homogenized, the glassware is rinsed with
solvent and the homogenate is diluted with sufficient solvent to make a mixture of
5 ml. of solvent per gram fresh weight of glands. The pH is adjusted to 8.5 and the
mixture is stirred magnetically in the cold room for one hour, then centrifuged at
16,000 x g for one hour.
The clear, deep red supernatant solution is decanted and saved in a graduated cyl-
inder. The residue is re-suspended in 0.1 ammonium sulfate, pH 8.5 (5 ml. per gm.
fresh weight of glands), and the suspension is stirred for an hour and centrifuged as
before. The clear pink supernatant solution is added to the first extract and the vol-
ume is noted. The residue is stored in the deep freeze.
To the combined extracts, vigorously stirred, sufficient solid ammonium sulfate is
added slowly to make the system 0.8 M in ammonium sulfate, and the pH is adjusted
to 7.0. After an hour, the solution is centrifuged, as before, and the supernatant solu-
tion is decanted. The precipitate (Fraction A) is taken up in distilled water, dialyzed
against distilled water until salt-free, and lyophilized.
To the clear red supernatant solution, vigorously stirred, sufficient solid ammonium
sulfate is added slowly to make the system 1.8 M in ammonium sulfate. The resulting
precipitate flocculates and settles rapidly, and after one-half hour it is centrifuged as
before. The supernatant solution is decanted and set aside, and the precipitate is taken
up in water (5 ml. water per gm. fresh weight of glands), and dissolved by adjusting
the pH to 8.5. The clear solution, vigorously stirred., is adjusted to pH 4.0, and suf-
ficient solid ammonium sulfate is added slowly to make the system 1.25 M in am-
monium sulfate. The precipitate, which flocculates and settles readily,
off as before. The supernatant solution is decanted, and the precipitate is taken up in
water, neutralized, dialyzed and lyophilized (Fraction B).
Useful side fractions are recovered from the supernatant solutions at pH 7 and pH
4 by adding solid ammonium sulfate sufficient to make the concentration of the salt
4 M. The precipitate at pH 7 (Fraction C) is a source of FSH; that at pH 4 (Frac-
tion D) is a source of TSH. Both precipitates are recovered by centrifuging, dialyz-
ing, and lyophilizing as usual.
Fraction B contains the bulk of the growth hormone in the extract. It is purified
further by suspending it (lg/100 ml. solvent) in 0.05 M ammonium sulfate, pH 9.0,
and stirring for an hour, then centrifuging for
residue is taken up in water, dialyzed and lyophilized (Fraction E). The clear, nearly
colorless supernatant solution is adjusted to pH 7.0, and solid ammonium sulfate is
added slowly, with vigorous stirring, in an amount sufficient to make the system
1.8 M in ammonium sulfate. The solution is allowed to stand for an hour, then cen-
trifuged. The precipitate is taken up in water, dialyzed and lyophilized (Fraction F).
Solid ammonium sulfate is added to the supernatant solution in an amount sufficient
to raise the concentration of ammonium sulfate to 3 M, and after standing overnight,
1 hour at 16,000 x g. The colored
Caninte growth hormone
the mixture is centrifuged. The precipitate is taken up in water, dialyzed and lyophil-
ized (Fraction G). The supernatant solution is discarded.
Fraction F is purified further by DEAE-cellulose chromatography. The fraction is
dissolved in 0.01 M Tris-formate buffer, pH 8.0 (lg/100 ml.) and is first treated with
sufficient 10% di-isopropylfluorophosphate (DFP) in isopropanol to make the mix-
ture 0.001 M in DFP. This treatment is necessary in order to suppress enzymatic
activity which, as reported by Lewis and Cheever,' can cause partial decomposition
of the hormone during chromatography. The DFP-treated solution is stirred vigor-
ously under a hood for 15-30 minutes and is then pumped onto a column of DEAE-
cellulose equilibrated with 0.01 M Tris-formate buffer, pH 8.0. The column is then
washed with buffer, and the hormone (Fraction H) is eluted by a step-wise increase
in buffer concentration to 0.1 M. A second small peak is eluted at 0.2 M Tris-formate,
pH 8.0, and a large inert peak is eluted at 0.2 M Tris-formate, pH 8.0, 0.2 M NaCl.
The contents of the tubes containing each fraction are combined, dialyzed and lyophil-
ized. The elution pattern in a typical experiment is shown in Figure 1.
FIG. 1. Chromatography of canine growth hormone (Fraction F) on DEAE-cel-
lulose. The preparation was dissolved (1%) in 0.01 M Tris-formate, pH 8.0, and ap-
plied to the column. After washing with the same buffer, the buffer concentration is
increased in steps, as indicated. The growth hormone is eluted (fraction a, as shown)
at 0.1 M Tris-formate, pH 8.0. The column was operated at room temperature. Flow
rate, 60 ml/hour; fraction size, 10 ml. Tubes comprising fractions a, b, c, and d, as
shown, were pooled, and the contents were dialyzed and lyophilized.
YALE JOURNAL OF BIOLOGY AND MEDICINE
RESULTS AND DISCUSSION
Table 1 presents a summary of the results of eight experiments with
amounts of fresh dog pituitaries ranging from 1.9 to 132 gm. The yields
of each fraction have been fairly consistent from run to run, and the yields
and specific activities of the fractions containing the growth hormone (B,
F, and H) have been particularly high. Fractions C, D, E, and G have not
been assayed for growth hormone content because the patterns obtained in
polyacrylamide gel disc electrophoresis indicated either that growth hor-
mone was absent or was present in negligible amounts (much less, for
example, than in Fraction A). The disc gel patterns of the main fractions
TABLE 1. YIELDS AND GROWTH HORMONE ACTIVITIES OF FRACTIONS
Yield Growth actizity
(mg/gm. fresh weight)
I.U./gm. fresh weight LU./n1g.
NOTE: The yields are the averages of eight extractions of samples weighing 1.9, 20,
65, 103, 75, 132, and 48 gm. respectively. Figures in parenthesis give the range of the
yields. A fresh dog pituitary was found to weigh, on the average, 90 mg.
are illustrated in Figure 2. It will be seen that Fraction H is very nearly
homogeneous; the principal very dense growth hormone band is accom-
panied by a trace of faster moving material. From preliminary experiments
carried out on a sample of Fraction F not treated with DFP, this more
rapidly moving component can be identified with the principal product of
enzymatic activity on the fraction during chromatography. In such experi-
ments the hormone is obtained in two peaks of nearly equal size, and of
identical specific activity. The electrophoretic mobility of the main com-
ponent in the second peak is the same as that of the trace component seen
in Fraction H (Figure 2). The nature of the chemical change is not known
in detail; it may be attended by loss of one or more amide groups; and it
is not accompanied by any loss of biological activity. Fraction H may be
further purified by repeating the chromatography on DEAE-cellulose.
Elution is accomplished by increasing the buffer concentration only to
Volutite 41, October,1968
FIG. 2. Polyacrylamide disc gel electrophoresis patterns of fractions derived in the
extraction and chromatography of dog pituitaries. In each case, 0.1 mg. of the frac-
tion was applied to the gel. Fractions B, F, and H (fraction a of Figure 1) represent
the important growth hormone containing fractions. Fractions E and G yielded clear
gels containing no stained bands except the marker, resembling in this respect, Frac-
tion D, and they are therefore not shown.
Canine growth hormone
0.04 M Tris-formate, pH 8.0. The protein obtained from the center of the
eluted peak is essentially homogeneous by disc-gel electrophoresis.
Fraction F, which is obtained in very high yield, with about the same
potency as the bovine International Standard, may be particularly useful
for experiments in which a high order of purity is not required. One such
fraction has been tested for its content of other hormonal activities. These
proved to be (per mg. of the fraction), for LH, 0.0071 x NIH-LH-S1
(0.0052-0.0097); for FSH, less than 0.03 x NIH-FSH-S1; for TSH,
0.0070 I.U. (0.0050-0.0098). Other activities have not yet been accounted
for, but from the absence of a distinct band in the expected position after
disc electrophoresis, it may be estimated that the prolactin activity of this
fraction cannot be more than 0.5-0.7 I.U./mg.
There has not yet been time for exhaustive tests for hormonal contami-
nants of the highly purified dog growth hormone, Fraction H. (These
tests are very costly of material, and in the light of the excellent results
obtained with the immediate precursor of this fraction, the tests have been
deferred.) Prolactin activity (approximately 3 I.U./mg.) is recovered in
the small peak eluted at 0.2 M Tris-formate, pH 8.0. From the yield and
specific activity of this fraction it may be calculated that most of the pro-
lactin activity in the starting material is accounted for. Fraction H itself,
therefore, is not expected to have any significant prolactin activity.
Lyophilized purified canine growth hormone is a fluffy white powder
forming colorless solutions. It is easily soluble in dilute buffers or salt
solutions at pH 8.0 or pH 4.0. The hormone precipitates maximally at
pH 6.3. In disc electrophoresis it migrates the same distance, relative to
the front, that porcine growth hormone does. Since the isoelectric point of
the porcine hormone is 6.3,14 that of the canine hormone is probably the
The principal amino-terminal residue, as determined by the fluorodinitro-
benzene method, is phenylalanine (94% of the total amino acids detected).
A small amount of alanine was the only other amino acid detected. The
data of two experiments are presented in Table 2.
Digestion with carboxypeptidase A leads, in 15 minutes, to the rapid
liberation of phenylalanine, accompanied by small amounts of several other
amino acids. Of these, and only in the presence of 1 o sodium dodecylsul-
fate, only alanine is set free in substantially greater quantities after 24
hours of digestion. The carboxyl-terminal sequence of canine growth hor-
mone is therefore-ala-phe-COOH, the same as that found for bovine,
porcine, and equine growth hormone."0'15 The data are summarized in
YALE JOURNAL OF BIOLOGY AND MEDICINE
TABLE 2. AMINO-TERMINAL RESIDUES OF CANINE GROWTH HORMONE
Amino acid found
(nanomols per mg. of protein)
I. 8.3 mg. protein
II. 8.9 mg. protein
(Values are uncorrected for loss.)
These are duplicate observations on a single sample of the purified hormone (Frac-
TABLE. 3. CARBOXYL-TERMINAL RESIDUES OF CANINE GROWTH HORMONE
Nanomols per mg. of protein
The sample, 8.70 mg., was dissolved in 2.00 ml. of 0.1% sodium dodecyl sulfate
with the aid of 0.23 ml. of 2 M ammonium bicarbonate, and 0.02 ml. of a solution of
(DFP-treated, Worthington Lot No. COA-70A)
0.2 mg. of the enzyme, was added. Incubation was at room temperature. Samples
(1.00 ml.) were withdrawn at 15 minutes and 24 hours, mixed with 1.00 ml. of 20%
trichloroacetic acid, and centrifuged. 1.90 ml. of the supernatant fluid were withdrawn
and placed on the long column of the amino acid analyzer.
A sample of the purified hormone has been treated, by the method of
Wallis,"6 for the isolation of the carboxyl-terminal peptide set free by diges-
tion with trypsin. The amino acid composition of this peptide-phe2, ser2,
ala, val, glu,Y2cys-isidentical to that found by Mills,' for the same pep-
tide in porcine and (as Mills has since ascertained"7) equine growth hor-
mone. The sequence of this peptide from the canine hormone is yet to be
determined, but its composition invites the presumption that it is the same
as that in pig and horse growth hormone.
The amino acid composition of canine growth hormone is presented in
Table 4. Since the molecular weight of the canine hormone has not yet
been determined, the data are presented in terms of mols amino acid found
per 1,000 gm. of protein. Recent observations have indicated that the
Canine growth hormone
TABLE 4. AMINO ACID COMPOSITION OF CANINE GROWTH HORMONE
Mols per 1,000 gm. protein
mols per 22,000 gm.
The sample, 22.54 mg., was dissolved in 25.0 ml. of 0.001 N NaOH. Six 2.0 ml.
samples were taken for hydrolysis for 24, 48, and 72 hours. Three 4.0 ml. samples
were taken for nitrogen determinations by micro-Kjeldahl. Nitrogen content of the
lyophilized protein was 15.19%o. The solution contained 0.1370 mg.N/ml. The hy-
drolyzed samples were evaporated to dryness, redissolved in 2.10 ml. of the first buf-
fer, and 1.00 ml. was applied to each column. Nitrogen content per sample: 0.1304 mg.
The data for the long columns were corrected on the basis of recovery of added nor-
leucine. The average recovery of nitrogen from the six runs was 103.2%. From the
total weight of amino acid residues detected and the nitrogen content of the applied
samples, a value of 16.39% was computed for the nitrogen content of the protein.
This value was used in calculating mols of amino acid found per 1,000 gm. of protein.
The values for half-cystine, serine, and threonine were extrapolated to zero time by
the method of least squares; that for leucine was calculated from the averages for the
48- and the 72 hour hydrolysates. Tryptophan was not determined. The figures at
each time interval are the averages of duplicate determinations, which did not in any
case differ from one another by more than 5%.
molecular weights of bovine, porcine, and human growth hormone may all
be of the same order, approximately 22,000.18.19 The last column of Table 4
gives the mols of amino acid per 22,000 grn. of protein, calculated to the
nearest integer, for the canine hormone. These figures differ from our
analysis of porcine growth hormone by two residues in respect of aspartic
acid and leucine, one residue in respect of glutamic acid, arginine, phenyl-
alanine, serine, threonine, proline and alanine, and are the same for lysine,
histidine, tyrosine, half-cystine, methionine, glycine, valine and isoleucine.
The amino acid composition of the two hormones is therefore very similar.
YALE JOURNAL OF BIOLOGY AND MEDICINE
Volume 41, October, 1968
The simple method of preparation of canine growth hormone described
in this paper is easily adapted to small quantities of starting material. Be-
cause of the high yield rate, it should be useful in preparing hormone for
physiological experiments and especially for use in radioimmunoassay.
Such studies are under way in the laboratories of those investigators who
so kindly supplied the dog pituitaries used in developing the method, and
they will be reported on in due course.
These experiments were undertaken at the suggestion of the late Professor Richard
C. de Bodo, who provided the first small lots of frozen dog pituitaries. His colleagues,
Drs. Norman Altszuler and Richard Steele, continued support of the work by the
purchase of glands from Pel-Freeze. Two small lots of glands were also obtained
from Dr. Roger Unger and Dr. Eugene Segre' of Syntex. The author is grateful
for the opportunity provided by these investigators. I am indebted to Dr. Leo E.
Reichert, Jr. for bioassays of the LH, FSH, and TSH activities of the fractions, to
Dr. John B. Mills for the observations on the terminal tryptic peptide of the hormone,
to Mr. Stewart Howard for the amino acid analyses, and to Misses Sylvia Scapa,
Annette Westphal, Mary Willcox, Margaret Yager, and Mrs. June Boyett for expert
Bliss, C. J.: Confidence limits for measuring the precision of bioassays. Bio-
metrics, 1956, 12, 491-526.
Marx, W., Simpson, M. E., and Evans, H. M.: Bioassay of the growth hormone
of the anterior pituitary. Endocrinology, 1942, 30, 1-10.
Parlow, A. F., Wilhelmi, A. E., and Reichert, L. E., Jr.: Further studies on the
fractionation of human pituitary glands. Endocrinology, 1965, 77, 1126-1134.
Riddle, O., Bates, R. W., and Dykshorn, S. W.: The preparation, identification
and assay of prolactin-a hormone of the anterior pituitary. Amer. J. Physiol.,
1933, 105, 191-216.
Steelman, S. L. and Pohley, F. M.: Assay of the follicle-stimulating hormone
based on the augmentation with human chorionic gonadotropin. Endocrinol-
ogy, 1953, 53, 604-616.
Parlow, A. F.: Bioassay of pituitary luteinizing hormone by depletion of ovarian
ascorbic acid. In, Human Pituitary gonadotropins, edited by A. Albert.
Springfield, C. C. Thomas, 1961, pp. 301-310.
Lamberg, B. A.: Radioactive phosphorus as indicator in a chick assay of thyro-
tropic hormone. Acta med. Scand., 1953, 145, Suppl. 279.
Ornstein, L. and Davis, B.
Products Industries, 1961.
Fraenkel-Conrat, H., Harris, J. I., and, Levy, A. L.: Recent developments in
techniques for terminal and sequence studies in peptides and proteins.
Methods in Biochemical Analysis, edited by D. Glick. New York, Interscience
Publishers, 1955, 2, 359-425.
Santome', J. A., Wolfenstein, C., and Paladini, A. C.: C-Terminal sequence of
amino acids in bovine growth hormone. Biochem. biophys. Res. Commun.,
1965, 20, 482-485.
Spackman, D. H., Stein, W. H., and Moore, S.: Automatic recording apparatus
for use in the chromatography of amino acids. Analyt. Chem., 1958, 30, 1190-
Hiller, A., Plazin, J., and Van Slyke, D. D.: A study of conditions for Kjeldahl
determination of nitrogen in proteins. J. biol. Chem., 1948, 176, 1401-1420.
J.: Disc Electrophoresis. Rochester,
Canine growth hormone
Lewis, U. J. and Cheever, E. V.: Evidence for two types of conversion reactions
for prolactin and growth hormone. J. bio. Chem., 1965, 240, 247-252.
Papkoff, H., Li, C. H., and Liu, W-K: The isolation and characterization of
growth hormone from porcine pituitaries. Arch. Biochem., 1962, 96, 216-225.
Mills, J. B.: C-Terminal sequence of pig growth hormone. Nature, 1967, 213,
Wallis, M.: A C-terminal sequence from ox growth hormone. Biochim. biophys.
Acta (Amst.), 1966,115, 423-428.
Mills, J. B.: Personal communication.
Dellacha, J. M., Enero, M. A., and Faiferman, I.: Molecular weight of bovine
growth hormone. Experientia (Basel), 1966, 22, 16-17.
Ellis, G. J., Marler, E., Chen, H. C., and Wilhelmi, A. E.: Molecular weight of
bovine, porcine and human growth hormone by sedimentation equilibrium.
Fed. Proc., 1966, 25, 348.