Radioimmunoassay of Saliva Estrone Sulfate in Pregnant Sows
Tadatoshi OHTAKI, Masaharu MORIYOSHI, Ken NAKADA, Toshihiko NAKAO, and Keiichiro KAWATA
Department of Veterinary Obstetrics and Gynecology, Faculty of Veterinary Medicine, Rakuno Gakuen University, Ebetsu, Hokkaido
(Received 28 November 1996/Accepted 12 May 1997)
ABSTRACT. The aim of this study was to establish radioimmunoassay (RIA) for saliva estrone sulfate (E 1S), and to elucidate changes in
saliva E1S during pregnancy in the sow. Saliva E1 S was extracted using a commercially available solid phase column, and the E1S fraction
obtained was subjected to RIA. The sensitivity of the RIA was 29.7 pg/tube. The intra- and inter-assay coefficients of variation were 5.5–
8.4% and 13.1–19.5%, respectively. Mean recovery for E1 S added to saliva samples was as high as 99.9%. A significant positive
correlation (r=0.54, n=69, p<0.01) existed between saliva and plasma E1 S concentrations. During gestation, the changing patterns of
saliva and plasma E1 S concentrations were essentially the same, and two peaks of E1S concentrations were observed, one around day 30
and another just before parturition, although E1S concentrations in saliva remained at only 2.4–38.1% (mean 11.4%) of those in plasma
E1S. Thus, the present study has made it possible to measure saliva concentrations of E 1 S and demonstrated a high degree of positive
correlation between saliva and plasma E1S concentrations. These results suggest that diagnosis of early pregnancy and of normal or
abnormal fetal development could be made by measurements of E1 S in saliva. — KEY WORDS: estrone sulfate, pregnancy, radioimmunoassay,
J. Vet. Med. Sci. 59(9): 759–763, 1997
In pregnant sows, estrone sulfate (E 1S) in peripheral blood saliva from the cotton. Saliva samples obtained were mixed
is first detectable around day 16 of pregnancy and fluctuates with sodium azide (5 mg/ml) and kept frozen at 20°C
showing 2 peaks, one on days 23–30 and another just before until assayed for E1S. Immediately after saliva collection,
parturition . Since Robertson and King  suggested blood samples were also collected from the median tail vein
that E1S concentration in the maternal blood reflects the into heparinized test tubes. Plasma was separated by
amount of estrogens synthesized and secreted by the immediate centrifugation (1,700 × g, 15 min) and kept
blastocyst or fetus , early pregnancy diagnosis [2, 19, frozen at 20°C until assayed for E1 S.
20] and the diagnosis of normal or abnormal fetal Extraction of E1S: Saliva and plasma E1S was extracted
development  could be made in terms of maternal blood by a slight modification of the method described by
E1S concentrations. Nakamura . Briefly, saliva and plasma samples (1–4
Because saliva samples can be more easily and safely ml) were diluted with 2 ml borate buffer (0.2 M, pH 8.0)
collected compared to blood samples, various steroid containing 0.05% bovine serum albumin and 0.06% bovine
hormones in saliva have been measured in humans, cows, γ-globulin (BSA buffer), and then poured into a 5 ml
horses and pigs [4, 5, 8, 9, 11–13, 15]. In pigs, saliva disposable syringe attached to solid phase columns (Sep-
progesterone and cortisol have been determined for early Pak®PLUS C18 Cartridge, Millpore Co., Ltd., U.S.A.). The
pregnancy diagnosis  and for an adrenocortical function sample solutions were allowed to flow into the columns at a
test , respectively. An attempt has been made to rate of 0.2–1 ml per min. The columns were then washed
determine saliva E1S in horse  but not in the pig. with 4 ml distilled water and with 3 ml diethylether,
Therefore, the present study was undertaken to measure successively and the effluents were discarded. Finally, five
saliva concentrations of E1S by radioimmunoassay (RIA) ml acetone was poured into the columns, and the resultant
and to correlate the changing pattern of saliva E1S with that effluent (E1S fraction) was recovered into test tubes. The
of plasma E1S in pregnant sows. effluents were evaporated to dryness at 50°C under nitrogen
gas, and the residues were dissolved in 200 µl of BSA
MATERIALS AND METHODS buffer.
RIA for E1 S: Saliva and plasma E1S were measured by a
Saliva and blood collections: Five hybrid parous sows modification of RIA described by Nakamura , as shown
raised in this laboratory were used for the present study. in Fig. 1, using 1, 3, 5 -Estratrien-3-ol-17-one sulfate
Saliva samples were collected from each sow at intervals of (E1 S) as a standard, estrone sulfate ammonium salt [6.7–
3H(N)] (3H-E S; specific activity, 1.48–2.22 TBq/mM) as a
2–10 days during pregnancy, as described previously . 1
Briefly, a wooden chopstick (approximately 20 cm long) tracer and anti-estrone-3-sulfate rabbit serum (anti-E 1S
tipped with 2 g of absorbent cotton was inserted into the serum). The standard E1S preparation was provided by Dr.
mouth. Subsequently, the cotton soaked with sufficient A. Kanbegawa (Teikoku Hormone Manufacturing Co.. Ltd.,
amounts of saliva was removed from the stick and Japan) and 3H-E1S was prepared by Daiichi Pure Chemicals
compressed in a 10–20 ml disposable syringe to squeeze the Co.. Ltd., Japan. The anti-E1 S serum was generated by
760 T. OHTAKI, ET AL.
Fig. 2. The standard curve for radioimmunoassay of
estrone sulfate (E1 S) .
solutions were mixed 250 µl of dextran-coated charcoal
solution [0.2 M borate buffer solution containing 0.5%
charcoal (Norit A; Kishida Chemical Industries, Ltd., Japan)
and 0.05% dextran (DextranT70; Pharmacia Fine Chemicals
AB, Uppsala, Sweden), pH 8.0], left at 4°C for 10 min and
subsequently centrifuged at 4°C for 15 min (1,700 × g).
Two-hundred µl of the supernatant was emulsified with 3
ml scintillator (Scintisol ® EX-H, Wako Pure Chemical
Industries, Ltd., Japan), and radioactivity was assayed by a
Fig. 1. Flowchart of the procedure in RIA method. scintillation counter (Liquid scintillation system LSC-703,
Aloka Co., Ltd., Japan). The quantities of E1S in saliva and
Cosmo Bio Co., Ltd. Japan using anti 6-oxo-estrone-3- plasma samples were estimated by interpolation from a
sulfate-6-CMO-BSA as the antigen. According to the standard curve.
information from the manufacturer, the cross-reactivities of
the anti-E1S serum with E1S, estrone glucuronide, estrone, RESULTS
estradiol-3-sulfate, estradiol-3-glucuronide and estriol
sulfate were 100, 5.4, 4.0, 1.5, 0.8 and 0.8%, respectively. Parameter for reliability of E1S RIA: The standard curve
Prior to use, standard E1S and 3H-E1 S were diluted with obtained by the assay of standard E 1S in quantities ranging
BSA buffer at concentration of 3.9–1,000 pg/200 µl and from 3.9 to 1,000 pg is shown in Fig. 2. The sensitivity
250,000 cpm/ml, respectively. Anti-E1S serum was diluted was estimated as 29.7 pg/tube. Recovery of E1S in swine
1:36,000 with BSA buffer for use. saliva samples was determined by addition of 500, 1,000 or
Standard E 1S (3.9–1,000 pg) and saliva or plasma extracts 3,000 pg E1 S to 2 ml saliva samples, as shown in Table 1.
dissolved in 200 µl BSA buffer were each mixed with 100 Mean recovery ranged from 92.3 to 108.4%, and for all the
µl anti-E 1S serum solution and allowed to react at 4°C for 9 measurements the value was 99.9%.
12–24 hr. The mixtures were then incubated with 100 µl The intra- and inter-assay coefficients of variation were
3H-E S solution at 4°C for 12–24 hr. The incubated each calculated from quintuplicate results on 2 pooled saliva
Table 1. Recovery of varying amounts of estrone sulfate (E 1S) added to
swine saliva samples
Added E1 S
No. of assays Assay value (pg/ml) Recovery rate (%)
0 3 11 ± 0.6* –
250 3 241 ± 21.6 92.3
500 3 506 ± 57.4 99.0
1,500 3 1,638 ± 162.4 108.4
Total mean 99.9
* Mean ± S.E.
RIA OF SALIVA E1 S IN PREGNANT SOWS 761
from a plot of the two parameters is given in Fig. 3. A
significant positive correlation existed between saliva and
plasma E1S concentrations. The correlation coefficient was
Changes in saliva and plasma E 1S concentrations during
pregnancy: As shown in Fig. 4, mean E1S concentrations in
saliva and plasma fluctuated exhibiting two peaks. The
first peak of E 1S concentrations was evident on days 26–32
in saliva and on days 24–30 in plasma. The steroid in both
saliva and plasma remained at baseline levels during mid-
pregnancy. Saliva and plasma E1S, however, began to rise
gradually around days 70–80 and reached a peak
concentration (the second peak) on days 110–114 (just
before parturition). Thus, the changing patterns of saliva
E 1S concentrations resembled closely the plasma pattern,
although E 1S concentrations in saliva remained at only 2.4–
38.1% (mean 11.4%) of those in plasma.
Fig. 3. Correlation between saliva and plasma estrone sulfate DISCUSSION
(E1S) concentrations in sows .
In the present study, first the reliability of the RIA method
samples with different mean E1S concentrations. The intra- as a procedure for measuring saliva E1 S in sows has been
assay coefficient of variation was 5.5% for the sample with demonstrated. Variations in individual recovery rates for
the lower E1S concentrations (302 pg/ml) and for the other E 1S added to saliva samples are in small range, and mean
sample with the higher E1S concentrations (2,233 pg/ml) it recovery for steroid is as high as 99.9%. The intra- and
was 8.4%, while the inter-assay coefficient of variation for inter-assay coefficients of variation obtained for saliva E1S
the sample with the lower E1S concentrations (324 pg/ml) RIA in the present study are rather small and closely
was 19.5% and 13.1% for the sample with the higher E1 S comparable with the values reported for plasma E 1S assay
concentrations (2,598 pg/ml). in cows  and humans . The sensitivity of the assay is
Correlation between saliva and plasma E 1S: Correlation 29.7 pg/tube and high enough to measure E1S concentrations
between saliva and plasma E1 S concentrations were not only in the plasma but also in the saliva.
examined in 69 pregnant sows. The regression line obtained Two peaks of plasma E1S concentrations are evident in
Fig. 4. Changes in saliva and plasma estrone sulfate (E 1S) concentrations during pregnancy in sows.
762 T. OHTAKI, ET AL.
the present study, the first peak around day 30 and the REFERENCES
second peak just before parturition. This agrees quite well
with the finding of Robertson and King , although they 1. Ciotti, P. M., Franceschetti, F., Bulletti, C., Jasonni, V. M.,
carried out serial estimations of plasma E1S in only a single and Bolelli,G. F. 1989. Rapid and specific RIA of serum
gilt at implantation, throughout pregnancy, and at estrone sulfate with selective solid phase extration. J. Steroid
Biochem. 32: 473–474.
2. Cunningham, N. F., Hattersley. J. J. P., and Wrathall, A. E.
The present study has first described changes in saliva 1983. Pregnancy diagnosis in sows based on serum oestrone
E 1S concentrations during pregnancy in the sow. Moriyoshi sulphate concentration. Vet. Rec. 113: 229–233.
et al.  have reported that saliva and plasma progesterone 3. Edqvist, L. -E. and Stabenfeldt, G. H. 1991. Clinical repro-
concentrations are nearly comparable in pregnant sows. ductive endocrinology. pp. 668–673. In: Clinical Biochemistry
However, E 1S concentrations in saliva remain at only 2.4– of Domestic Animals, 4th ed. (Kaneko, J. J. ed.), (Kubota, S.
38.1% of those in plasma in present study. This is consistent and Tomoda, I. translaters), Kindai-shuppan, Tokyo (in Japa-
with the previous finding that only a small fraction of blood nese).
E 1S can pass into saliva in pregnant horses . Thus, in 4. Gao, Y., Short, R. V., and Fletcher, T. P. 1988. Progesterone
contrast to the striking similarity in progesterone concentration in plasma, saliva and milk of cows in different
concentrations between swine saliva and plasma, a great reproductive states. Br. Vet. J. 144: 262–268.
5. Harris, B., Lovett, L., Roberts, S., Read, G.F., and Riad-
difference in E 1S concentrations between saliva and plasma
Fahmy, D. 1993. Cardiff puerperal mood and hormone study.
are apparent in pregnant sows and horses. In general, steroid 1. Saliva steroid hormone profiles in late pregnancy and the
hormones are transported in the blood through association puerperium: Endocrine factors and parturition. Hormone Res.
with various types of proteins, and the free form usually 39: 138–145.
less than 10% of the total amount of hormone in the plasma 6. Hatzidakis, G., Katrakill, K., and Krambovitis, E. 1993. De-
. The steroid hormones must be in the free or unbound velopment of a direct and specific enzymeimmunoassay for
form before it can pass into saliva . This is the measurement of oestrone sulfate in bovine milk. J. Reprod.
accomplished by the establishment of an equilibrium Fertil. 98: 235–240.
between bound and free form in plasma. The lower 7. Kasman, L. H., Hughes, J. P., Stabenfeldt, G. H., Starr, M. D.,
concentrations of saliva E 1S measured in the present study and Lasley, B. L. 1988. Estrone sulfate concentrations as an
indicator of fetal demise in horses. Am. J. Res. 49: 184–187.
might be related to fairly tight association of the greater
8. Kirkpatrick, S. W., Campbell, P. S., and Robinson, S. L.
amount of the steroid with plasma proteins in the blood  1993. Salivary testosterone in children with and without learn-
which prevents the plasma E1S molecules from passing into ing disabilities. Physiol. Behav. 53: 583–586.
saliva. 9. Moriyoshi, M., Tanaka, Y., Nakao, T., and Kawata, K. 1994.
It has been suggested that E1S concentrations in maternal Early pregnancy diagnosis in the sow by saliva progesterone
blood reflect not only the amount of estrogens synthesized measurement using a bovine milk progesterone test EIA kit.
and secreted by the blastocyst or fetus  but also the J. Reprod. Dev. 40: j87–j92 (in Japanese with English sum-
litter size . On the other hand, Kasman et al.  mary).
suggested that the amount of E 1S in the urine may be of 10. Nakamura, K. 1990. Radioimmunoassay of estrone sulfate in
best use as an indicator of fetal demise in nontractable, bovine plasma. School of Veterinary Medicine, Obihiro Uni-
versity of Agriculture and Veterinary Medicine, Thesis for
feral, or exotic equine species. As described above, E 1 S
Bachelar’s Degree (in Japanese).
concentrations remain much lower in saliva than in the 11. Parrott, R. F., Misson, B. H., and Baldwin, B. A. 1989. Sali-
blood, and concentrations of the steroid in individual saliva vary cortisol in pigs following adrenocorticotrophic hormone
samples tend to represent slightly wide variations. However, stimulation: Comparison with plasma levels. Br. Vet. J. 145:
E 1S concentrations in the saliva are positively correlated 362–366.
with those in the plasma, and the changing patterns of saliva 12. Quissell, D. O. 1993. Steroid hormone analysis in human
and plasma E 1S concentrations are essentially the same. saliva. Ann. New York Acad. Sci. 694: 143–145.
Since, compared to the blood or urine collection, the saliva 13. Riad-Fahmy, D., Read, G. F., Walker, R. F., Walker, S. M.,
collection can be performed more readily and safely, with and Griffiths, K. 1987. Determination of ovarian steroid hor-
little or no stress, diagnosis of early pregnancy and of normal mone levels in saliva. An overview. J. Reprod. Med. 32:
or abnormal fetal development could be made in terms of
14. Robertson, H. A. and King, G. J. 1974. Plasma concentra-
measurements of E 1S in saliva but not in blood. For this
tions of progesterone, oestrone, oestradiol-17 beta and of
purpose, the more sensitive and simple method for oestrone sulphate in the pig at implantation, during preg-
measurement of saliva E1S such as Hatzidakis et al.  has nancy and at parturition. J. Reprod. Fertil. 40: 133–141.
reported should be preferably developed in the near future. 15. Sist, M. D., Youngblood, M. A., Williams, J. F., and Glade,
M. J. 1988 Salivary and serum estorone sulfate levels in preg-
ACKNOWLEDGMENT. The authors wish to thank Dr. nant mares. J. Equine Vet. Sci. 8: 164–167.
A. Kanbegawa for providing the standard estrone sulfate. 16. Stabenfeldt, G. H. 1992. The endocrine system. pp. 369–387.
In: Textbook of Veterinary Physiology (Cunningham, J. G.
ed.), W. B. Saunders Company, U.S.A.
17. Stone, B. A., Seamark, R. F., Godfrey, B. M., Quinn, P., and
Lloyd, B. 1986. Oestrone sulphate levels in plasma of sows
RIA OF SALIVA E1 S IN PREGNANT SOWS 763
as a basis for prediction of litter size at term. Anim. Reprod. application to early pregnancy diagnosis in pigs. Br. Vet. J.
Sci. 11: 51–62. 141: 60–68.
18. Stoner, C. S., Bazer, F. W., Thatccher, W. W., Wilcox, C. J., 20. Tamamura, F., Kawata, K., Nakao, T., and Tsunoda, N. 1982.
Combs, G. E., Knight, J. W., Wettemann, R. P., and White, Enzyme immunoassay of serum estrone sulphate applied to
C. E. 1986. Relationship between estrone sulfate in plasma swine early pregnancy diagnosis. J. Coll. Dairying 9: 337–
and litter size at forrowing for sows and gilts. Theriogenology 356.
25: 709–720. 21. Vining, R. F., McGinley, R. A., and Symons, R. G. 1983.
19. Sugiyama, S., Nakao, T., Tsunoda, N., and Kawata, K. 1985. Hormones in saliva: Mode of entry and consequent implica-
An enzymeimmunoassay of serum oestrone sulphate and its tions for clinical interpretation. Clin. Chem. 29: 1752–1756.