Bovine Serum Growth Hormone, Corticoids and Insulin During Lactation 1 J. A. KOPROWSKI, 2 AND H. ALLEN TUCKER Animal Reproduction Laboratory, Department of Dairy Science, Michigan State University, East Lansing, Michigan 48823 ABSTRACT. Serum GH and insulin were measured in 26 cows and serum corticoids were measured in 12 cows. Blood samples were collected 2-4 hr before, immediately after and 1 hr after for the duration of a lactation (44 weeks). Milking was associated with the release of corticoids and insulin but not GH. Serum GH was negatively associated with milk yield, especially after 24 weeks of lactation. This negative, relationship may be more directly caused by changes associated with concurrent pregnancy rather than with lactation per se, because serum GH concentrations decreased as lactation advanced, but increased as gestation progressed. IN GENERAL, prolactin, growth hormone (GH), corticoids, insulin, thyroid hormones and parathyroid hormone are considered essential for maintenance of lactation in laboratory species (1). We have recently reported a positive association between serum prolactin and milk yield in dairy cattle (2), and this report describes the changes in serum concentrations of GH, corticoids and insulin in some of the same serum samples previously assayed for prolactin. In addition to correlating these estimates of serum hormones with milk yield, the influence of stage of lactation, season, gestation, fetal sex, parity and stage of the estrous cycle on GH, corticoids and insulin were determined. Materials and Methods Sera from 26 cows which completed 44 weeks of lactation were assayed for GH and insulin. Twelve of the 26 cows were randomly selected and their sera, collected throughout lactation, Received January 9, 1973. 1 Journal Article No. 6207 from the Michigan Agricultural Experiment Station. This research was supported in part by USPHS Grant No. AM- 15899. 2 National Institutes of Health Predoctoral Fellow (S-F01-GM 42220). 645 Serum GH was greater (p < 0.05) during estrus than during the luteal phase of the estrous cycle of lactating cows. Serum insulin concentrations were inversely related to GH levels (increasing as lactation advanced, but decreasing as pregnancy progressed), and insulin was consistently negatively related to milk yields. Serum corticoids were variable, but on the average they changed very little as lactation advanced. Except for the overall correlation between serum corticoids measured immediately after and milk production (r = 0.19), serum corticoids generally were not significantly correlated with milk yields. {Endocrinology 93: 645, 1973) were assayed for total corticoids. Parturition occurred in these cows between August, 1969 and April, 1970. Beginning 1 month postpartum, each cow was bled at 4-week intervals through 44 weeks of lactation. In addition, sera of 12 of the lactating cows on the day of first postpartum estrus (day 0), on days 2, 4, 7, 9, 11, 15, 18, 20 and again on the day of the second postpartum estrus (day 00) were assayed for GH. On the day of bleeding, blood samples were collected 2-4 hr before the PM, immediately (within 5 min) after the machine was removed, and 1 hr after the PM. The cows were milked twice daily at equal intervals, and the total milk produced by each cow on the day before and on the day after blood sampling was used as the measure of milk production. Cows were fed normal forages consisting of corn silage and alfalfa haylage ad libitum, with alfalfa hay limited to 2.3 kg daily. Grain concentrates were fed at 8-9 AM and at 1-2 PM at the rate of 1 kg concentrate per 3 kg milk produced. The total ration was balanced for body maintenance and milk production requirements. Serum GH was also measured in 28 nonlactating Holstein heifers in samples collected between 1200 and 1400 hr at 30-day intervals beginning 30 days after conception and continuing through 270 days of gestation. The cows and heifers were bled via tail venipuncture using a 20 ml BD Vacutainer (165 X 16 mm, Becton, Dickinson and Co., Rutherford,
646 KOPROWSKI AND TUCKER Endo Vol 93 1973 No 3 N.J.) and a 1.5 inch 20 gauge needle. To minimize potential metabolism of corticoids by red blood cells blood samples were centrifuged within IS min of collection at 6,500 X g for 15 min, and the supernatant was stored at 20 C until assayed for hormone content. Upon thawing for hormone assay, most samples contained large fibrin aggregates which were removed by centrifugation. Bovine serum GH was estimated by radioimmunoassay as developed in our laboratory by Purchas et al. (3). Bovine GH (NIH-B12, National Institutes of Health, Bethesda) was used as the reference standard. Competitive protein binding (5) was used to quantify total serum corticoids. Insulin concentrations were determined by double antibody radioimmunoassay (4), similar to that used for GH. Guinea pig antibodies against bovine insulin (Lot No. 21) were obtained from Miles Laboratories (Kankakee, Illinois), standard bovine insulin (Lot No. 795372) was provided by Eli Lilly & Co. (Indianapolis, Indiana), and 125 I bovine insulin was purchased as needed from Amersham Searle (Arlington Heights, Illinois). The second antibody prepared against guinea pig gamma globulin was from the same sheep used in the GH assay (3). Bovine prolactin, GH, LH and ovine FSH at concentrations up to 500 ng did not interfere with the insulin radioimmunoassay. Recoveries of varying quantities of standard inte 16 20 24 28 32 Week of lactation 36 40 44 FIG. 1. Average total corticoid concentrations during lactation in serum collected 2-4 hr before (A A), S min after ( ), and 1 hr after (O O). Pooled standard error of mean was 0.3. Number of observations per point was 12. sulin added to bovine serum ranged from 96 to 102%. The data were subjected to analysis of variance, correlation analyses and covariance analysis using stage of lactation, pregnancy and season as covariates. Results Serum GH, corticoids and insulin during lactation unadjusted. Serum GH (ng/ml) did not change significantly (p > 0.05) as lactation progressed averaging 4.2 ± 0.2 before, 4.4 ± 0.2 immediately after, and 4.2 ± 0.2 1 hr after. Prior to corticoids averaged 3.4 ± 0.5 ng/ml at 4 weeks of lactation, peaked at 20 weeks (9.3 ± 2.4 ng/ml) then decreased gradually to 4.3 ± 0.5 ng/ml at 44 weeks of lactation (Fig. 1). Corticoids in samples collected immediately after averaged 10.8 ±1.4 ng/ml at 4 weeks, increased to 16.2 ± 1.7 ng/ml at 12 weeks then gradually decreased to 8.7 ± 1.6 ng/ml at the end of lactation. At 1 hr after the corticoids were lower and more stable than in pre- or immediately post samples, averaging between 2.2 ± 0.2 and 3.3 ± 0.7 ng/ml throughout lactation. Overall, serum corticoids averaged 5.7 ± 0.4 ng/ml for pre- samples, 11.1 ± 0.6 ng/ml for samples immediately after and 2.6 ± 0.1 ng/ml 1 hr after, and differences among these means were significant (p < 0.01). Serum insulin increased approximately 2-3-fold between the 4th and 12th week of lactation but thereafter remained relatively stable (Fig. 2). Overall, serum insulin (\n U/ml) averaged 74.5 ±3.5 before, 85.8 ± 4.2 immediately after and 69.7 ± 3.1 1 hr after. Insulin in the immediate post- sample was significantly (p < 0.01) greater than that in the pre- or 1 hr post- samples. Relationship between milk production and serum GH, corticoids and insulin. Disregarding stage of lactation, the overall correlation coefficients between serum GH in the three
BOVINE SERUM HORMONES 647 E 80 4 8 12 16 20 24 26 32 36 40 44 Week of lactation FIG. 2. Average insulin concentrations during lactation in serum collected 2-4 hr before (A A), 5 min after ( ), and 1 hr after (O O). Pooled standard error of mean was 2.1. Number of observations per point was 26. blood samples and milk yield were 0.01, 0.10 and 0.07 (p > 0.05). However, during the first 20 weeks of lactation within-stageof-lactation correlations between GH and milk yields were generally positively related, but none was significantly different from zero (Table 1). In contrast, between 24 and 44 weeks of lactation, the correlation coefficients were consistently negative and several were significantly different from zero. When stage of lactation was disregarded, the overall correlations between milk yield and serum corticoids were 0.14 and 0.01 (p > 0.05) for samples collected before and 1 hr after, respectively, whereas the coefficient was 0.19 (p < 0.05) for samples collected immediately after. Within stage of lactation, there were no consistent relationships between serum corticoids and milk yield, although a few coefficients were significantly different from zero (Table 1). The overall correlations between milk yields and serum insulin at the three sampling times were 0.31, 0.21 and 0.35 (p < 0.01). Similarly, the coefficients within each stage of lactation between milk production and serum insulin were consistently negative (Table 1). Serum GH, corticoids and insulin during lactation covariance adjustment. Since stage of pregnancy and season of the year could affect the concentrations of GH, corticoids and insulin, covariance analysis was used to adjust the data. The covariates were stage of pregnancy and months of the year (season). As in the unadjusted data, induced the release of corticoids and insulin and serum GH concentrations were unaffected. The adjusted data among the three sampling TABLE 1. Within-stage-of-lactation correlation coefficients between milk yield and three estimates of serum GH, corticoids and insulin Serum GH Serum corticoids Serum insulin Week of lactation 4 8 12 16 20 24 28 32 36 40 44 0.20 0.16 0.32 0.01 0.33-0.06-0.07-0.07-0.38-0.46 a -0.31 Pre Immediately after 0.20 0.10 0.12 0.19-0.19-0.12-0.06-0.39 n -O.45 a -0.29 1 hr after 0.20 0.11-0.01 0.03 0.26-0.09-0.10-0.43 a -0.36-0.43 11-0.28 Pre -O.65 a 0.27-0.21-0.10-0.03 0.16-0.22-0.18 0.24 0.15 0.67 a Immediately after -0.25-0.26 0.28 -O.72 b 0.50 0.09 0.36 0.01 1 hr after -0.03-0.05-0.44-0.27 0.35 0.30 0.27-0.25 0.53-0.16 0.30-0.18-0.16-0.04-0.22-0.06-0.33-0.27-0.24 -O.65 b Pre Immediately after 0.07-0.11-0.17-0.01 -O.52 b 0.01-0.21-0.25-0.07-0.42 a 1 hr after -0.01-0.12-0.15-0.19-0.08-0.19-0.12-0.26-0.39 a -0.36 a p< O.OS. b p<0.01.
648 KOPROWSKI AND TUCKER Endo Vol 93 1973 No 3 times were generally parallel within each hormone throughout lactation; therefore, the pattern of response during lactation was similar for the three sampling times. For simplicity and statistical purposes the adjusted data for each hormone were averaged across the three sampling times (Fig. 3). Adjusted GH means decreased linearly (p < 0.01) from 5.8 ng/ml at 4 weeks of lactation to 3.2 ng/ml at 44 weeks (Fig. 3). Except for an initial increase during the first 12 weeks of lactation, little change occurred in adjusted serum corticoid averages for the remainder of lactation (Fig. 3). In contrast, serum insulin increased linearly from 5 H-U/ml at 4 weeks to 138 ^U at 44 weeks of lactation. Effects of stage of gestation on serum GH, corticoids and insulin. Covariance analysis was used to determine the effects of gestation by using stage of lactation and season as the covariates. Results for the three sampling times at each 4-week interval were averaged, as described above, to obtain the results in Fig. 4. Serum GH increased linearly (p < 0.01) whereas insulin decreased linearly (p < 0.01). The irregular decline in corticoid concentration was not significant (p > 0.05). Effects of season, fetal sex and parity on serum GH, corticoids and insulin. Neither season of the year nor fetal sex significantly 4 8 12 16 20 24 28 32 36 40 44 Wttk of lactotion FIG. 3. Adjusted average serum GH, corticoids and insulin concentrations during lactation using stage of pregnancy and season of year as covariates. 2 3 4 5 6 Month of pregnancy FIG. 4. Adjusted average serum GH, corticoids and insulin concentrations during pregnancy using stage of lactation and season of year as covariates. affected (p > 0.05) serum GH, corticoids or insulin. Serum GH and corticoids were not influenced by parity (p > 0.05), whereas multiparous cows had greater (p < 0.01) concentrations of insulin (78.2 [x U/ml) than primiparous cows (59.3 M- U/ml). Serum GH during the estrous cycle in lactating cows. Concentrations of GH in sera collected before, immediately after, and 1 hr after were generally parallel, and were not significantly different from each other (p > 0.05). Therefore, the data were averaged across sampling times at the various days of the estrous cycle listed in Table 2. Serum GH was significantly higher (p < 0.05) during the estrogenic phase (day of estrus, day 20 and day of subsequent estrus) of the estrous cycle than during the luteal phase (days 2 through 18). Serum GH during gestation in nonlactating heifers. Serum GH did not change significantly (p > 0.05) in pregnant, nonlactating heifers (Table 3). As in the pregnantlactating cow, fetal sex did not significantly affect (p > 0.05) serum GH in nonlactatingpregnant heifers. Discussion Data from this study show that causes release of corticoids and insulin, but not of GH. Failure to find increases in serum
BOVINE SERUM HORMONES 649 TABLE 2. Average serum GH during the estrous cycle in lactating cows Day of estrous cycle 0 b 2 4 7 9 11 15 18 20 00 c Serum GH a 4.9 ± 0.8 d 3.1 ±0.2 3.3 ±0.2 3.6 ±0.3 3.2 ± 0.3 2.9 ±0.2 3.4 ± 0.4 3.5 ±0.3 3.9 ± 0.4 d 3.9 ± 0.4 d (ng/ml) 11 Mean ± SE with 36 observations per mean. b Day of first postpartum estrus. c Day of second postpartum estrus. d Orthogonal contrasts indicated that GH during the estrogenic phase of the estrous cycle was significantly greater (p < 0.05) than GH during the luteal phase (day 2-1S). GH after at any stage of lactation is in agreement with our previous study (6). The induced release of corticoids agrees with other reports (5,7), but to our knowledge this is the first report that induces the release of insulin. In contrast to our finding that the -induced release of prolactin gradually disappears as lactation advances (2), we found no evidence that cows lose their ability to discharge corticoids or insulin in response to. Most likely the induced release of corticoids in cows involves a neural reflex from the mammary gland to the hypothalamo-pituitary axis which results in the sequential release of ACTH-releasing factor, ACTH and finally corticoids (8,9). What triggers the release of insulin at is not known although changes in serum glucose, volatile fatty acids or other metabolites would be likely candidates in view of their known roles in affecting serum insulin (10,11). The time of feeding grain concentrates is another factor which may have affected serum insulin. The time between the PM feeding and collection of the serum sample immediately after ranged from 2-4 hr. Trenkle (12) observed a 23% increase in serum insulin within 4 hr of feeding sheep a mixed hay-grain diet, although in another similar experiment no change in serum insulin was observed. It seems unlikely that feeding pattern altered either pre- or 1 hr post- estimates of serum insulin. This reasoning is based on the fact that grain concentrate feeding occurred around the time pre- blood samples were being collected and preceded 1 hr post- samples by 3-5 hr. Yet serum insulin was almost identical in both samples (Fig. 2). Obviously the interrelations between serum insulin, metabolites, feeding patterns and stimuli in ruminants need additional clarification. The overall correlation between GH and milk yield was essentially zero, and when viewed on a within-stage-of-lactation basis there was a strong tendency toward a negative relationship between GH and milk yield. This latter response is puzzling, especially since GH administered to cattle stimulates lactation (1,13). However, stage of concurrent pregnancy also advanced as lactation progressed in these cows, and there was a coincident increase in serum GH associated with advancing pregnancy (Fig. 4). When the data for lactating cows were adjusted for stage of pregnancy, GH decreased as lactation advanced (Fig. 3). Perhaps the negative correlations within stage of lactation reflect a change in serum GH which is associated with pregnancy rather than lactation per se. However, these changes in serum GH must be associated with a type of interaction between lactation and pregnancy, because pregnancy did not affect serum GH in the absence of lactation (Table 3). TABLE 3. Average serum GH during pregnancy in heifers Days pregnant Serum GH 11 (ng/ml) 30 60 90 120 150 180 210 240 270 2.8 ±0.1 2.7 ±0.1 3.1 ±0.2 2.6 ±0.1 2.6 ±0.1 2.6 ±0.1 2.7 ±0.1 2.6 ±0.2 2.8 ±0.2 a Mean ± SE with 28 observations per mean.
650 KOPROWSKI AND TUCKER Endo Vol 93 1973 No 3 The significant overall correlation coefficient between milk yield and serum corticoids measured immediately after agrees with our previous observation that the highest correlations between milk yield and prolactin also occur in serum collected immediately after (2). Our failure to find consistent patterns in the correlations calculated within-stage-of-lactation in cattle agrees with similar findings between serum corticoids and lactational performance in lactating rats (14). These findings are consistent with failure to show any effect on lactation in cows after administering low doses of corticoids throughout lactation (15). The consistent negative correlations between milk production and serum insulin, are in agreement with the findings of several workers (1,16,17) that milk yields are reduced in response to insulin administration. However, the causes of the rising serum insulin during lactation, or the fall during gestation remain to be determined, but as previously discussed various metabolites influence serum insulin (10,11). It should be noted that grain concentrates were reduced as lactation and concurrent pregnancy advanced because milk yields were declining during this period. But it seems unlikely that the reduction in concentrates per se caused the changes in serum insulin during lactation or during pregnancy because the insulin responses were in opposite directions during these two physiological states. The inverse relationship between serum GH and insulin in cows during lactation (Fig. 3) or during pregnancy (Fig. 4) agrees with the data of Trenkle (18), who also observed an inverse relationship between serum insulin and GH in steers. Serum glucose and short chain fatty acid concentrations and utilization rates decrease in ruminants as lactation advances (19-20). Infusions of glucose or some short chain fatty acids stimulate insulin release (11), but these metabolites have little effect on serum GH in ruminants (10,21). Injections of insulin lower blood glucose (16), whereas GH injections have little effect on serum glucose (22) or glucose utilization (23). Considering these relationships between metabolites and hormones, serum GH might be expected to decrease and serum insulin to < increase with advancing lactation as we observed in the present study. But which factors are causes and which are effects remain unanswered. In any event, the changing patterns in serum concentrations between the two hormones depending upon whether the ( data are adjusted for stage of lactation or stage of pregnancy illustrates the necessity of cautious interpretation of serum hormone concentrations as explanations for control of mammary secretions. The greater concentrations of serum GH i during the estrogenic phase of the estrous cycle of cows is in agreement with similar measurements made in mice (24). Trenkle (18) also observed an increase in serum GH of steers fed diethylstilbestrol. But the role of serum GH in regulating lactational ^ performance of cattle during the estrous cycle is obscure because the greater concentrations of GH at this time did not affect milk yields (2). Estrogen excretion rates (25) and serum estrogens (26,27) increase during pregnancy. Thus, the increased GH in our lactating cows as pregnancy progresses may be related to this elevated estrogen secretion although it is not known why a similar increase in GH was not observed in the nonlactating pregnant heifers. Acknowledgments The valuable technical assistance of Mrs. Annemieke Ambrosier is gratefully acknowledged. References 1. Cowie, A. T., and J. S. Tindal, The Physiology of Lactation, The Williams & Wilkins Com- * pany, Baltimore, 1971, p. 161-165. 2. Koprowski, J. A., and H. A. Tucker, Endocrinology 92: 180, 1973. 3. Purchas, R. W., K. L. Macmillan, and H. D. Hafs, / Anim Sci 31: 358, 1970. 4. Grigsby, J. S., M.S. Thesis, Michigan State University, 1973. * 5. Smith, V. G., E. M. Convey, and L. A. Edgerton, J Dairy Sci 55: 1170, 1972.
BOVINE SERUM HORMONES 651 6. Tucker, H. A., / Anim Sci (Suppl 1) 32: 137, 1971. 7. Wagner, W. C, Am J Vet Med 154: 1395, 1969. 8. Denamur, R., M. Stoliaroff, and J. Desclin, Comp Rend 260: 3175, 1965. 9. Voogt, J. L., M. Sar, and J. Meites, Am J Physiol 216: 655, 1969. 10. Hertelendy, F., L.-Machlin, and D. M. Kipnis, Endocrinology 84: 192, 1969. 11. McAtee, J. W., and A. Trenkle, / Anim Sci 33: 43S, 1971. 12. Trenkle, A., / Nutrition 100: 1323, 1970. 13. Bullis, D. D., L. J. Bush, and P. B. Barto, / Dairy Sci 48: 338, 1965. 14. Thatcher, W. W., and H. A. Tucker, Proc Soc Exp Biol Med 134: 915, 1970. 15. Head, H. H., W. W. Thatcher, and C. J. Wilcox, / Dairy Sci 55: 700, 1972 (Abstract). 16. Kronfeld, D. S., G. P. Mayer, J. M. Robertson, and F. Raggi, / Dairy Sci 46: 559, 1963. 17. Schmidt, G. H., / Dairy Sci 49: 381, 1966. 18. Trenkle, A., J Anim Sci 31: 389, 1970. 19. Surve, A. H., / Dairy Sci 51: 954, 1968 (Abstract). 20. Bergman, E. N., and D. E. Hogue, Am J Physiol 213: 1378, 1967. 21. Trenkle, A., / Anim Sci 32: 111, 1971. 22. Manns, J. G., and J. M. Boda, Endocrinology 76: 1109, 1965. 23. Head, H. H., M. Ventura, D. W. Webb, and C. J. Wilcox, / Dairy Sci 53: 1496, 1970. 24. Sinha, Y. N., F. W. Selby, N. J. Lewis, and W. P. Vanderlaan, Endocrinology 91: 784, 1972. 25. Randel, R. D., and R. E. Erb, / Anim Sci 33: 115, 1971. 26. Wettemann, R. P., and H. D. Hafs, / Anim Sci 36: 51, 1973. 27. Smith, V. G., L. A. Edgerton, H. D. Hafs, and E. M. Convey, J Anim Sci 36: 391, 1973.