0021-972X/91/7203-0623S03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1991 by The Endocrine Society
Vol. 72, No. 3 Printed in U.S.A.
Human Milk Stimulates Prostacyclin Production by Cultured Human Vascular Endothelial Cells* ARI RISTIMAKI, OLAVI YLIKORKALA, KRISTINA PESONEN, JAAKKO PERHEENTUPA, AND LASSE VIINIKKA Children's Hospital (A.R., K.P., J.P., L. V.) and the Departments of Obstetrics and Gynecology (0. Y.), University of Helsinki, SF-00290 Helsinki, Finland
ABSTRACT. Prostacyclin (PGI2) is an antithrombotic and vasodilatory factor, which is produced mainly by the vascular endothelium. Little is known about how this process is regulated. We investigated the effect of human milk on PGI2 synthesis by human vascular endothelial cells by measuring its stable metabolite, 6-keto-prostaglandin Fia, by RIA. Human milk induced dose- and time-dependent stimulation of PGI2 production, whereas cow's milk was ineffective. The lowest concentration of human milk that stimulated the production of PGI2 was 0.1%, and 10% induced a 2.4- to 3.4-fold increase. The effect of human milk was detectable after 2 h and was blocked by inhibitors of transcription, translation, and cyclooxygenase. Boiling abolished
the activity, but acetone extraction enhanced it. A 10% concentration of acetone-extracted human milk stimulated the release of endothelial cell PGI2 by 6.6-fold. In human milk samples we found no correlation between the amount of immunoreactive epidermal growth factor (EGF) and the activity stimulating PGI2 synthesis. Furthermore, EGF antibodies did not inhibit the activity. This is the first demonstration that human milk stimulates PGI2 production by endothelial cells. We conclude that human milk is a potent inducer of PGI2 production by human vascular endothelial cells and that the stimulatory activity is not due to EGF. (J Clin Endocrinol Metab 72: 623-627, 1991)
P
ROSTACYCLIN (PGI2), a major arachidonic acid metabolite of human vascular endothelial cells, is a powerful vasodilator and a potent inhibitor of platelet aggregation (1). It may also accelerate wound healing (2), induce neovascularization (3), and regulate the fetal and neonatal circulation (4). Despite the importance of PGI2, little is known about the regulation of its production. Human milk is the ideal nutrient for a newborn infant. It contains numerous biologically active compounds that benefit the breast-fed baby. Many of these factors in human milk are polypeptides, which act as host defense agents, enzymes, and hormones (5). In addition, human milk contains growth factors, of which the major one is epidermal growth factor (EGF) (6). Recently, human milk was shown to stimulate PGI2 synthesis by cultured human skin fibroblasts (7) and rabbit aorta (8). The stimulatory activity appeared to be associated with a protein less than 10,000 in mol wt, which was suggested to be a growth factor (8). Interestingly, EGF enhances PGI2 production by human vascular endothelial cells (9). Since nothing is known about the effect of milk on the
production of PGI2 by its biologically most important source, vascular endothelium (1), we investigated the effect of human milk on the synthesis of PGI2 in cultured human vascular endothelial cells.
Materials and Methods Samples Human milk samples were obtained from the Human Milk Bank of the Children's Hospital, University of Helsinki. They were from milk excreted 2 days (n = 6), 2 weeks (n = 6), or 2 months (n = 6) after delivery. Pasteurized cow's milk and cow's milk-derived infant formula (Tutteli, Valio, Turenki-Janakkala, Finland) were commercial products. All milk samples were defatted, and the cells were removed by centrifugation at 15,000 X g for 30 min at 0 C, after which they were stored at -20C. Assays of EGF and PGh The concentration of EGF was measured by RIA (10). The PGI2 concentration was evaluated by measuring its stable hydrolysis product, 6-keto-prostaglandin Fia (6-keto-PGFi«) (11), by RIA (4).
May 3, 1990. Address all correspondence and requests for reprints to: Ari Ristimaki, Kuusitie 5 A 45, 00270 Helsinki, Finland. * This work was supported by the Sigrid Juselius Foundation, the Academy of Finland, the Foundation for Pediatric Research (Helsinki, Finland), the Ida Montini Foundation, the Finnish Foundation for Cancer Research, and the Finnish Culture Foundation.
Endothelial cell cultures With the approval of the local committee of ethics, endothelial cell cultures were prepared from human umbilical cord 623
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
RISTIMAKI ET AL.
624
veins, according to the procedure of Jaffe et al. (12) with slight modifications (13). Briefly, after the cords had been cannulated and washed with 50 mL phosphate-buffered saline, pH 7.4 (PBS; Orion Diagnostica, Espoo, Finland), the cells were digested with 0.1% collagenase (CLS 3, Worthington Biochemical Corp., Freehold, NJ) in PBS for 10 min at room temperature, flushed with 20 mL culture medium, which was medium 199 supplemented with 20% fetal calf serum (Flow Laboratories, Ayshire, United Kingdom), 100 U/mL penicillin, 100 Mg/mL streptomycin, 0.25 jig/mL amphotericin-B, and 2 mmol/L Lglutamine (Gibco, Grand Island, NY), collected by centrifugation (400 X g; 10 min), and resuspended in 5 mL culture medium. Primary cultures were grown in 25-cm2 culture flasks (Nunclon, Roskilde, Denmark) maintained at 37 C in 5% CO2 in air. The medium was changed 12-24 h after seeding and thereafter every second or third day. Only those cultures were used which grew confluent in less than 7 days. The confluent cultures in 25-cm2 flasks were detached with 0.02% EDTA and 0.05% trypsin (Orion Diagnostica) and passed on 96-well (Nunclon) or on 48-well (Costar, Cambridge, MA) plates. All culture flasks and plates were pretreated with 0.2% gelatin (Merck, Darmstadt, Germany) overnight. The cells exhibited von Willebrand factor, as shown by indirect immunofluorescence (Dakopatts, Glostrup, Denmark). Experimental design The effect of milk on PGI2 production was studied in 96-well plates using confluent monolayers of endothelial cells. Immediately before each experiment, the culture medium was aspirated from the confluent monolayers, and these were washed once with 0.25 mL culture medium. Fresh culture medium (0.25 mL) was added without (control) or with 0.1-10% concentrations of human milk or its derivatives, cow's milk, or infant formula and incubated for 1-32 h. Concentration dependence was studied with cow's milk and 3 samples of human milk, each originating from pooled milk from 6 mothers. Pool I milk was excreted 2 days, pool II 2 weeks, and pool III 2 months after delivery. Time dependence was studied with a pool of all 18 human milk samples. Aliquots of this pool of 18 milk samples were extracted with acetone, boiled, pasteurized, or acidified. For acetone extraction, 1 mL of the pool of human milk was added to 40 mL acetone cooled to -40 C (7). The precipitate was washed once with acetone and diethyl ether, dried under an air flow, and suspended in 1 mL PBS with 0.1% BSA (Sigma Chemical Co., St. Louis, MO). The recovery of immunoreactive EGF was 39.1 ± 5.3% (mean ± SE) after acetone extraction (3 separate extractions). For heat treatments, 5 mL of the pool were placed in each test tube with a cap, which was penetrated with a bloodneedle. The test tubes were incubated in a water bath at 60 C for pasteurizing or 100 C for boiling for 30 min. After the heat treatments, the evaporation was replaced by adding PBS. The recovery of immunoreactive EGF was 83.6% after pasteurizing, and 1.5% after boiling. Milk was subjected to acidic conditions by adding hydrochloric acid (1 mmol/L) to 5 mL milk until pH 4.0 and pH 2.0 were obtained. To the control tube a volume of PBS was added equal to the volume of hydrochloric acid added to the pH 2 tube. The acidified milk tubes and the control were
JCE & M • 1991 Vol 72 • No 3
then incubated for 1 h in a 37 C water bath, and after the treatment the original pH 6.5 was obtained by adding sodium hydroxide (5 mmol/L). Cycloheximide (0.2 fig/mL; Sigma), actinomycin-D (0.2 /*g/ mL; Sigma), and indomethacin (0.5 /ug/mL; Dumex, Copenhagen, Denmark) were added simultaneously without (control) or with milk and incubated for 8 h. Human EGF antibodies (Amgen, Thousand Oaks, CA) were added simultaneously with or without 1% or 10% concentrations of human milk and incubated for 24 h (9). At the end of each incubation, which was performed at 37 C in 5% CO2 in air, the medium was placed in a polypropylene tube, frozen immediately, and stored at -20 C until assayed. To study the effect of human milk on the number of cells in confluent cultures, endothelial cells were grown on 48-well plates (Costar) in 0.5 mL culture medium. The cellular DNA content was quantified fluorometrically, using the specific fluorescence of diamidinophenylindole (Sigma) (9, 14) after incubation in 1% and 10% concentrations of human milk for 24 h. Statistical methods Statistical significance was calculated with Student's t test in the case of a single comparison. For multiple comparison the t test was used only if one-way analysis of variance showed a significant difference. The Spearman correlation was used for calculating the significance of the correlation.
Results Milk-induced 6-keto-PGFia release Human breast milk from the various stages of lactation stimulated 6-keto-PGFla release from human umbilical vein endothelial cells in a concentration-dependent manner, whereas cow's milk was ineffective (Fig. 1). Stimulation could be detected at milk concentrations of 0.10.5% and was maximal at concentrations of 5-10%. Stimulation was evident after incubation for 2 h (Fig. 2). The infant formula was ineffective (data not shown). The activity in human milk that stimulated 6-ketoPGFi« release was resistant to pasteurizing, but boiling abolished it and even resulted a slight inhibition (Fig. 3). Acetone extraction enhanced the stimulating activity (Fig. 3). Acidification of human milk to pH 4.0 or pH 2.0 did not change the stimulatory capacity of the milk; a 1% concentration of human milk stimulated the release of 6-keto-PGFla by 2.23 ± 0.03-fold, pH 4.0 milk by 2.51 ± 0.06-fold, and pH 2.0 milk by 2.32 ± 0.12-fold (values are the mean ± SE from four replicate determinations). EGF and 6-keto-PGFia in milk samples The EGF concentration of human milk decreased as the time from parturition increased (Table 1). The infant formula and cow's milk did not contain any EGF (data
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
HUMAN MILK STIMULATES OF PGI2 PRODUCTION 400
5
300
800
Q control Q human / pool 1 PJ human/ pool II | human/ pool I I I | cow
u D
"S 1
c o w
JL
600 -
JL
200
a
0 0
1
0.1
0.5
400-
contrc human milk acetone extr. pasteurized boiled P < O.OS
n m • * *** P < 0.001
•
oDo
6-Ket
100 •
625
200 •
0 1.0
2.5
5.0
10
10 Concentration of human milk (%)
Concentration of milk (%)
FIG. 1. Dose-dependent stimulation of 6-keto-PGF]a release from cultured human umbilical vein endothelial cells induced by three pools of human milk (details in Materials and Methods). Cow's milk was ineffective. Values are the mean ± SE of four replicate determinations, expressed as percentages of basal 6-keto-PGFla release. The statistical significance of increases was calculated against the control value. The lowest concentration of human milk that caused a statistically significant increase in 6-keto-PGFi0 release was 0.1% for pool I (P < 0.005) and pool III (P < 0.01), and 0.5% for pool II (P < 0.05).
FIG. 3. Dose-dependent effect of human milk pooled from 18 samples before (control) and after pasteurizing, acetone extraction, or boiling on 6-keto-PGFia release from confluent human umbilical vein endothelial cell monolayers. Milk and its fractions were stimulatory unless boiled; boiled milk was inhibitory. Values are the mean ± SE of 9-12 replicate determinations from 3 separate experiments, expressed as percentages of basal 6-keto-PGFitt release. The statistical significance of stimulated values was calculated against the control. TABLE 1. EGF and 6-keto-PGFla concentrations of human milk samples 2 days'
2 weeks6
2 months*
6 6.99 ± 1.62 2.74 ± 0.25 1.93 ± 0.20 EGF (nmol/L) 6-Keto-PGFla (nmol/L) 6 0.60 ±0.13 0.56 ±0.14 0.69 ± 0.19 Values are the mean ± SE. EGF and 6-keto-PGFirt were measured by RIA from human milk samples, as explained in Materials and Methods. " Number of samples. 6 Time between delivery and donation of samples. 20 • 1
2
4
8
16
Time (h)
FlG. 2. Time course of human milk-stimulated 6-keto-PGFia release. Cultured human umbilical vein endothelial cells were incubated for periods of time indicated without (control) or with 2.5% human milk pooled from 18 samples. Values are the mean ± SE of four replicate determinations. The first statistically significant difference appeared after incubation for 2 h (P < 0.003).
2 days 2 weeks 2 months
n
32
* 15
ID-
• D
not shown). The initial 6-keto-PGFia concentration was similar in milk samples of various ages (Table 1). The correlation between the stimulation of 6-ketoPGFi« release and EGF concentrations of human milk samples was not significant (Fig. 4). Antibodies to human EGF had no effect on the release of 6-keto-PGFla induced by human milk (data not shown). Effect of inhibitors on 6-keto-PGFla release Cycloheximide and actinomycin-D blocked the stimulatory effect of human milk, indicating that de nouo synthesis of RNA and protein were needed (Table 2).
*
• D
A
o-t150
200
250
6-Keto-PGF , a (% of control)
FlG. 4. The effect of human milk samples (1.0%) on 6-keto-PGFi« release from cultured human umbilical vein endothelial cells in relation to their EGF content. Milk samples were from milk secreted 2 days, 2 weeks, or 2 months after delivery. Values for 6-keto-PGFia release are expressed as percentages of the control value and are the mean of two replicate determinations. There was no correlation (P < 0.05) between EGF concentrations and the amount of stimulation of 6-keto-PGFi«.
They did not affect the basal rate of 6-keto-PGFi« release. Indomethacin inhibited the stimulatory effect of human milk and decreased basal 6-keto-PGFla release
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
RISTIMAKI ET AL.
626
TABLE 2. Effect of inhibitors on human milk-induced 6-keto-PGFlo release 6-Keto-PGFla (nmol/L)
Additions None Milk
16.46 ± 0.43 30.80 ± 1.30°
Cycloheximide (0.2 ^g/mL) Cycloheximide + milk
13.95 ± 1.30 17.76 ± 1.08
Actinomycin-D (0.2 jtg/mL) Actinomycin-D + milk
17.41 ± 0.24 15.74 ± 2.81
Indomethacin (0.5 Mg/mL) Indomethacin + milk
0.78 ± 0.11° 1.08 ± 0.08°
Confluent monolayers of human umbilical vein endothelial cells were incubated for 8 h without (control) or with 2.5% of human milk and/or with the indicated amounts of inhibitors. 6-Keto-PGFia was measured by RIA from conditioned medium. Values are the mean ± SE of three replicate determinations. For details, see Materials and Methods. 0 P < 0.001 vs. the control.
by 95%, confirming the key role of cyclooxygenase (Table 2). Human milk had no effect on the DNA content of the confluent endothelial cell monolayers (data not shown).
Discussion EGF is present in human milk throughout lactation in decreasing concentrations (from 54.4 to 0.83 nmol/L) (15-17). In our milk samples also, the concentration of EGF decreased during lactation. 6-Keto-PGFia levels in human milk have been reported to be 0.05-3.60 nmol/L (18-22). Our present rinding is well within that range. In addition, the concentrations of 6-keto-PGFia in our milk samples were so low that the amounts of it added to the incubation medium with the milk did not contribute to the concentrations of 6-keto-PGFia measured in endothelial cell-conditioned medium. Human milk stimulated the production of PGI2 by human skin fibroblasts (7) and rabbit aorta rings (8). We now show that human milk stimulates the production of PGI2 by human vascular endothelial cells. We also show that the stimulation of PGI2 production is dependent on both RNA and protein synthesis. In this respect, the stimulation of PGI2 synthesis by milk closely resembles its stimulation by peptide growth factors (9,13, 23). That the stimulating activity was not abolished by acetone extraction, pasteurizing, acidification, or dialysis, but could be reduced or totally abolished by boiling or proteolysis, further suggests that the simulator in milk is a relatively heat- and pH-resistant peptide or protein (Refs. 7 and 8 and the present report). All of this would be compatible with an earlier suggestion that the agent stimulating PGI2 synthesis in human
JCE&M«1991 Vol 72 • No 3
milk could be a growth factor, probably EGF (8). However, the present, more detailed study of the phenomenon shows that the activity cannot be attributed to EGF. We found no correlation between the concentrations of EGF and the stimulation of PGI2 production, the addition of antibodies to EGF did not decrease the effect of the milk, and stimulation was much greater (9, 13) than would be expected from the concentration of EGF in the milk. Extraction of human milk with acetone increased its PGI2-stimulating activity, whereas boiling not only abolished the activity, but caused the basal production of PGI2 to decrease. The reason for this finding is not known, but the most probable explanation would appear to be the established inhibition of PGI2 synthesis by lipid peroxides (1), which are removed by acetone extraction. Boiling, in turn, denatures stimulating proteins but
leaves inhibitory peroxides active, the net effect being the inhibition of PGI2 synthesis. Interestingly, acetoneextracted milk caused a greater stimulation than any other substance known to induce PGI2 production by human vascular endothelial cells via protein synthesis. These factors (which have been tested under conditions to similar to those in our present report) include EGF, transforming growth factor-a, transforming growth factor-/?, interleukin-1, and tumor necrosis factor-a (9, 13, 23-25). In animals at birth, PGI2 synthetic pathways are immature (26), and premature weaning leads to a decrease in the activity of the enzymes involved in lipid homeostasis (27) and PGI2 synthesis (28). In addition, some clinical data support the importance of PGI2 in human neonatal vascular adaptation (29, 30). Since human milk itself does not contain physiologically significant levels of prostaglandins (22, 31), the interesting possibility remains that the synthesis of PGI2 in the neonate may be regulated by factors that exist in mother's milk but not in cow's milk or infant formulas based on cow's milk. This hypothesis is supported by our finding, that a pH similar to or even much more acidic than those observed in infant stomach after a milk meal (32-34) did not abolish the activity from the milk. However, no comparative information is available on PGI2 synthesis in infants fed human milk or artificial formulas. In conclusion, human milk is a potent inducer of endothelial cell PGI2 production, and the stimulatory activity of acetone-extracted milk exceeds that of other PGI2-stimulating factors known to act through protein synthesis. The stimulation is not caused by EGF.
Acknowledgments We wish to thank Kristiina Nokelainen and Anu Harju for excellent technical assistance.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
HUMAN MILK STIMULATES OF PGI2 PRODUCTION References 1. Moncada S, Vane J. Pharmacology and endogenous roles of prostaglandin endoperoxides, thromboxane A2, and prostacyclin. Pharmacol Rev. 1979;30:293-331. 2. Lelcuk S, Merhav A, Klausner J, Rozin R. Prostacyclin (PGI2) and thromboxane (Tx) A2; mediators of wound healing. Isr J Med Sci. 1987;23:841-3. 3. Ohtsu A, Fujii K, Kurozumi S. Induction of angiogenic response by chemically stable prostacyclin analogs. Prostaglandin Leukotr Essent Fatty Acids. 1988;33:35-9. 4. Makila U-M, Jouppila P, Kirkinen P, Viinikka L, Ylikorkala 0. Relation between umbilical prostacyclin production and blood-flow in the fetus. Lancet. 1983;l:728-9. 5. Lonnerdal B. Biochemistry and physiological function of human milk proteins. Am J Clin Nutr. 1985;42:1299-317. 6. Shing Y, Davidson S, Klagsbrun M. Purification of polypeptide growth factors from milk. Methods Enzymol. 1987;146:42-8. 7. Subbiah M, Yunker R, Yamamoto M, Kottke B, Bale L. Human breast milk stimulates prostaglandin synthesis in cultured human skin fibroblasts. Biochem Biophys Res Commun. 1985;129:972-6. 8. Bydlowsky S, Yunker R, Campaigne B, Kotogal U, Subbiah M. Human breast milk contains a factor which markedly stimulates aortic prostacyclin synthesis. Biochim Biophys Acta. 1986;878:16. 9. Ristimaki A, Ylikorkala 0, Perheentupa J, Viinikka L. Epidermal growth factor stimulates prostacyclin production by cultured human vascular endothelial cells. Thromb Hemost. 1988;59:248-50. 10. Pesonen K, Viinikka L, Myllyla G, Kiuru J, Perheentupa J. Characterization of material with epidermal growth factor immunoreactivity in human serum and platelets. J Clin Endocrinol Metab. 1989;68:486-91. 11. Johnson R, Morton D, Kinner J, et al. The chemical structure of prostaglandin X (prostacyclin). Prostaglandins. 1976;12:915-28. 12. Jaffe E, Nachman R, Becker C, Minick C. Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J Clin Invest. 1973,52:274556. 13. Ristimaki A. Transforming growth factor a stimulates prostacyclin production by cultured human vascular endothelial cells more potently than epidermal growth factor. Biochem Biophys Res Commun. 1989;160:1100-5. 14. Sorger T, Germinario R. A direct solubilization procedure for the determination of DNA and protein in cultured fibroblasts monolayers. Anal Biochem. 1983;131:254-6. 15. Read L, Upton F, Francis G, Wallace J, Dahlenberg G, Ballard F. Changes in the growth-promoting activity of human milk during lactation. Pediatr Res. 1984;18:133-8. 16. Jansson L, Karlson F, Westermark B. Mitogenic activity and epidermal growth factor content in human milk. Acta Pediatr Scand. 1985;74:250-3. 17. Corps A, Blakeley D, Carr J, Rees L, Brown K. Synergistic stimu-
18. 19. 20. 21. 22. 23. 24. 25.
26. 27. 28. 29.
30. 31. 32. 33. 34.
627
lation of Swiss mouse 3T3 fibroblasts by epidermal growth factor and other factors in human mammary secretions. J Endocrinol. 1987;112:151-9. Lucas A, Mitchell M. Prostaglandins in human milk. Arch Dis Child. 1980;55:950-2. Alzina V, Puig M, de Echaniz L, Villa I, Ferreira R. Prostaglandins in human milk. Biol Neonate. 1986;50:200-4. Chappell J, Clandinin M, Barbe G, Armstrong D. Prostanoid content of human milk: relationship to milk fatty acid content. Endocrinol Exp. 1983;17:351-8. Friedman Z. Prostaglandins in breast milk. Endocrinol Exp. 1986;20:285-91. Wu-Wang C-Y, Neu J. Low levels of prostaglandins in human milk after purification by high performance liquid chromatography. Prostagland Leukotr Med. 1986;24;207-18. Ristimaki A, Ylikorkala 0, Viinikka L. Effect of growth factors on human vascular endothelial cell prostacyclin production. Arteriosclerosis. 1990;10:653-7. Rossi V, Breviario F, Ghezzi P, Dejana E, Mantovani A. Prostacyclin synthesis induced in vascular cells by interleukin-1. Science 1985;229:174-6. Kawakami M, Ishibashi S, Ogawa H, Murase T, Takaku F, Shibata S. Cachectin/TNF as well as interleukin-1 induces prostacyclin synthesis in cultured vascular endothelial cells. Biochem Biophys Res Commun. 1986;141:482-7. Deckmyn H, Font L, Van Hemelen C, Carreras L, Defreyn G, Vermylen J. Low prostacyclin synthase activity of fetal rat aorta. Progressive increase after birth. Life Sci. 1983;33:1491-7. Subbiah M, Yunker R, Menkhaus A, Poe B. Premature weaninginduced changes of cholesterol metabolism in guinea pigs. Endocrinol Metab. 1985;12:E251-6. Bydlowsky S, Yunker R, Subbiah M. Ontogeny of 6-keto-PGFla synthesis in rabbit aorta: effect of premature weaning. Circulation. 1985;72:III-95. Seyberth H, Miiller H, Ulmer H, Wille L. Urinary excretion rates of 6-keto-PGFia in preterm infants recovering from respiratory distress with and without patent ductus arteriosus. Pediatr Res. 1984;18:520-4. Ruth V, Ylikorkala 0, Viinikka L, Raivio K. Urinary excretion of prostacyclin metabolites in infants born after maternal preeclampsia or with birth asphyxia. Biol Neonate. 1989;56:83-9. Neu J, Wu-Wang C-Y, Measel C, Gimotty P. Prostaglandin concentrations in human milk. Am J Clin Nutr. 1988;47:649-52. Mason S. Some aspects of gastric function in the newborn. Arch Dis Child. 1962;37:387-91. Harries J, Fraser A. The acidity of the gastric contents of premature babies during the first fourteen days of life. Biol Neonate. 1968;12;186-93. Fredrikzon B, Hernell 0, Blackberg L, Olivecrona T. Bile saltstimulated lipase in human milk: evidence of activity in vivo and of a role in the digestion of milk retinol esters. Pediatr Res. 1978;12:1048-52.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
Vol. 72, No. 3 Printed in U.S.A.
Human Milk Stimulates Prostacyclin Production by Cultured Human Vascular Endothelial Cells* ARI RISTIMAKI, OLAVI YLIKORKALA, KRISTINA PESONEN, JAAKKO PERHEENTUPA, AND LASSE VIINIKKA Children's Hospital (A.R., K.P., J.P., L. V.) and the Departments of Obstetrics and Gynecology (0. Y.), University of Helsinki, SF-00290 Helsinki, Finland
ABSTRACT. Prostacyclin (PGI2) is an antithrombotic and vasodilatory factor, which is produced mainly by the vascular endothelium. Little is known about how this process is regulated. We investigated the effect of human milk on PGI2 synthesis by human vascular endothelial cells by measuring its stable metabolite, 6-keto-prostaglandin Fia, by RIA. Human milk induced dose- and time-dependent stimulation of PGI2 production, whereas cow's milk was ineffective. The lowest concentration of human milk that stimulated the production of PGI2 was 0.1%, and 10% induced a 2.4- to 3.4-fold increase. The effect of human milk was detectable after 2 h and was blocked by inhibitors of transcription, translation, and cyclooxygenase. Boiling abolished
the activity, but acetone extraction enhanced it. A 10% concentration of acetone-extracted human milk stimulated the release of endothelial cell PGI2 by 6.6-fold. In human milk samples we found no correlation between the amount of immunoreactive epidermal growth factor (EGF) and the activity stimulating PGI2 synthesis. Furthermore, EGF antibodies did not inhibit the activity. This is the first demonstration that human milk stimulates PGI2 production by endothelial cells. We conclude that human milk is a potent inducer of PGI2 production by human vascular endothelial cells and that the stimulatory activity is not due to EGF. (J Clin Endocrinol Metab 72: 623-627, 1991)
P
ROSTACYCLIN (PGI2), a major arachidonic acid metabolite of human vascular endothelial cells, is a powerful vasodilator and a potent inhibitor of platelet aggregation (1). It may also accelerate wound healing (2), induce neovascularization (3), and regulate the fetal and neonatal circulation (4). Despite the importance of PGI2, little is known about the regulation of its production. Human milk is the ideal nutrient for a newborn infant. It contains numerous biologically active compounds that benefit the breast-fed baby. Many of these factors in human milk are polypeptides, which act as host defense agents, enzymes, and hormones (5). In addition, human milk contains growth factors, of which the major one is epidermal growth factor (EGF) (6). Recently, human milk was shown to stimulate PGI2 synthesis by cultured human skin fibroblasts (7) and rabbit aorta (8). The stimulatory activity appeared to be associated with a protein less than 10,000 in mol wt, which was suggested to be a growth factor (8). Interestingly, EGF enhances PGI2 production by human vascular endothelial cells (9). Since nothing is known about the effect of milk on the
production of PGI2 by its biologically most important source, vascular endothelium (1), we investigated the effect of human milk on the synthesis of PGI2 in cultured human vascular endothelial cells.
Materials and Methods Samples Human milk samples were obtained from the Human Milk Bank of the Children's Hospital, University of Helsinki. They were from milk excreted 2 days (n = 6), 2 weeks (n = 6), or 2 months (n = 6) after delivery. Pasteurized cow's milk and cow's milk-derived infant formula (Tutteli, Valio, Turenki-Janakkala, Finland) were commercial products. All milk samples were defatted, and the cells were removed by centrifugation at 15,000 X g for 30 min at 0 C, after which they were stored at -20C. Assays of EGF and PGh The concentration of EGF was measured by RIA (10). The PGI2 concentration was evaluated by measuring its stable hydrolysis product, 6-keto-prostaglandin Fia (6-keto-PGFi«) (11), by RIA (4).
May 3, 1990. Address all correspondence and requests for reprints to: Ari Ristimaki, Kuusitie 5 A 45, 00270 Helsinki, Finland. * This work was supported by the Sigrid Juselius Foundation, the Academy of Finland, the Foundation for Pediatric Research (Helsinki, Finland), the Ida Montini Foundation, the Finnish Foundation for Cancer Research, and the Finnish Culture Foundation.
Endothelial cell cultures With the approval of the local committee of ethics, endothelial cell cultures were prepared from human umbilical cord 623
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
RISTIMAKI ET AL.
624
veins, according to the procedure of Jaffe et al. (12) with slight modifications (13). Briefly, after the cords had been cannulated and washed with 50 mL phosphate-buffered saline, pH 7.4 (PBS; Orion Diagnostica, Espoo, Finland), the cells were digested with 0.1% collagenase (CLS 3, Worthington Biochemical Corp., Freehold, NJ) in PBS for 10 min at room temperature, flushed with 20 mL culture medium, which was medium 199 supplemented with 20% fetal calf serum (Flow Laboratories, Ayshire, United Kingdom), 100 U/mL penicillin, 100 Mg/mL streptomycin, 0.25 jig/mL amphotericin-B, and 2 mmol/L Lglutamine (Gibco, Grand Island, NY), collected by centrifugation (400 X g; 10 min), and resuspended in 5 mL culture medium. Primary cultures were grown in 25-cm2 culture flasks (Nunclon, Roskilde, Denmark) maintained at 37 C in 5% CO2 in air. The medium was changed 12-24 h after seeding and thereafter every second or third day. Only those cultures were used which grew confluent in less than 7 days. The confluent cultures in 25-cm2 flasks were detached with 0.02% EDTA and 0.05% trypsin (Orion Diagnostica) and passed on 96-well (Nunclon) or on 48-well (Costar, Cambridge, MA) plates. All culture flasks and plates were pretreated with 0.2% gelatin (Merck, Darmstadt, Germany) overnight. The cells exhibited von Willebrand factor, as shown by indirect immunofluorescence (Dakopatts, Glostrup, Denmark). Experimental design The effect of milk on PGI2 production was studied in 96-well plates using confluent monolayers of endothelial cells. Immediately before each experiment, the culture medium was aspirated from the confluent monolayers, and these were washed once with 0.25 mL culture medium. Fresh culture medium (0.25 mL) was added without (control) or with 0.1-10% concentrations of human milk or its derivatives, cow's milk, or infant formula and incubated for 1-32 h. Concentration dependence was studied with cow's milk and 3 samples of human milk, each originating from pooled milk from 6 mothers. Pool I milk was excreted 2 days, pool II 2 weeks, and pool III 2 months after delivery. Time dependence was studied with a pool of all 18 human milk samples. Aliquots of this pool of 18 milk samples were extracted with acetone, boiled, pasteurized, or acidified. For acetone extraction, 1 mL of the pool of human milk was added to 40 mL acetone cooled to -40 C (7). The precipitate was washed once with acetone and diethyl ether, dried under an air flow, and suspended in 1 mL PBS with 0.1% BSA (Sigma Chemical Co., St. Louis, MO). The recovery of immunoreactive EGF was 39.1 ± 5.3% (mean ± SE) after acetone extraction (3 separate extractions). For heat treatments, 5 mL of the pool were placed in each test tube with a cap, which was penetrated with a bloodneedle. The test tubes were incubated in a water bath at 60 C for pasteurizing or 100 C for boiling for 30 min. After the heat treatments, the evaporation was replaced by adding PBS. The recovery of immunoreactive EGF was 83.6% after pasteurizing, and 1.5% after boiling. Milk was subjected to acidic conditions by adding hydrochloric acid (1 mmol/L) to 5 mL milk until pH 4.0 and pH 2.0 were obtained. To the control tube a volume of PBS was added equal to the volume of hydrochloric acid added to the pH 2 tube. The acidified milk tubes and the control were
JCE & M • 1991 Vol 72 • No 3
then incubated for 1 h in a 37 C water bath, and after the treatment the original pH 6.5 was obtained by adding sodium hydroxide (5 mmol/L). Cycloheximide (0.2 fig/mL; Sigma), actinomycin-D (0.2 /*g/ mL; Sigma), and indomethacin (0.5 /ug/mL; Dumex, Copenhagen, Denmark) were added simultaneously without (control) or with milk and incubated for 8 h. Human EGF antibodies (Amgen, Thousand Oaks, CA) were added simultaneously with or without 1% or 10% concentrations of human milk and incubated for 24 h (9). At the end of each incubation, which was performed at 37 C in 5% CO2 in air, the medium was placed in a polypropylene tube, frozen immediately, and stored at -20 C until assayed. To study the effect of human milk on the number of cells in confluent cultures, endothelial cells were grown on 48-well plates (Costar) in 0.5 mL culture medium. The cellular DNA content was quantified fluorometrically, using the specific fluorescence of diamidinophenylindole (Sigma) (9, 14) after incubation in 1% and 10% concentrations of human milk for 24 h. Statistical methods Statistical significance was calculated with Student's t test in the case of a single comparison. For multiple comparison the t test was used only if one-way analysis of variance showed a significant difference. The Spearman correlation was used for calculating the significance of the correlation.
Results Milk-induced 6-keto-PGFia release Human breast milk from the various stages of lactation stimulated 6-keto-PGFla release from human umbilical vein endothelial cells in a concentration-dependent manner, whereas cow's milk was ineffective (Fig. 1). Stimulation could be detected at milk concentrations of 0.10.5% and was maximal at concentrations of 5-10%. Stimulation was evident after incubation for 2 h (Fig. 2). The infant formula was ineffective (data not shown). The activity in human milk that stimulated 6-ketoPGFi« release was resistant to pasteurizing, but boiling abolished it and even resulted a slight inhibition (Fig. 3). Acetone extraction enhanced the stimulating activity (Fig. 3). Acidification of human milk to pH 4.0 or pH 2.0 did not change the stimulatory capacity of the milk; a 1% concentration of human milk stimulated the release of 6-keto-PGFla by 2.23 ± 0.03-fold, pH 4.0 milk by 2.51 ± 0.06-fold, and pH 2.0 milk by 2.32 ± 0.12-fold (values are the mean ± SE from four replicate determinations). EGF and 6-keto-PGFia in milk samples The EGF concentration of human milk decreased as the time from parturition increased (Table 1). The infant formula and cow's milk did not contain any EGF (data
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
HUMAN MILK STIMULATES OF PGI2 PRODUCTION 400
5
300
800
Q control Q human / pool 1 PJ human/ pool II | human/ pool I I I | cow
u D
"S 1
c o w
JL
600 -
JL
200
a
0 0
1
0.1
0.5
400-
contrc human milk acetone extr. pasteurized boiled P < O.OS
n m • * *** P < 0.001
•
oDo
6-Ket
100 •
625
200 •
0 1.0
2.5
5.0
10
10 Concentration of human milk (%)
Concentration of milk (%)
FIG. 1. Dose-dependent stimulation of 6-keto-PGF]a release from cultured human umbilical vein endothelial cells induced by three pools of human milk (details in Materials and Methods). Cow's milk was ineffective. Values are the mean ± SE of four replicate determinations, expressed as percentages of basal 6-keto-PGFla release. The statistical significance of increases was calculated against the control value. The lowest concentration of human milk that caused a statistically significant increase in 6-keto-PGFi0 release was 0.1% for pool I (P < 0.005) and pool III (P < 0.01), and 0.5% for pool II (P < 0.05).
FIG. 3. Dose-dependent effect of human milk pooled from 18 samples before (control) and after pasteurizing, acetone extraction, or boiling on 6-keto-PGFia release from confluent human umbilical vein endothelial cell monolayers. Milk and its fractions were stimulatory unless boiled; boiled milk was inhibitory. Values are the mean ± SE of 9-12 replicate determinations from 3 separate experiments, expressed as percentages of basal 6-keto-PGFitt release. The statistical significance of stimulated values was calculated against the control. TABLE 1. EGF and 6-keto-PGFla concentrations of human milk samples 2 days'
2 weeks6
2 months*
6 6.99 ± 1.62 2.74 ± 0.25 1.93 ± 0.20 EGF (nmol/L) 6-Keto-PGFla (nmol/L) 6 0.60 ±0.13 0.56 ±0.14 0.69 ± 0.19 Values are the mean ± SE. EGF and 6-keto-PGFirt were measured by RIA from human milk samples, as explained in Materials and Methods. " Number of samples. 6 Time between delivery and donation of samples. 20 • 1
2
4
8
16
Time (h)
FlG. 2. Time course of human milk-stimulated 6-keto-PGFia release. Cultured human umbilical vein endothelial cells were incubated for periods of time indicated without (control) or with 2.5% human milk pooled from 18 samples. Values are the mean ± SE of four replicate determinations. The first statistically significant difference appeared after incubation for 2 h (P < 0.003).
2 days 2 weeks 2 months
n
32
* 15
ID-
• D
not shown). The initial 6-keto-PGFia concentration was similar in milk samples of various ages (Table 1). The correlation between the stimulation of 6-ketoPGFi« release and EGF concentrations of human milk samples was not significant (Fig. 4). Antibodies to human EGF had no effect on the release of 6-keto-PGFla induced by human milk (data not shown). Effect of inhibitors on 6-keto-PGFla release Cycloheximide and actinomycin-D blocked the stimulatory effect of human milk, indicating that de nouo synthesis of RNA and protein were needed (Table 2).
*
• D
A
o-t150
200
250
6-Keto-PGF , a (% of control)
FlG. 4. The effect of human milk samples (1.0%) on 6-keto-PGFi« release from cultured human umbilical vein endothelial cells in relation to their EGF content. Milk samples were from milk secreted 2 days, 2 weeks, or 2 months after delivery. Values for 6-keto-PGFia release are expressed as percentages of the control value and are the mean of two replicate determinations. There was no correlation (P < 0.05) between EGF concentrations and the amount of stimulation of 6-keto-PGFi«.
They did not affect the basal rate of 6-keto-PGFi« release. Indomethacin inhibited the stimulatory effect of human milk and decreased basal 6-keto-PGFla release
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
RISTIMAKI ET AL.
626
TABLE 2. Effect of inhibitors on human milk-induced 6-keto-PGFlo release 6-Keto-PGFla (nmol/L)
Additions None Milk
16.46 ± 0.43 30.80 ± 1.30°
Cycloheximide (0.2 ^g/mL) Cycloheximide + milk
13.95 ± 1.30 17.76 ± 1.08
Actinomycin-D (0.2 jtg/mL) Actinomycin-D + milk
17.41 ± 0.24 15.74 ± 2.81
Indomethacin (0.5 Mg/mL) Indomethacin + milk
0.78 ± 0.11° 1.08 ± 0.08°
Confluent monolayers of human umbilical vein endothelial cells were incubated for 8 h without (control) or with 2.5% of human milk and/or with the indicated amounts of inhibitors. 6-Keto-PGFia was measured by RIA from conditioned medium. Values are the mean ± SE of three replicate determinations. For details, see Materials and Methods. 0 P < 0.001 vs. the control.
by 95%, confirming the key role of cyclooxygenase (Table 2). Human milk had no effect on the DNA content of the confluent endothelial cell monolayers (data not shown).
Discussion EGF is present in human milk throughout lactation in decreasing concentrations (from 54.4 to 0.83 nmol/L) (15-17). In our milk samples also, the concentration of EGF decreased during lactation. 6-Keto-PGFia levels in human milk have been reported to be 0.05-3.60 nmol/L (18-22). Our present rinding is well within that range. In addition, the concentrations of 6-keto-PGFia in our milk samples were so low that the amounts of it added to the incubation medium with the milk did not contribute to the concentrations of 6-keto-PGFia measured in endothelial cell-conditioned medium. Human milk stimulated the production of PGI2 by human skin fibroblasts (7) and rabbit aorta rings (8). We now show that human milk stimulates the production of PGI2 by human vascular endothelial cells. We also show that the stimulation of PGI2 production is dependent on both RNA and protein synthesis. In this respect, the stimulation of PGI2 synthesis by milk closely resembles its stimulation by peptide growth factors (9,13, 23). That the stimulating activity was not abolished by acetone extraction, pasteurizing, acidification, or dialysis, but could be reduced or totally abolished by boiling or proteolysis, further suggests that the simulator in milk is a relatively heat- and pH-resistant peptide or protein (Refs. 7 and 8 and the present report). All of this would be compatible with an earlier suggestion that the agent stimulating PGI2 synthesis in human
JCE&M«1991 Vol 72 • No 3
milk could be a growth factor, probably EGF (8). However, the present, more detailed study of the phenomenon shows that the activity cannot be attributed to EGF. We found no correlation between the concentrations of EGF and the stimulation of PGI2 production, the addition of antibodies to EGF did not decrease the effect of the milk, and stimulation was much greater (9, 13) than would be expected from the concentration of EGF in the milk. Extraction of human milk with acetone increased its PGI2-stimulating activity, whereas boiling not only abolished the activity, but caused the basal production of PGI2 to decrease. The reason for this finding is not known, but the most probable explanation would appear to be the established inhibition of PGI2 synthesis by lipid peroxides (1), which are removed by acetone extraction. Boiling, in turn, denatures stimulating proteins but
leaves inhibitory peroxides active, the net effect being the inhibition of PGI2 synthesis. Interestingly, acetoneextracted milk caused a greater stimulation than any other substance known to induce PGI2 production by human vascular endothelial cells via protein synthesis. These factors (which have been tested under conditions to similar to those in our present report) include EGF, transforming growth factor-a, transforming growth factor-/?, interleukin-1, and tumor necrosis factor-a (9, 13, 23-25). In animals at birth, PGI2 synthetic pathways are immature (26), and premature weaning leads to a decrease in the activity of the enzymes involved in lipid homeostasis (27) and PGI2 synthesis (28). In addition, some clinical data support the importance of PGI2 in human neonatal vascular adaptation (29, 30). Since human milk itself does not contain physiologically significant levels of prostaglandins (22, 31), the interesting possibility remains that the synthesis of PGI2 in the neonate may be regulated by factors that exist in mother's milk but not in cow's milk or infant formulas based on cow's milk. This hypothesis is supported by our finding, that a pH similar to or even much more acidic than those observed in infant stomach after a milk meal (32-34) did not abolish the activity from the milk. However, no comparative information is available on PGI2 synthesis in infants fed human milk or artificial formulas. In conclusion, human milk is a potent inducer of endothelial cell PGI2 production, and the stimulatory activity of acetone-extracted milk exceeds that of other PGI2-stimulating factors known to act through protein synthesis. The stimulation is not caused by EGF.
Acknowledgments We wish to thank Kristiina Nokelainen and Anu Harju for excellent technical assistance.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
HUMAN MILK STIMULATES OF PGI2 PRODUCTION References 1. Moncada S, Vane J. Pharmacology and endogenous roles of prostaglandin endoperoxides, thromboxane A2, and prostacyclin. Pharmacol Rev. 1979;30:293-331. 2. Lelcuk S, Merhav A, Klausner J, Rozin R. Prostacyclin (PGI2) and thromboxane (Tx) A2; mediators of wound healing. Isr J Med Sci. 1987;23:841-3. 3. Ohtsu A, Fujii K, Kurozumi S. Induction of angiogenic response by chemically stable prostacyclin analogs. Prostaglandin Leukotr Essent Fatty Acids. 1988;33:35-9. 4. Makila U-M, Jouppila P, Kirkinen P, Viinikka L, Ylikorkala 0. Relation between umbilical prostacyclin production and blood-flow in the fetus. Lancet. 1983;l:728-9. 5. Lonnerdal B. Biochemistry and physiological function of human milk proteins. Am J Clin Nutr. 1985;42:1299-317. 6. Shing Y, Davidson S, Klagsbrun M. Purification of polypeptide growth factors from milk. Methods Enzymol. 1987;146:42-8. 7. Subbiah M, Yunker R, Yamamoto M, Kottke B, Bale L. Human breast milk stimulates prostaglandin synthesis in cultured human skin fibroblasts. Biochem Biophys Res Commun. 1985;129:972-6. 8. Bydlowsky S, Yunker R, Campaigne B, Kotogal U, Subbiah M. Human breast milk contains a factor which markedly stimulates aortic prostacyclin synthesis. Biochim Biophys Acta. 1986;878:16. 9. Ristimaki A, Ylikorkala 0, Perheentupa J, Viinikka L. Epidermal growth factor stimulates prostacyclin production by cultured human vascular endothelial cells. Thromb Hemost. 1988;59:248-50. 10. Pesonen K, Viinikka L, Myllyla G, Kiuru J, Perheentupa J. Characterization of material with epidermal growth factor immunoreactivity in human serum and platelets. J Clin Endocrinol Metab. 1989;68:486-91. 11. Johnson R, Morton D, Kinner J, et al. The chemical structure of prostaglandin X (prostacyclin). Prostaglandins. 1976;12:915-28. 12. Jaffe E, Nachman R, Becker C, Minick C. Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J Clin Invest. 1973,52:274556. 13. Ristimaki A. Transforming growth factor a stimulates prostacyclin production by cultured human vascular endothelial cells more potently than epidermal growth factor. Biochem Biophys Res Commun. 1989;160:1100-5. 14. Sorger T, Germinario R. A direct solubilization procedure for the determination of DNA and protein in cultured fibroblasts monolayers. Anal Biochem. 1983;131:254-6. 15. Read L, Upton F, Francis G, Wallace J, Dahlenberg G, Ballard F. Changes in the growth-promoting activity of human milk during lactation. Pediatr Res. 1984;18:133-8. 16. Jansson L, Karlson F, Westermark B. Mitogenic activity and epidermal growth factor content in human milk. Acta Pediatr Scand. 1985;74:250-3. 17. Corps A, Blakeley D, Carr J, Rees L, Brown K. Synergistic stimu-
18. 19. 20. 21. 22. 23. 24. 25.
26. 27. 28. 29.
30. 31. 32. 33. 34.
627
lation of Swiss mouse 3T3 fibroblasts by epidermal growth factor and other factors in human mammary secretions. J Endocrinol. 1987;112:151-9. Lucas A, Mitchell M. Prostaglandins in human milk. Arch Dis Child. 1980;55:950-2. Alzina V, Puig M, de Echaniz L, Villa I, Ferreira R. Prostaglandins in human milk. Biol Neonate. 1986;50:200-4. Chappell J, Clandinin M, Barbe G, Armstrong D. Prostanoid content of human milk: relationship to milk fatty acid content. Endocrinol Exp. 1983;17:351-8. Friedman Z. Prostaglandins in breast milk. Endocrinol Exp. 1986;20:285-91. Wu-Wang C-Y, Neu J. Low levels of prostaglandins in human milk after purification by high performance liquid chromatography. Prostagland Leukotr Med. 1986;24;207-18. Ristimaki A, Ylikorkala 0, Viinikka L. Effect of growth factors on human vascular endothelial cell prostacyclin production. Arteriosclerosis. 1990;10:653-7. Rossi V, Breviario F, Ghezzi P, Dejana E, Mantovani A. Prostacyclin synthesis induced in vascular cells by interleukin-1. Science 1985;229:174-6. Kawakami M, Ishibashi S, Ogawa H, Murase T, Takaku F, Shibata S. Cachectin/TNF as well as interleukin-1 induces prostacyclin synthesis in cultured vascular endothelial cells. Biochem Biophys Res Commun. 1986;141:482-7. Deckmyn H, Font L, Van Hemelen C, Carreras L, Defreyn G, Vermylen J. Low prostacyclin synthase activity of fetal rat aorta. Progressive increase after birth. Life Sci. 1983;33:1491-7. Subbiah M, Yunker R, Menkhaus A, Poe B. Premature weaninginduced changes of cholesterol metabolism in guinea pigs. Endocrinol Metab. 1985;12:E251-6. Bydlowsky S, Yunker R, Subbiah M. Ontogeny of 6-keto-PGFla synthesis in rabbit aorta: effect of premature weaning. Circulation. 1985;72:III-95. Seyberth H, Miiller H, Ulmer H, Wille L. Urinary excretion rates of 6-keto-PGFia in preterm infants recovering from respiratory distress with and without patent ductus arteriosus. Pediatr Res. 1984;18:520-4. Ruth V, Ylikorkala 0, Viinikka L, Raivio K. Urinary excretion of prostacyclin metabolites in infants born after maternal preeclampsia or with birth asphyxia. Biol Neonate. 1989;56:83-9. Neu J, Wu-Wang C-Y, Measel C, Gimotty P. Prostaglandin concentrations in human milk. Am J Clin Nutr. 1988;47:649-52. Mason S. Some aspects of gastric function in the newborn. Arch Dis Child. 1962;37:387-91. Harries J, Fraser A. The acidity of the gastric contents of premature babies during the first fourteen days of life. Biol Neonate. 1968;12;186-93. Fredrikzon B, Hernell 0, Blackberg L, Olivecrona T. Bile saltstimulated lipase in human milk: evidence of activity in vivo and of a role in the digestion of milk retinol esters. Pediatr Res. 1978;12:1048-52.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 24 November 2015. at 16:57 For personal use only. No other uses without permission. . All rights reserved.
