- Email: [email protected]
Dietary polyunsaturated fatty acid prevents malformations in offspring of diabetic rats
Dietary polyunsaturated fatty acid prevents malformations in offspring of diabetic rats E. Albert Reece, MD," Ying-King Wu, MD, PhD, a Arnon Wiznitzer, MD, a Carol Homko, RN, MS," Jing Yao, MD," Michael Borenstein, PhD, b and Gary Sloskey, PharmD b Philadelphia, Pennsylvania OBJECTIVE: The purpose of the current study was to determine whether a dietary source of arachidonic acid could serve as a pharmacologic prophylaxis to obviate the teratogenic effects of hyperglycemia. STUDY DESIGN: Eighty-day-old Sprague-Dawley rats were mated, and after conception were randomly allocated to five groups: two groups were nondiabetic normal controls and three groups had diabetes experimentally induced with streptozocin. Of the two control groups, one was fed a normal diet (group 1) and the other group (group 2) received a normal diet and 1.0 ml of safflower oil, a polyunsaturated fatty acid known to increase serum arachidonic acid levels. In the three diabetic groups (groups 3, 4, and 5) glucose levels were allowed to remain >350 mg/dl by withholding daily insulin therapy. Group 3 received a normal diet without supplementation; group 4 received a normal diet plus normal saline solution sham feedings, whereas group 5 received a normal diet supplemented with 1.0 ml of safflower oil. The oral agents (normal saline solution and polyunsaturated fatty acid) were administered with a tuberculin syringe. RESULTS: Diabetic rats not receiving insulin therapy and receiving normal diets produced offspring with malformation rates of 20% compared with control rates of 4.8%. Supplemental normal saline solution or safflower oil given orally to controls did not alter the growth or malformation rates. These rates were similarly unaffected in the diabetic rats receiving oral supplementation of normal saline solution. However, with safflower oil supplementation to diabetic rats the incidence of neural tube defects was decreased from 20.0% to 7.6% (p < 0.0001). An inverse relationship was observed between the malformation rate and the serum arachidonic acid level: 17.83 (SD 5.84 ~g/ml) in the nondiabetic controls, with a malformation rate of 4.8%, versus 14.18 (SD 2.58 ~g/ml) in the diabetic rats, with a malformation rate of 20.0% (p < 0.05). With safflower oil supplementation serum levels of arachidonic increased from 14.18 + 2.58 ~g/ml to 19,99 + 7.99 ~g/ml (p < 0.05); this was associated with a concomitant decline in the malformation rate. CONCLUSION: These data demonstrate that diabetic embryopathy is associated with a deficiency state in essential fatty acid, corroborating our previous in vitro findings. Furthermore, the use of a dietary polyunsaturated fatty acid that specifically increases arachidonic levels significantly reduced the incidence of diabetic embryopathy. These findings may serve as a basis for developing strategies of pharmacologic prophylaxis against diabetes-induced congenital malformations. (Am J Obstet Gynecol 1996;175:818-23.)
Key words: Diabetes, pregnancy, embryopathy, polyunsaturated fatty acids, arachidonic acid
Congenital malformations a m o n g infants of diabetic mothers occur at a rate about four to five times h i g h e r than that r e p o r t e d in the general population. Clinically there is a great diversity seen in the types of malformations associated with i n s u l i n - d e p e n d e n t diabetes mellitus. However, the organ systems most c o m m o n l y affected include the central nervous and cardiovascular systems. A threefold to 20-fold increase in neural tube defects has From the Department of Obstetrics, Gynecology, and Reproductive Sciences, Temple University School of Medicine, a and the Temple University School of Pharmacy.h Presented at the Sixteenth Annual Meeting of the Society of Perinatal Obstetricians, Kamuela, Hawaii, February 4-10, 1996. Reprint requests: E. Albert Reece, MD, Department of Obstetrics, Gynecolog3, and Reproductive Sciences, Temple University School of Medicine, 3401 N. Broad St., 7-OPB, Philadelphia, PA 19140. Copyright 9 1996 by Mosby-Year Book, Inc. 0002-9378/96 $5.00 + 0 6/6/75230 818
b e e n r e p o r t e d a m o n g the offspring of w o m e n with diabetes. This increased m a l f o r m a t i o n rate accounts for about 40% of the deaths a m o n g infants of diabetic mothers. 1 It has b e e n shown that altered metabolic fuels, including hyperglycemia, a m o n g p r e g n a n t diabetic w o m e n is associated with an increased m a l f o r m a t i o n rate. Conversely, recent studies have d e m o n s t r a t e d that preconception glycemic control results in a decreased rate of malformations, z-~ Unfortunately, however, meticulous family p l a n n i n g is necessary to effect this benefit. We previously r e p o r t e d our results with the use of increasing concentrations of v-glucose during o r g a n o g e n esis to induce malformations in a dose-related fashion in the postimplantation rat e m b r y o culture model. 6' 7 In o u r rat m o d e l we specifically studied the effects of hyperglycemia on neural tube development. F u r t h e r m o r e , when arachidonic acid was added in increasing concentrations,
R e e c e et al.
Volume 175, Number 4, Part l AmJ Obstet Gynecol
819
Table I. Relationship between safflower oil dosage and pregnancy outcomes
Dose
No. of embryos
0.5 ml/day SO 1.0 ml/day SO 2.0 ml/day SO Diabetic controls fed normal diet
187 162 67 280
AAlevel(vg/ml), day 12- day 0 2.7• 3.0• 86.96• -3.65•
5.4* 9.3 27.4* 8.2
AA, Arachidonic acid; NTD,neural tube defects; CT~L,crown-rump length; *p < 0.05.
malformation rates declined, even in the presence of the 100% teratogenic dose of excess D-glucose. s The purpose of the current in vivo study was to test the hypothesis that diabetic embryopathy is associated with an arachidonic acid deficiency state and that daily dietary supplementation with a polyunsaturated fatty acid high in arachidonic acid will obviate the teratogenic effects of hyperglycemia. Methods Test animals. Seventy- to 90-day-old Sprague-Dawley rats with weights between 250 and 300 gm were used for the current study. Rats were mated overnight and vaginal smears were examined the next morning. If spermatozoa were present, pregnancy day 0 was determined. Diabetes was induced in some (experimental group), whereas others remained nondiabetic (controls). Diabetes was induced by an intravenous injection of streptozocin (65 mg/kg) on pregnancy day 6 because preconceptionally induced diabetes may be associated with infertility in rats. Furthermore, streptozocin used during pregnancy is readily eliminated within 24 hours and thus will be removed from the circulation before the critical period of organogenesis. 9 Glucose levels between 300 and 380 m g / d l were achieved by simply withholding daily insulin therapy. (Normal glucose level in the rat is approximately 150 mg/dl.) Rats were randomly assigned to one of five groups. Group 1 animals were nondiabetic controls receiving a normal diet, and group 2 animals were fed a normal diet supplemented with 1 ml of safflower oil daily starting on pregnancy day 6 until the day embryos were explained. The remaining three groups had diabetes experimentally induced with strePtozocin. Group 3 animals were diabetics receiving a normal diet, group 4 were diabetics receiving a normal diet plus an oral sham feeding of normal saline solution, and group 5 animals were diabetics fed a normal diet and supplemented with safflower oil as in control group 2. The oral agents (normal saline solution and safflower oil) were administered with a tuberculin syringe daily starting on day 6 of gestation. Determination of arachidonic acid dosage. A pilot study was undertaken to determine the most effective dose for supplementation. The pilot study was conducted in a
Absorptions] (%)
NTD] (%) 16.2" 7.6 5.5* 20.0
CRL (mm) 0.54• 0.52• 0.35• 0.63•
SO, safflower oil.
m a n n e r similar to the larger study described in detail in this section. Four diabetic groups were chosen and these animals received varying doses of safflower oil (2 ml/day, 1 ml/day, and 0.5 ml/day) or normal diet alone. A dose of 1 ml per daywas found to be most effective and was the dose used in groups 2 and 5 (Table I). Assessment parameters. Blood samples were collected by cardiac puncture on day 0 and day 12 of pregnancy and analyzed for arachidonic acid. Blood was obtained daily from the tail vein and analyzed for glucose levels. Arachidonic acid assays were performed by capillary gas chromatography with hexadecanoic acid as the internal standard by a technique described elsewhere) ~ O n pregnancy day 12 the rats were killed; the embryos were examined and measurements were taken u n d e r light microscopy with x l 0 magnification. The crownrump length of each embryo was measured. The embryos were categorized as morphologically normal or showing neural tube defects or other malformations. These included microcephaly, abnormal cervical spine, reverse tail flexion, tubular heart formation, and pericardial effusion. The normal embryos demonstrated correct body flexure and both the anterior and posterior neural pores were closed. Embryos with malrotation, delayed closure of a single neural pore, or an abnormally open neural tube were defined as having a neural tube defect. Embryos in various stages Of resorption were denoted as resorbed and no further analysis was performed on the latter group. Statistical analyses was performed with Student t tests, )C, and analysis of variance where appropriate. Data are expressed as mean + SD. Results In the pilot study animals receiving either 1 or 2 ml/day showed a significant decrease in neural tube defects compared with the diabetic control group (7.6% and 5.5%, respectively). Supplementation with 2 ml/day dramatically increased serum arachidonic acid levels and reduced the neural tube defect rate. However, the incidence of embryonic resorption was significantly higher (27%) than in either of the other two groups (5.4% with 0.5 ml/day and 9.3% with 1 ml/day, p < 0.05) (Table I). In addition, 2 ml/day safflower oil had an adverse effect of
820
Reece et al.
October 1996 AmJ Obstet Gynecol
Table II. Relationship between safflower oil dosage and pregnancy o u t c o m e s
Controls Group 1, normal Group 2, normal Diabetics Group 3, normal Group 4, normal saline solution Group 5, normal
Resorptions (%)
NTD (%)
0.62• 0.63_+0.1
10/163 (6.1) 8/97 (9.3)
4.8 4.1
360.94 _+48 350.95 _+48
0.63_+0.2 0.68_+0.4
23/280 (8.2) 4/78 (5.1)
20.0 28.2
353 _+36
0.52_+0.2
15/162 (9.3)
7.6
No. of embryos
Glucose level (mg/dl)
diet diet and PUFA
163 97
133.25 _+22 143.5 -+ 24
diet diet and normal sham feeding diet and PUFA
280 78 162
CRL (mm)
CRL, Crown-rump length; NTD, neural tube defects; PUFA, polyunsaturated fatty acid. *Comparison of arachidonic acid levels between day 0 and day 12 (paired t tests).
Table IlL S e r u m arachidonic acid levels a m o n g offspring of controls and diabetic rats AA levels (izg/ml) Day 0
I
Day 12
Significance*
1
Controls Group 1, normal Group 2, normal Diabetics Group 3, normal Group 4, normal Group 5, normal
diet diet and PUFA
17.58_+4.86 14.15_+3.62
17.83_+5.84 15.22•
p>0.1 p>0.1
diet diet and normal saline solution sham feeding diet and PUFA
17.83 + 5.84 17.15_+4.15 16.99 _+8.59
14.18 + 2.58 13.43_+3.52 19.99 -+ 7.99
p< 0.05 p< 0.05 p< 0.05
AA, Arachidonic acid; PUFA, polyunsaturated fatty acid. *Comparison of arachidonic acid levels between day 0 and day 12 (paired t tests). fetal growth as shown by a statistically significant reduction in c r o w n - r u m p length in this g r o u p c o m p a r e d with controls and animals receiving lower doses of supplementation (Table II). T h e r e f o r e 1 m l / d a y safflower oil supplem e n t a t i o n was used for further studies, and the remainder of the data will be based on this dosage. Arachidonic acid levels after supplementation with polyunsaturated fatty acids. Arachidonic acid in the maternal s e r u m was analyzed on days 0 and 12 of pregnancy. T h e arachidonic acid levels on day 0 and day 12 in maternal s e r u m of the nondiabetic controls are shown in Table III. T h e diabetic rats fed a n o r m a l diet only d e m o n s t r a t e d a m a r k e d r e d u c t i o n in arachidonic acid levels by day 12 to a m e a n level that was statistically significant (14.18 _+ 2.58 D g / m l vs 17.83 + 5.84 p g / m l , p < 0.0001) f r o m that observed in controls. Arachidonic acid b l o o d levels between days 0 and 12 after 1 m l / d a y of safflower oil supplementation were significantly h i g h e r than in both the diabetic animals fed a n o r m a l diet and n o r m a l controls (Table III). However, there was no significant difference in ser u m arachidonic acid levels on day 12 between supplem e n t e d controls (group 2) and s u p p l e m e n t e d diabetic animals (group 5). Pregnancy outcomes. G r o u p 1 ( n o n d i a b e t i c / n o r m a l diet) c o n t a i n e d 163 embryos. A l t h o u g h each rat pregnancy had multiple embryos, embryos ranged f r o m 8 to 20 per pregnancy with an average of 15 per rat. T h e m e a n m a t e r n a l glucose level was 133 + 22 m g / d l . T h e resorp-
tion rate in this group was 6.13% (10/163); the crownr u m p length of the embryos was 0.62 _+0.06 cm, and the incidence of neural tube defects was 4.8%. G r o u p 2 ( n o n d i a b e t i c / n o r m a l diet plus safflower oil) c o n t a i n e d 97 embryos. T h e m e a n glucose level was 143.5 _+ 24 m g / d l . T h e resorption rate was 9.3% and the crown-rump length was 0.63 + 0.1 cm for this group. T h e incidence of neural tube defects was 4.1%. G r o u p 3 ( d i a b e t i c / n o r m a l diet) c o n t a i n e d 280 embryos. T h e m e a n m a t e r n a l glucose level was 360 _+48 m g / d l , which was significantly h i g h e r than in the control group (p < 0.05). In this g r o u p the resorption rate was n o t significantly different from that of the control g r o u p of 8.2% (23/280), n o r was the m e a n c r o w n - r u m p length of the embryos different (0.63 _+0.18 cm). However, neural tube defects were increased by sevenfold c o m p a r e d with controls (20.0% vs 4.8%, p < 0.05). This resulted in nearly o n e fourth of the embryos p e r rat pregnancy being malf o r m e d (Fig. 1). G r o u p 4 ( d i a b e t i c / n o r m a l diet plus sham feeding) contained 78 embryos. Blood glucose levels were 350 _+48 m g / d l in this group, which had a resorption rate of 5.1% and a m e a n crown-rump length of 0.68 _+0.4 cm. Arachidonic acid levels fell f r o m 17.15 _+4.15 p g / m l to 13.43 + 3.52 p g / m l (p < 0.05), whereas the incidence of neural tube defects increased significantly to 28.2% in this group. G r o u p 5 ( d i a b e t i c / n o r m a l diet plus safflower oil) con-
Volume 175, Number 4, Part 1 AmJ Obstet Gynecol
Reeoe et al.
821
Fig. 1. Twelve-day-old normal embryo is seen on right next to embryo with neural robe defect (left). Arrows outline defect. tained 162 embryos with m e a n maternal glucose levels of 353 _4-36 m g / d l . This g r o u p had a resorption rate of 9.3% (15/162) and a m e a n e m b r y o n i c c r o w n - r u m p length of 0.52 + 0.2 cm, n o t significantly different f r o m the control groups. However, neural tube defects in this g r o u p were significantly decreased c o m p a r e d with the diabetic g r o u p fed n o r m a l diet without arachidonic s u p p l e m e n t a t i o n (20.0% vs 7.6%, p < 0.005) (Fig. 2). This r e d u c e d incid e n c e of neural tube defect was n o t statistically different f r o m that of the nondiabetic controls.
Comment In our previous studies with the postimplantation rat e m b r y o culture we d e m o n s t r a t e d that arachidonic acid s u p p l e m e n t a t i o n r e d u c e d the rates of neural tube defects even in the presence of a 100 % teratogenic dose of excess D-glucose. 8 In the c u r r e n t study we replicated in vivo the results of our findings obtained in vitro. We d e m o n strated that in diabetic rats fed a n o r m a l diet the arachidonic acid s e r u m level is significantly decreased, whereas the m a l f o r m a t i o n rate is significantly increased above control rates. Conversely, the malformation rate was significantly r e d u c e d in diabetic rats whose diets were s u p p l e m e n t e d with a dietary source of arachidonic acid. These data support our hypothesis that diabetic embryopathy is associated with an arachidonic acid deficiency state and that daily s u p p l e m e n t a t i o n with a polyunsaturated fatty acid high in arachidonic acid will obviate the teratogenic effects of hyperglycemia and o t h e r metabolic fuels. In the c u r r e n t experiments we used safflower oil as a dietary source of arachidonic acid, which served as a p h a r m a c o l o g i c prophylaxis against the teratogenic effects of hyperglycemia. Safflower oil is a polyunsaturated fatty acid that is converted to linoleic acid and eventually
to arachidonic acid by a tightly linked enzyme s e q u e n c e ? 1 We d e m o n s t r a t e d a direct relationship between safflower oil s u p p l e m e n t a t i o n and arachidonic acid levels in a dose-related fashion. As can be seen in Table I, as arachidonic acid levels increased, the incidence of neural tube defects decreased in the offspring of the diabetic rats. However, w h e n arachidonic acid levels in the diabetic rats a p p r o x i m a t e d levels seen in nondiabetic controls, the incidence of e m b r y o n i c neural tube defects still r e m a i n e d significantly elevated above the background. It appears that diabetic animals n e e d e d h i g h e r levels of arachidonic acid to c o m p e n s a t e for or override the metabolic effects of hyperglycemia to result in a r e d u c t i o n in the incidence of neural tube defects toward b a c k g r o u n d rates. However, d o u b l i n g the dose of safflower oil to 2 m l / d a y resulted in a significantly h i g h e r level of arachidonic acid but also a markedly high resorption rate and a r e d u c t i o n in the m e a n e m b r y o n i c size. O n the basis of these data we f o u n d 1 m l / d a y to be the o p t i m u m therapeutic dose. This dosage significantly r e d u c e d neural tube defects in offspring of diabetic rats without significantly increasing resorption rates or affecting e m b r y o n i c growth. Several h u m a n studies have shown that a high intake of the essential fatty acids, either linoleic or 7-1inoleic acid, are associated with an i m p r o v e d plasma lipid profile and a r e d u c e d risk of both cardiovascular and retinal complications in patients with diabetes. 1215 Howard-Williams et al. 13 prospectively followed 149 individuals with newly diagnosed type I1 diabetes over a 7-year period. Subjects were r a n d o m l y assigned to either a low carbohydrate diet or m o d i f i e d fat diet that was rich in linoleic acid. Both o p h t h a h n o l o g i c and metabolic assessments were obtained at entry into the study and peri-
822
R e e c e et al.
October 1996 AmJ Obstet Gynecol
1S-
=, 10-
5::= O" o
-S"
-10 " 40
~,~
3O
I(J =..
uJ a uJ m
20
Iiv. ~ w Z
tO
CONTROL
DIABETIC
SAFFLOWEROIL
Fig. 2. Upper portion of figure demonstrates arachidonic acid levels (day 0 to day 12) in three groups. Lower portion shows incidence of neural tube defects in corresponding groups. odically thereafter. The authors found that poorly controlled patients with low levels of linoleic acid had a significantly greater frequency of retinopathy than wellcontrolled patients or patients with similarly unsatisfactory control but higher levels of linoleic acid. Houstmuller et al. 14reported a similar reduction in the progression of retinopathy in poorly controlled diabetic patients receiving a high polyunsaturated fatty acid diet. Over a 6-year period they found a 92% reduction in the development or progression of retinopathy. A multicenter trial was conducted in the United Kingdom to examine the effects of a high intake of polyunsaturated fatty acid on the course of mild diabetic neuropathy. 15 One h u n d r e d eleven subjects were randomly assigned to receive either placebo or polyunsaturated fatty acid supplements in this double-blind study. A statistically significant difference was demonstrated in 13 of the 16 neurologic parameters studied. However, treatment with polyunsaturated fatty acid was more effective in the well-controlled versus the poorly controlled patients. The results of this trial are consistent with the two studies described above and support the hypothesis that polyunsaturated fatty acid is protective against diabetesrelated complications. Currently no universally accepted mechanism for the etiopathologic features of diabetic-induced birth defects is recognized. 1~"]7 Although we and others have demonstrated that hyperglycemia has a teratogenic effect during organogenesis in both in vitro and in vivo models, the
exact mechanism has not been precisely elucidated. In our previous studies and in this current report we have demonstrated that hyperglycemia results in a functional deficiency of arachidonic acid. It is well established that the yolk sac membrane and cell membranes contain prostanoid metabolites of arachidonic acid as essential components. We postulate that arachidonic acid deficiency results in damage to embryonic cells and yolk sac membranes. This permits the intracellular influx of high concentrations of glucose across the mitochondrial membranes. The glucose is oxidized and byproducts of free oxygen radicals are removed by scavenging enzymes such as superoxide dismutase. Eriksson et al. ]8 have shown that in the presence of hyperglycemia these enzymes are saturated, resulting in the formation of excess free oxygen radicals that may have a direct effect on deoxyribonucleic acid and may inhibit the production of prostacyclin. Therefore the beneficial effects of arachidonic acid supplementation may be due to a restoration of membrane integrity with a restored balance of the phospholipid and prostanoids within the yolk sac and cell membranes. In conclusion, these data confirm our previous findings that hyperglycemia is a teratogen and, when embryos are exposed to it during organogenesis, it induces malformations. In addition, diabetic embryopathy is associated with an essential fatty acid deficiency state. Finally, the use of a dietary polyunsaturated fatty acid, such as safflower oil, increases blood levels of arachidonic acid,
Volume 175, Number 4, Part l AmJ Obstet Gynecol
thus m i n i m i z i n g the t e r a t o g e n i c effects o f hyperglycemia. T h e clinical a p p l i c a t i o n o f these data is yet to b e elucid a t e d b u t offer significant promise. It is h o p e d t h a t t h e s e f i n d i n g s may serve as a p r e l u d e to studies in h u m a n s e x p l o r i n g the p o t e n t i a l o f a dietary prophylaxis u s e d against d i a b e t e s - i n d u c e d c o n g e n i t a l m a l f o r m a t i o n s .
REFERENCES
1. Karlsson K, KjeUmer I. The outcome of diabetic pregnancies in relation to the mother's blood sugar level. Am J Obstet Gynecol 1972;112:213. 2. Fuhrinann K, Reiher H, Semmler K, et al. Prevention of congenital malformations in infants of insulin-dependent diabetic mothers. Diabetes Care 1983;6:219-23. 3. Goldrnan JA, Dicker D, Feldberg, et al. Pregnancy outcome in patients with insulin-dependent diabetes mellitus with preconceptional diabetic control: a comparative study. A m J Obstet Gynecol 1986;155:293-7. 4. Greene MF, Hare JW, ClohertyJE et al. First-trimester hemoglobin Aic and risk of major malformation and spontaneous abortion in diabetic pregnancy. Teratology 1989;39: 225-31. 5. KitzmillerJL, Gavin LA, Gin GD, et al. Preconception management of diabetes continued through early pregnancy prevents the excess frequency of major congenital anomalies in infants of diabetic mothers. JAMA 1991;265:731-6. 6. Pinter E, Reece EA, Leranth C, et al. Yolk sac failure in embryopathy due to hyperglycemia: ultrastructural analysis of yolk sac differentiation in rat conceptuses under hyperglycemic culture conditions. Teratology 1986;33:73-84. 7. Reece EA, Pinter E, Leranth CZ, et al. Ultrastructural anal-
Reece et al.
823
ysis of malformations of the embryonic neural axis induced by in vitro hyperglycemic conditions. Teratology 1985;32: 363-73. 8. Pinter E, Reece EA, Leranth C, et al. Arachidonic acid prevents hyperglycemia-associated yolk sac damage and embryopathy. Am J Obstet Gynecol 1986;155:691-702. 9. Olin BR, editor. Drug facts and comparisons. Philadelphia: JB Lippincott, 1990 Aug:2165. 10. McDonald-Gibson RG. Quantitative measurement ofarachidonic acid in tissues or fluids. In: Prostaglandins related substance: a practical approach. Oxford (UK): IRL Press, 1987:259-65. ] 1_ Suttie JW. Lipid chemistry: introduction to biochemistry. New York: Holt, Reinhart and Winston, 1972. 12. King RC, DobreeJH, Kok DA, Foulds WS, Dangerfield WG. Exudative diabetic retinopathy: spontaneous changes and effects of a corn oil diet. B r J Ophthalmol 1963;47:666-72. 13. Howard-Williams J, Patel P, Jelfs R, et al. Polyunsaturated fatty acids and diabetic retinopathy. B r J Ophthalmol 1985; 69:15-8. 14_ Houtsmuller AJ, van Hal-FerwerdaJ, Kahn KJ, Henkes HE. Favorable influences of linoleic acid on the progression of diabetic micro- and macro-angiopathy in adult onset diabetes mellitus. Prog Lipid Res 1982;20:377-86. 15. Keen H, Payan J, AllawiJ, et al. Treatment of diabetic neuropathy with 7-1inolenic acid. Diabetes Care 1993;16:8-15. 16. Reece EA, Pinter E, Homko C, Wu Y-K, Naftolin E The yolk sac theory: closing the circle on why diabetes associated malformations occur.J Soc Gynecol Invest 1994;1:3-13. 17. Reece EA, Homko CJ, Wu Y-K, Wiznitzer A. Metabolic fuel mixtures and diabetic embryopathy. Clin Perinatol 1993;20: 517-32. 18. Eriksson uJ, Hakan Borg LA, Forsberg H, StyrudJ. Diabetic embryopathy: studies with animal and in vitro models. Diabetes 1991;40(Suppl 2):94-8.
Key words: Diabetes, pregnancy, embryopathy, polyunsaturated fatty acids, arachidonic acid
Congenital malformations a m o n g infants of diabetic mothers occur at a rate about four to five times h i g h e r than that r e p o r t e d in the general population. Clinically there is a great diversity seen in the types of malformations associated with i n s u l i n - d e p e n d e n t diabetes mellitus. However, the organ systems most c o m m o n l y affected include the central nervous and cardiovascular systems. A threefold to 20-fold increase in neural tube defects has From the Department of Obstetrics, Gynecology, and Reproductive Sciences, Temple University School of Medicine, a and the Temple University School of Pharmacy.h Presented at the Sixteenth Annual Meeting of the Society of Perinatal Obstetricians, Kamuela, Hawaii, February 4-10, 1996. Reprint requests: E. Albert Reece, MD, Department of Obstetrics, Gynecolog3, and Reproductive Sciences, Temple University School of Medicine, 3401 N. Broad St., 7-OPB, Philadelphia, PA 19140. Copyright 9 1996 by Mosby-Year Book, Inc. 0002-9378/96 $5.00 + 0 6/6/75230 818
b e e n r e p o r t e d a m o n g the offspring of w o m e n with diabetes. This increased m a l f o r m a t i o n rate accounts for about 40% of the deaths a m o n g infants of diabetic mothers. 1 It has b e e n shown that altered metabolic fuels, including hyperglycemia, a m o n g p r e g n a n t diabetic w o m e n is associated with an increased m a l f o r m a t i o n rate. Conversely, recent studies have d e m o n s t r a t e d that preconception glycemic control results in a decreased rate of malformations, z-~ Unfortunately, however, meticulous family p l a n n i n g is necessary to effect this benefit. We previously r e p o r t e d our results with the use of increasing concentrations of v-glucose during o r g a n o g e n esis to induce malformations in a dose-related fashion in the postimplantation rat e m b r y o culture model. 6' 7 In o u r rat m o d e l we specifically studied the effects of hyperglycemia on neural tube development. F u r t h e r m o r e , when arachidonic acid was added in increasing concentrations,
R e e c e et al.
Volume 175, Number 4, Part l AmJ Obstet Gynecol
819
Table I. Relationship between safflower oil dosage and pregnancy outcomes
Dose
No. of embryos
0.5 ml/day SO 1.0 ml/day SO 2.0 ml/day SO Diabetic controls fed normal diet
187 162 67 280
AAlevel(vg/ml), day 12- day 0 2.7• 3.0• 86.96• -3.65•
5.4* 9.3 27.4* 8.2
AA, Arachidonic acid; NTD,neural tube defects; CT~L,crown-rump length; *p < 0.05.
malformation rates declined, even in the presence of the 100% teratogenic dose of excess D-glucose. s The purpose of the current in vivo study was to test the hypothesis that diabetic embryopathy is associated with an arachidonic acid deficiency state and that daily dietary supplementation with a polyunsaturated fatty acid high in arachidonic acid will obviate the teratogenic effects of hyperglycemia. Methods Test animals. Seventy- to 90-day-old Sprague-Dawley rats with weights between 250 and 300 gm were used for the current study. Rats were mated overnight and vaginal smears were examined the next morning. If spermatozoa were present, pregnancy day 0 was determined. Diabetes was induced in some (experimental group), whereas others remained nondiabetic (controls). Diabetes was induced by an intravenous injection of streptozocin (65 mg/kg) on pregnancy day 6 because preconceptionally induced diabetes may be associated with infertility in rats. Furthermore, streptozocin used during pregnancy is readily eliminated within 24 hours and thus will be removed from the circulation before the critical period of organogenesis. 9 Glucose levels between 300 and 380 m g / d l were achieved by simply withholding daily insulin therapy. (Normal glucose level in the rat is approximately 150 mg/dl.) Rats were randomly assigned to one of five groups. Group 1 animals were nondiabetic controls receiving a normal diet, and group 2 animals were fed a normal diet supplemented with 1 ml of safflower oil daily starting on pregnancy day 6 until the day embryos were explained. The remaining three groups had diabetes experimentally induced with strePtozocin. Group 3 animals were diabetics receiving a normal diet, group 4 were diabetics receiving a normal diet plus an oral sham feeding of normal saline solution, and group 5 animals were diabetics fed a normal diet and supplemented with safflower oil as in control group 2. The oral agents (normal saline solution and safflower oil) were administered with a tuberculin syringe daily starting on day 6 of gestation. Determination of arachidonic acid dosage. A pilot study was undertaken to determine the most effective dose for supplementation. The pilot study was conducted in a
Absorptions] (%)
NTD] (%) 16.2" 7.6 5.5* 20.0
CRL (mm) 0.54• 0.52• 0.35• 0.63•
SO, safflower oil.
m a n n e r similar to the larger study described in detail in this section. Four diabetic groups were chosen and these animals received varying doses of safflower oil (2 ml/day, 1 ml/day, and 0.5 ml/day) or normal diet alone. A dose of 1 ml per daywas found to be most effective and was the dose used in groups 2 and 5 (Table I). Assessment parameters. Blood samples were collected by cardiac puncture on day 0 and day 12 of pregnancy and analyzed for arachidonic acid. Blood was obtained daily from the tail vein and analyzed for glucose levels. Arachidonic acid assays were performed by capillary gas chromatography with hexadecanoic acid as the internal standard by a technique described elsewhere) ~ O n pregnancy day 12 the rats were killed; the embryos were examined and measurements were taken u n d e r light microscopy with x l 0 magnification. The crownrump length of each embryo was measured. The embryos were categorized as morphologically normal or showing neural tube defects or other malformations. These included microcephaly, abnormal cervical spine, reverse tail flexion, tubular heart formation, and pericardial effusion. The normal embryos demonstrated correct body flexure and both the anterior and posterior neural pores were closed. Embryos with malrotation, delayed closure of a single neural pore, or an abnormally open neural tube were defined as having a neural tube defect. Embryos in various stages Of resorption were denoted as resorbed and no further analysis was performed on the latter group. Statistical analyses was performed with Student t tests, )C, and analysis of variance where appropriate. Data are expressed as mean + SD. Results In the pilot study animals receiving either 1 or 2 ml/day showed a significant decrease in neural tube defects compared with the diabetic control group (7.6% and 5.5%, respectively). Supplementation with 2 ml/day dramatically increased serum arachidonic acid levels and reduced the neural tube defect rate. However, the incidence of embryonic resorption was significantly higher (27%) than in either of the other two groups (5.4% with 0.5 ml/day and 9.3% with 1 ml/day, p < 0.05) (Table I). In addition, 2 ml/day safflower oil had an adverse effect of
820
Reece et al.
October 1996 AmJ Obstet Gynecol
Table II. Relationship between safflower oil dosage and pregnancy o u t c o m e s
Controls Group 1, normal Group 2, normal Diabetics Group 3, normal Group 4, normal saline solution Group 5, normal
Resorptions (%)
NTD (%)
0.62• 0.63_+0.1
10/163 (6.1) 8/97 (9.3)
4.8 4.1
360.94 _+48 350.95 _+48
0.63_+0.2 0.68_+0.4
23/280 (8.2) 4/78 (5.1)
20.0 28.2
353 _+36
0.52_+0.2
15/162 (9.3)
7.6
No. of embryos
Glucose level (mg/dl)
diet diet and PUFA
163 97
133.25 _+22 143.5 -+ 24
diet diet and normal sham feeding diet and PUFA
280 78 162
CRL (mm)
CRL, Crown-rump length; NTD, neural tube defects; PUFA, polyunsaturated fatty acid. *Comparison of arachidonic acid levels between day 0 and day 12 (paired t tests).
Table IlL S e r u m arachidonic acid levels a m o n g offspring of controls and diabetic rats AA levels (izg/ml) Day 0
I
Day 12
Significance*
1
Controls Group 1, normal Group 2, normal Diabetics Group 3, normal Group 4, normal Group 5, normal
diet diet and PUFA
17.58_+4.86 14.15_+3.62
17.83_+5.84 15.22•
p>0.1 p>0.1
diet diet and normal saline solution sham feeding diet and PUFA
17.83 + 5.84 17.15_+4.15 16.99 _+8.59
14.18 + 2.58 13.43_+3.52 19.99 -+ 7.99
p< 0.05 p< 0.05 p< 0.05
AA, Arachidonic acid; PUFA, polyunsaturated fatty acid. *Comparison of arachidonic acid levels between day 0 and day 12 (paired t tests). fetal growth as shown by a statistically significant reduction in c r o w n - r u m p length in this g r o u p c o m p a r e d with controls and animals receiving lower doses of supplementation (Table II). T h e r e f o r e 1 m l / d a y safflower oil supplem e n t a t i o n was used for further studies, and the remainder of the data will be based on this dosage. Arachidonic acid levels after supplementation with polyunsaturated fatty acids. Arachidonic acid in the maternal s e r u m was analyzed on days 0 and 12 of pregnancy. T h e arachidonic acid levels on day 0 and day 12 in maternal s e r u m of the nondiabetic controls are shown in Table III. T h e diabetic rats fed a n o r m a l diet only d e m o n s t r a t e d a m a r k e d r e d u c t i o n in arachidonic acid levels by day 12 to a m e a n level that was statistically significant (14.18 _+ 2.58 D g / m l vs 17.83 + 5.84 p g / m l , p < 0.0001) f r o m that observed in controls. Arachidonic acid b l o o d levels between days 0 and 12 after 1 m l / d a y of safflower oil supplementation were significantly h i g h e r than in both the diabetic animals fed a n o r m a l diet and n o r m a l controls (Table III). However, there was no significant difference in ser u m arachidonic acid levels on day 12 between supplem e n t e d controls (group 2) and s u p p l e m e n t e d diabetic animals (group 5). Pregnancy outcomes. G r o u p 1 ( n o n d i a b e t i c / n o r m a l diet) c o n t a i n e d 163 embryos. A l t h o u g h each rat pregnancy had multiple embryos, embryos ranged f r o m 8 to 20 per pregnancy with an average of 15 per rat. T h e m e a n m a t e r n a l glucose level was 133 + 22 m g / d l . T h e resorp-
tion rate in this group was 6.13% (10/163); the crownr u m p length of the embryos was 0.62 _+0.06 cm, and the incidence of neural tube defects was 4.8%. G r o u p 2 ( n o n d i a b e t i c / n o r m a l diet plus safflower oil) c o n t a i n e d 97 embryos. T h e m e a n glucose level was 143.5 _+ 24 m g / d l . T h e resorption rate was 9.3% and the crown-rump length was 0.63 + 0.1 cm for this group. T h e incidence of neural tube defects was 4.1%. G r o u p 3 ( d i a b e t i c / n o r m a l diet) c o n t a i n e d 280 embryos. T h e m e a n m a t e r n a l glucose level was 360 _+48 m g / d l , which was significantly h i g h e r than in the control group (p < 0.05). In this g r o u p the resorption rate was n o t significantly different from that of the control g r o u p of 8.2% (23/280), n o r was the m e a n c r o w n - r u m p length of the embryos different (0.63 _+0.18 cm). However, neural tube defects were increased by sevenfold c o m p a r e d with controls (20.0% vs 4.8%, p < 0.05). This resulted in nearly o n e fourth of the embryos p e r rat pregnancy being malf o r m e d (Fig. 1). G r o u p 4 ( d i a b e t i c / n o r m a l diet plus sham feeding) contained 78 embryos. Blood glucose levels were 350 _+48 m g / d l in this group, which had a resorption rate of 5.1% and a m e a n crown-rump length of 0.68 _+0.4 cm. Arachidonic acid levels fell f r o m 17.15 _+4.15 p g / m l to 13.43 + 3.52 p g / m l (p < 0.05), whereas the incidence of neural tube defects increased significantly to 28.2% in this group. G r o u p 5 ( d i a b e t i c / n o r m a l diet plus safflower oil) con-
Volume 175, Number 4, Part 1 AmJ Obstet Gynecol
Reeoe et al.
821
Fig. 1. Twelve-day-old normal embryo is seen on right next to embryo with neural robe defect (left). Arrows outline defect. tained 162 embryos with m e a n maternal glucose levels of 353 _4-36 m g / d l . This g r o u p had a resorption rate of 9.3% (15/162) and a m e a n e m b r y o n i c c r o w n - r u m p length of 0.52 + 0.2 cm, n o t significantly different f r o m the control groups. However, neural tube defects in this g r o u p were significantly decreased c o m p a r e d with the diabetic g r o u p fed n o r m a l diet without arachidonic s u p p l e m e n t a t i o n (20.0% vs 7.6%, p < 0.005) (Fig. 2). This r e d u c e d incid e n c e of neural tube defect was n o t statistically different f r o m that of the nondiabetic controls.
Comment In our previous studies with the postimplantation rat e m b r y o culture we d e m o n s t r a t e d that arachidonic acid s u p p l e m e n t a t i o n r e d u c e d the rates of neural tube defects even in the presence of a 100 % teratogenic dose of excess D-glucose. 8 In the c u r r e n t study we replicated in vivo the results of our findings obtained in vitro. We d e m o n strated that in diabetic rats fed a n o r m a l diet the arachidonic acid s e r u m level is significantly decreased, whereas the m a l f o r m a t i o n rate is significantly increased above control rates. Conversely, the malformation rate was significantly r e d u c e d in diabetic rats whose diets were s u p p l e m e n t e d with a dietary source of arachidonic acid. These data support our hypothesis that diabetic embryopathy is associated with an arachidonic acid deficiency state and that daily s u p p l e m e n t a t i o n with a polyunsaturated fatty acid high in arachidonic acid will obviate the teratogenic effects of hyperglycemia and o t h e r metabolic fuels. In the c u r r e n t experiments we used safflower oil as a dietary source of arachidonic acid, which served as a p h a r m a c o l o g i c prophylaxis against the teratogenic effects of hyperglycemia. Safflower oil is a polyunsaturated fatty acid that is converted to linoleic acid and eventually
to arachidonic acid by a tightly linked enzyme s e q u e n c e ? 1 We d e m o n s t r a t e d a direct relationship between safflower oil s u p p l e m e n t a t i o n and arachidonic acid levels in a dose-related fashion. As can be seen in Table I, as arachidonic acid levels increased, the incidence of neural tube defects decreased in the offspring of the diabetic rats. However, w h e n arachidonic acid levels in the diabetic rats a p p r o x i m a t e d levels seen in nondiabetic controls, the incidence of e m b r y o n i c neural tube defects still r e m a i n e d significantly elevated above the background. It appears that diabetic animals n e e d e d h i g h e r levels of arachidonic acid to c o m p e n s a t e for or override the metabolic effects of hyperglycemia to result in a r e d u c t i o n in the incidence of neural tube defects toward b a c k g r o u n d rates. However, d o u b l i n g the dose of safflower oil to 2 m l / d a y resulted in a significantly h i g h e r level of arachidonic acid but also a markedly high resorption rate and a r e d u c t i o n in the m e a n e m b r y o n i c size. O n the basis of these data we f o u n d 1 m l / d a y to be the o p t i m u m therapeutic dose. This dosage significantly r e d u c e d neural tube defects in offspring of diabetic rats without significantly increasing resorption rates or affecting e m b r y o n i c growth. Several h u m a n studies have shown that a high intake of the essential fatty acids, either linoleic or 7-1inoleic acid, are associated with an i m p r o v e d plasma lipid profile and a r e d u c e d risk of both cardiovascular and retinal complications in patients with diabetes. 1215 Howard-Williams et al. 13 prospectively followed 149 individuals with newly diagnosed type I1 diabetes over a 7-year period. Subjects were r a n d o m l y assigned to either a low carbohydrate diet or m o d i f i e d fat diet that was rich in linoleic acid. Both o p h t h a h n o l o g i c and metabolic assessments were obtained at entry into the study and peri-
822
R e e c e et al.
October 1996 AmJ Obstet Gynecol
1S-
=, 10-
5::= O" o
-S"
-10 " 40
~,~
3O
I(J =..
uJ a uJ m
20
Iiv. ~ w Z
tO
CONTROL
DIABETIC
SAFFLOWEROIL
Fig. 2. Upper portion of figure demonstrates arachidonic acid levels (day 0 to day 12) in three groups. Lower portion shows incidence of neural tube defects in corresponding groups. odically thereafter. The authors found that poorly controlled patients with low levels of linoleic acid had a significantly greater frequency of retinopathy than wellcontrolled patients or patients with similarly unsatisfactory control but higher levels of linoleic acid. Houstmuller et al. 14reported a similar reduction in the progression of retinopathy in poorly controlled diabetic patients receiving a high polyunsaturated fatty acid diet. Over a 6-year period they found a 92% reduction in the development or progression of retinopathy. A multicenter trial was conducted in the United Kingdom to examine the effects of a high intake of polyunsaturated fatty acid on the course of mild diabetic neuropathy. 15 One h u n d r e d eleven subjects were randomly assigned to receive either placebo or polyunsaturated fatty acid supplements in this double-blind study. A statistically significant difference was demonstrated in 13 of the 16 neurologic parameters studied. However, treatment with polyunsaturated fatty acid was more effective in the well-controlled versus the poorly controlled patients. The results of this trial are consistent with the two studies described above and support the hypothesis that polyunsaturated fatty acid is protective against diabetesrelated complications. Currently no universally accepted mechanism for the etiopathologic features of diabetic-induced birth defects is recognized. 1~"]7 Although we and others have demonstrated that hyperglycemia has a teratogenic effect during organogenesis in both in vitro and in vivo models, the
exact mechanism has not been precisely elucidated. In our previous studies and in this current report we have demonstrated that hyperglycemia results in a functional deficiency of arachidonic acid. It is well established that the yolk sac membrane and cell membranes contain prostanoid metabolites of arachidonic acid as essential components. We postulate that arachidonic acid deficiency results in damage to embryonic cells and yolk sac membranes. This permits the intracellular influx of high concentrations of glucose across the mitochondrial membranes. The glucose is oxidized and byproducts of free oxygen radicals are removed by scavenging enzymes such as superoxide dismutase. Eriksson et al. ]8 have shown that in the presence of hyperglycemia these enzymes are saturated, resulting in the formation of excess free oxygen radicals that may have a direct effect on deoxyribonucleic acid and may inhibit the production of prostacyclin. Therefore the beneficial effects of arachidonic acid supplementation may be due to a restoration of membrane integrity with a restored balance of the phospholipid and prostanoids within the yolk sac and cell membranes. In conclusion, these data confirm our previous findings that hyperglycemia is a teratogen and, when embryos are exposed to it during organogenesis, it induces malformations. In addition, diabetic embryopathy is associated with an essential fatty acid deficiency state. Finally, the use of a dietary polyunsaturated fatty acid, such as safflower oil, increases blood levels of arachidonic acid,
Volume 175, Number 4, Part l AmJ Obstet Gynecol
thus m i n i m i z i n g the t e r a t o g e n i c effects o f hyperglycemia. T h e clinical a p p l i c a t i o n o f these data is yet to b e elucid a t e d b u t offer significant promise. It is h o p e d t h a t t h e s e f i n d i n g s may serve as a p r e l u d e to studies in h u m a n s e x p l o r i n g the p o t e n t i a l o f a dietary prophylaxis u s e d against d i a b e t e s - i n d u c e d c o n g e n i t a l m a l f o r m a t i o n s .
REFERENCES
1. Karlsson K, KjeUmer I. The outcome of diabetic pregnancies in relation to the mother's blood sugar level. Am J Obstet Gynecol 1972;112:213. 2. Fuhrinann K, Reiher H, Semmler K, et al. Prevention of congenital malformations in infants of insulin-dependent diabetic mothers. Diabetes Care 1983;6:219-23. 3. Goldrnan JA, Dicker D, Feldberg, et al. Pregnancy outcome in patients with insulin-dependent diabetes mellitus with preconceptional diabetic control: a comparative study. A m J Obstet Gynecol 1986;155:293-7. 4. Greene MF, Hare JW, ClohertyJE et al. First-trimester hemoglobin Aic and risk of major malformation and spontaneous abortion in diabetic pregnancy. Teratology 1989;39: 225-31. 5. KitzmillerJL, Gavin LA, Gin GD, et al. Preconception management of diabetes continued through early pregnancy prevents the excess frequency of major congenital anomalies in infants of diabetic mothers. JAMA 1991;265:731-6. 6. Pinter E, Reece EA, Leranth C, et al. Yolk sac failure in embryopathy due to hyperglycemia: ultrastructural analysis of yolk sac differentiation in rat conceptuses under hyperglycemic culture conditions. Teratology 1986;33:73-84. 7. Reece EA, Pinter E, Leranth CZ, et al. Ultrastructural anal-
Reece et al.
823
ysis of malformations of the embryonic neural axis induced by in vitro hyperglycemic conditions. Teratology 1985;32: 363-73. 8. Pinter E, Reece EA, Leranth C, et al. Arachidonic acid prevents hyperglycemia-associated yolk sac damage and embryopathy. Am J Obstet Gynecol 1986;155:691-702. 9. Olin BR, editor. Drug facts and comparisons. Philadelphia: JB Lippincott, 1990 Aug:2165. 10. McDonald-Gibson RG. Quantitative measurement ofarachidonic acid in tissues or fluids. In: Prostaglandins related substance: a practical approach. Oxford (UK): IRL Press, 1987:259-65. ] 1_ Suttie JW. Lipid chemistry: introduction to biochemistry. New York: Holt, Reinhart and Winston, 1972. 12. King RC, DobreeJH, Kok DA, Foulds WS, Dangerfield WG. Exudative diabetic retinopathy: spontaneous changes and effects of a corn oil diet. B r J Ophthalmol 1963;47:666-72. 13. Howard-Williams J, Patel P, Jelfs R, et al. Polyunsaturated fatty acids and diabetic retinopathy. B r J Ophthalmol 1985; 69:15-8. 14_ Houtsmuller AJ, van Hal-FerwerdaJ, Kahn KJ, Henkes HE. Favorable influences of linoleic acid on the progression of diabetic micro- and macro-angiopathy in adult onset diabetes mellitus. Prog Lipid Res 1982;20:377-86. 15. Keen H, Payan J, AllawiJ, et al. Treatment of diabetic neuropathy with 7-1inolenic acid. Diabetes Care 1993;16:8-15. 16. Reece EA, Pinter E, Homko C, Wu Y-K, Naftolin E The yolk sac theory: closing the circle on why diabetes associated malformations occur.J Soc Gynecol Invest 1994;1:3-13. 17. Reece EA, Homko CJ, Wu Y-K, Wiznitzer A. Metabolic fuel mixtures and diabetic embryopathy. Clin Perinatol 1993;20: 517-32. 18. Eriksson uJ, Hakan Borg LA, Forsberg H, StyrudJ. Diabetic embryopathy: studies with animal and in vitro models. Diabetes 1991;40(Suppl 2):94-8.