- Email: [email protected]
Cellular Growth in Intrauterine Malnutrition
Cellular Growth in Intrauterine Malnutrition MYRON WINICK, M.D.*
Growth may be defined simply as an increase in the size of an animal or an individual organ. This increase in organ size is the result of a continuous accretion of protein and in some cases lipids. Hence the rate of net protein synthesis determines the rate of growth. When protein synthesis and protein degradation reach equilibrium, growth ceases. Since net protein synthesis is linear in all organs in utero and for a varying period postnatally, growth would appear to be a homogeneous process. However this is only one aspect of growth. The protein in any organ is packaged within cells. The same total organ protein may be contained in numerous small cells or in a few larger cells. The constant amount of DNA within the diploid nucleus of any cell in a particular species makes possible accurate determination of the total number of cells by chemical methods. 2 During normal growth, total organ DNA content or cell number increases linearly, then begins to decelerate, and finally reaches a maximum long before the organ size, as determined by net protein accretion, has reached its maximum. 16 As a consequence of this pattern of DNA synthesis, three phases of growth can be described: proportional increase in weight, protein, and DNA content (hyperplasia); an increase in DNA slower than the increase in protein and weight (hyperplasia and concomitant hypertrophy); and finally no further increase in DNA content, with net protein and weight continuing to increase at the same rate (hypertrophy). The growing period, then, is not a homogeneous process when viewed in cellular terms. It is possible that stimuli which retard or accelerate growth may result in different effects depending on whether they influence the rate of cell division (DNA synthesis) or whether they affect the subsequent increase in cell size. *Associate Professor of Pediatrics and Director, Division of Growth and Development, Cornell University Medical College, New York, New York; recipient of a Career Development Award (No. 1-K04·HD-18869-01) from the National Institute of Child Health and Human Development Reported in part at the Eighth Annual Meeting of the Pan American Health Organization Advisory Committee on Medical Research, Washington, June 9-13, 1969.
Pediatric Clinics of North America- Vol. 17, No.1, February, 1970
69
70
MYRON WINICK
In rats it has been shown that early malnutrition will curtail the rate of cell division if it occurs during the period of hyperplasia. In brain, this period is prior to weaning. Not only will neonatal undernutrition curtail the rate of cell division, but it will result in a permanent deficit in brain cell numberY Malnutrition after weaning will prevent the subsequent increase in c;ell size, and this effect can be reversed by later rehabilitation. During the period of rapid cellular proliferation, undernutrition will affect the rate of cell division in any brain area where cells are dividing and in any cell type in the process of division. 14 Figure 1 demonstrates that in animals malnourished from birth, cerebellum, where cells are most rapidly dividing, is affected earliest and most severely. Figure 2 focuses on the cell types involved. In cerebral cortex, neurons are not dividing postnatally and hence are not affected by undernutrition. The number of glia, however, is reduced. In cerebellum, all cell types are reduced. The areas under the lateral ventricle and under the third ventricle are other areas where primitive neuronal division is curtailed. Thus discrete brain areas are disproportionally affected by neonatal undernutrition. The magnitude of the effect is proportional to the rate of ceJ) division in a particular area. In contrast, increased nutrition will accelerate cell division and result in a permanent increase in brain cell number, if it occurs during the proliferati ve phase. 15 ,19 Thus within certain limits the number of cells ultimately present within the adult brain may be programmed during early life, during that DNA MALNUTRITION, PER CENT OF NORMAL
100
80
60
40
20
o
8
10 14 17 21 CEREBRUM
HI PPOCAMPUS
CEREBELLUM
Figure 1. Effect of early malnutrition on DNA content of various parts of rat brain.
71
CELLULAR GROWTH IN INTRAUTERINE MALNUTRITION
_NORMAL ~ MALNUTRITION
Cereb.ellum
Lateral ventricle ,.......-"-.
25
20
~
c;; 15 u
Cerebral cortex
"0 <1>
]i .!!! ~
10
period in postnatal life when brain cells are still dividing. Can this type of programming take place during prenatal life? Certainly all fetal organs are in the proliferative growth phase and therefore would appear susceptible to permanent cellular effects. However, the fetus is protected from the environment by the mother and the placenta. Thus, one may ask, can the mother or the placenta protect the growing fetus from such adverse environmental stimuli as maternal malnutrition, maternal disease, or insufficient uterine vascular supply? In animals it is possible to monitor the effects of various maternal stresses on the cellular growth of both placenta and fetal organs. In the human the fetus is not available for these types of organ analysis and therefore the placenta, which is accessible after birth, is the only organ that can be examined when the fetus is born in a viable condition. Thus if changes in the pattern of cellular growth in placenta may be correlated with cellular growth of fetal organs in animals, perhaps similar changes in human placenta will give us a clue to the effects of similar maternal stresses on the human fetus. . Normal placental growth, both in the rat and in the human, proceeds in the same way as previously described in the organs of the rat. In the rat placenta, DNA synthesis and hence cell division stops at 17 days of a 21-day gestation. I8 In the human placenta, cell division continues until around the 34th to 36th week of gestation. 20 In both cases net protein synthesis continues to term. Thus the same three phases of cellular growth may be described as in other organs of the rat. Stimuli imposed
72
MYRON WINICK
prior to the 17th day or 36th week, respectively, should lead to a permanent reduction in placental cell number, whereas stimuli that are active only after cell division has stopped should affect the size of individual cells but not the total number of cells. Hence examination of the ultimate cellular make-up of the placenta should give us an idea of when, during the course of growth, a stimulus has been active. With these principles in mind, let us examine the effects of a number of precisely timed experimental stimuli on placental and fetal growth in the rat. Then we shall examine the ultimate effects of naturally occurring stimuli on the human placenta. Finally we will attempt to determine the time during pregnancy when these stimuli are active. Maternal protein restriction in rats will retard both placental and fetal growth. In placenta, cell number (DNA content) was reduced by 13 days after conception, cell size (protein/DNA) remained normal, and the RNA/DNA ratio was markedly elevated. Retardation in fetal growth first became apparent at 15 days. After this there was a progressive decrease in cell number in all the organs studied. By term there were only about 85 per cent of the number of brain cells as in control animals (Table 1). These data agree with previous data of Zamenhof, which showed a similar reduction in total brain cell number in term fetuses whose mothers were exposed to a slightly different type of nutritional deprivation. 22 Thus the cellular changes produced by severe prenatal food restriction are reflected in the placenta even earlier than in the fetus, but retardation of cell division in all fetal organs including brain can be clearly demonstrated. By employing radioautography after injecting the mother with tritiated thymidine, cell division can be assessed in various discrete brain regions.! Differential regional sensitivity can be demonstrated in this way by the 16th day of gestation in the brains of fetuses of proteinrestricted mothers (Fig. 3). The cerebral white and gray matter are mildly affected. The area adjacent to the third ventricle and the subiculum are moderately affected, whereas the cerebellum and the area directly adjacent to the lateral ventricle are markedly affected. These data again demonstrate that the magnitude of the effect produced on cell division is directly related to the actual rate of cell division at the time the stimulus is applied. Moreover they demonstrate that the maternal-placental barrier in the rat is not effective in protecting the fetal brain from discrete cellular effects caused by maternal food restriction. The subsequent course of these animals born of protein-restricted mothers can be examined. Chow has reported that even if these animals are raised normally on foster mothers they demonstrate a permanent impairment in their ability to utilize nitrogen. 3 Data from our own laboratory demonstrate that if these animals are nursed on normal foster mothers in normal-sized litters, they will remain with a deficit in total brain cell number at weaning. Thus we can again see early programming of the ultimate number of brain cells. This program, moreover, is written in utero in response to maternal nutrition. These same newborn pups of protein-restricted mothers may be subjected to postnatal nutritional manipulation. If they are raised in
73
CELLULAR GROWTH IN INTRAUTERINE MALNUTRITION
Table 1. Composition of Various Tissues in Newborn Rats Subjected to Maternal Undernutrition PER CENT OF NORMAL CONTROL TISSUE
Weight
Protein
RNA
DNA
Whole animal Brain Heart Lung Liver Kidney
87 91 84 82 82 84
81 85 84 85 80 81
83 82 79 85 85 82
81 84 81 89 85 85
litters of three on normal foster mothers until weaning, the deficit in total number of brain ce1Is may be almost entirely reversed. Although quantitatively the number of cells approaches normal, qualitatively the deficit at birth might very well be made up by an increase in cell number in different areas from those most affected in utero. Thus although it may appear that optimally nourishing pups after exposing them to prenatal undernutrition will reverse the cellular effects, this may not actually be so in specific brain areas. Perhaps most nearly analogous to the situation in humans is exposing, these pups, malnourished in utero, to subsequent postnatal deprivation. One can raise these animals on foster mothers in groups of 18. Animals so reared show a marked reduction in brain cell number by
25
_
NORMAL
~
MATERNAL PROTEIN RESTR I CTI ON
20
5
o
Cerebrum (gray)
Cerebrum (white)
Third ventricle
Subiculum Cerebellum
Lateral ventricle
Figure 3. Effect of prolonged maternal protein restriction on various brain regions in 16-day rat embryo.
74
MYRON WINICK
weaning. This effect is much more pronounced than the effect of either prenatal or postnatal undernutrition alone. Animals subjected to prenatal malnutrition alone as previously described show a 15 per cent reduction in total brain cell number at birth. Those subjected only to postnatal malnutrition show a similar 15 to 20 per cent reduction in cell number at weaning. In contrast, these "doubly deprived" animals demonstrate a 60 per cent reduction in total brain cell number by weaning (Fig. 4). These data demonstrate that malnutrition applied constantly throughout the entire period of brain cell proliferation will result in a profound reduction in brain cell number, greater than the sum of effects produced during various parts of the proliferative phase. It would appear that the duration of malnutrition as well as the severity during this early critical period is extremely important in determining the ultimate cellular make-up of the brain. In rats it has been shown that clamping the uterine artery supplying one horn of the uterus will curtail fetal growthY The cellular events accompanying this fetal growth failure can be examined. Regardless of when the clamping is performed, cell division is curtailed in all fetal organs except brain (Table 2). In contrast, DNA content of placenta is reduced only if the clamping is done before the 17th day; after this time cell size (proteinIDNA) is reduced. The RNAIDNA ratio in placenta increases following uterine artery ligation at any time. 13 The exact significance of this elevated RNA/DNA ratio is not known. It has been noted under a variety of conditions involving tissue "stress." In cardiac hypertrophy secondary to experimental aortic ligation,6 in uterine hypertrophy induced by estrogen,S and in muscles exposed to repeated nerve stimulation/ one of the first changes noted is an increase in the RNA/ DNA ratio. 100
TOTAL BRA I N CELL NUMBER
80
~....
g
60
'0 C QJ
Figure 4. Comparison of the effect on brain cell number of prenatal malnutrition, postnatal malnutrition, and "combined" malnutrition.
..., ~
c...
40 20
o
Prenatal Postnatal Prenatal malnutrition malnutrition & postnatal (birth) (weaning) malnutrition (weaning)
75
CELLULAR GROWTH IN INTRAUTERINE MALNUTRITION
Table 2. Composition of Various Tissues in Newborn Rats After Clamping the Uterine Artery of the Mother at 17 Days of Gestation PER CENT OF NORMAL CONTROL TISSUE
Weight
Protein
Whole animal Brain Heart Lung Liver Kidney
67 91 84 62 62 64
95 84 65.· 70 61
71
RNA
DNA
63 103 79 55 75 82
99 91 59 55 75
71
Organs of animals delivered on the 21st day from ligated horns and then foster nursed to weaning do not attain the expected number of cells by weaning. Thus again a change produced in utero will persist throughout life. In these animals, however, brain was normal at birth and remained normal at weaning. Recently Zamenhof has demonstrated that artificially reducing the number of fetuses in utero will result in offspring with an increased number of brain cells. 10 Certainly then, in rats, manipulation of the maternal environment during gestation.will result in permanent cellular changes in the offspring. The effects of similar naturally occurring environmental stresses during human pregnancies are more difficult to demonstrate. However, certain clear changes can be noted. Placentas from infants with "intrauterine growth failure" show fewer cells and an increased RNA/DNA ratio when compared to controls.12 Fifty per cent of placentas from an indigent population in Chile showed similar findings. Placentas from a malnourished population in Guatemala had fewer cells than normal. 4 In a single case of anorexia nervosa in which a severely emaciated mother carried to term and gave birth to a 2500 gm. infant, the placenta contained less than 50 per cent of the expected number of cells (Fig. 5). Thus both vascular insufficiency and maternal malnutrition will curtail cell division in human placenta. The cellular make-up of the placenta in both of these situations strongly suggests that both stimuli have been active for some time prior to the 34th to 36th week of gestation. The effects of these stimuli on cellular growth of the fetus are more difficult to assess. In both situations fetal growth is retarded and birth weight reduced. 9 Indirect evidence would suggest that cell division in the human fetus might be retarded by maternal undernutrition. If one examines available data on infants who died after exposure to severe postnatal malnutrition, three separate patterns emerge. Breast-fed infants malnourished during the second year have a reduced protein/DNA ratio but a normal brain DNA content. Full-term infants who subsequently died of severe food deprivation during the first year of life had a 15 to 20 per cent reduction in total brain cell number. Infants weighing 2000 gm. or less at birth who subsequently died of severe undernutrition during the first year of life showed a 60
76
MYRON WINICK
....
80
Q)
.J::l
E
::::l
c:
60
Q;
u
'" E .... 0
z
40
'f/<.
20 0
IUGF
Indigent population Malnourished Anorexia in Santiago, Chile population nervosa in Guatemala·
( ) Number of cases .:' Data of Dayton et aI. Figure 5. Comparison of placental DN A content in various types of maternal deprivation.
100
MARASMUS (low bi rth weight)
MARASMUS (normal bi rth weight)
KWASHIORKOR
n:I
E .... o
c:
~
« z
50
a
o Figure 6. Comparison of brain cell number in marasmus, marasmus plus low birth weight, and kwashiorkor.
r
CELLULAR GROWTH IN INTRAUTERINE MALNUTRITION
77
per cent reduction in total brain cell number (Fig. 6).21 It is possible that these children were deprived in utero and represent a clinical counterpart of the "doubly deprived" animal. It is also possible that these were true premature infants and that the premature is much more susceptible to postnatal malnutrition than the full-term infant. Regional growth in the human brain is somewhat different than in the rat brain. Cell division stops at about the same time (8 to 10 months) in cerebellum, cerebrum, and brain stem. Severe malnutrition during this proliferative phase retards cell division in all three of these regions. The rate of cell division is about the same in cerebellum and cerebral cortex, and both are severely affected by malnutrition. In comparison to the rat,5 cell division in human cerebrum is much more rapid during postnatal growth, and human cerebrum is much more affected by postnatal malnutrition. Thus postnatal malnutrition curtails cell division in human brain as it does in rat brain. Prenatal stimuli affect human placenta in much the same way as rat placenta. Although at this time human fetal data are still sketchy, the suggestion that cell division in fetal organs can be curtailed by maternal undernutrition appears to be valid enough to require further studies. In view of the tremendous public health implications of this possibility and in view of the evidence in animals which demonstrates that these changes are permanent, it would seem to me that every effort should be made quickly to confirm or rule out the possibility that undernutrition of the mother may permanently reduce the number of brain cells in her offspring.
REFERENCES 1. Altman, J., and Das, G.: Autoradiographic and histological studies of postnatal neuronogenesis. I. A longitudinal investigation of the kinetics, migration and transformation of cells incorporating tritiated thymidine in infant rats with special reference to postnatal neurogenesis in some brain regions. J. Compo Neurol., 126:337, 1966. 2. Boivin, A., Vendrely, R., and Vendrely, C.: L'Acide desoxyribonucleique du noyau cellulaire, despositaire des caracteres hereditaires; arguments d'ordre analytique. Compt. Rend. Acad. Sci., 226:1061,1948. 3. Chow, B., presented at Gordon Research Conference on Nutrition, New London, New Hampshire, 1967. 4. Dayton, D. H., Filer, L. J., and Canosa, C., reported at 53rd Annual Meeting, FASEB, Atlantic City, N.J., April 13-18, 1969. 5. Fish, I., and Winick, M.: Cellular growth in various regions of the developing rat brain. 6. Gluck, L., Talner, N. J., Stern, H., Gardner, T. H., and Kulovich, M. V.: Experimental cardiac hypertrophy; concentration of RNA in the ventricles. Science, 144:1244, 1964. 7. Logan, J. E., Mannell, W. A., and Rossiter, R.: Chemical studies of peripheral nerve during wallerian degeneration. Biochem. J., 52:482, 1952. 8. Moore, R. J., and Hamilton, T. H.: Estrogen-induced formation of uterine ribosomes. Proc. Nat. Acad. Sci., 52:439, 1964. 9. Smith, C. A.: Effects of maternal undernutrition upon the newborn infant in Holland (1944-45). J. Pediat., 30:229,1947. 10. Van Marthens, E., and Zamenhof, S.: Deoxyribonucleic acid of neonatal rat cerebrum increased by operative restriction of litter size. Exp. Neurol., 23:214, 1969. 11. Wiggleworth, J. S.: Experimental growth retardation in the fetal rat. J. Path. Bact., 88:1,1964. 12. Winick, M.: Cellular growth of human placenta. III. Intrauterine growth failure. J. Pediat., 71 :390-395, 1967. 13. Winick, M.: Diagnosis and Treatment of Fetal Disorders. New York, Springer-Verlag, 1969, pp. 83-101. 14. Winick, M., presented at 53rd Annual Meeting, FASEB, Atlantic City, N.]., April 13-18, 1969. 15. Winick, M., Fish, I., and Rosso, P.: Cellular recovery in rat tissues after a brief period of neonatal malnutrition. J. Nutrition., 95:623, 1968. 16. Winick, M., and Noble, A.: Quantitative changes in DNA, RNA, and protein during prenatal and postnatal growth in the rat. Devel. Biol., 12:451, 1965.
78
MYRON WINICK
17. Winick, M., and Noble, A.: Cellular response in rats during malnutrition at various ages. J. Nutrition, 89:300, 1966. 18. Winick, M., and Noble, A.: Quantitative changes in ribonucleic acids and protein during normal growth of rat placenta Nature, 212:34, 1966. 19. Winick, M., and Noble, A.: Cellular response with increased feeding in neonatal rats. J. Nutrition, 91 :179, 1967. 20. Winick, M., Noble, A., and Coscia, A.: Cellular growth in human placenta I. Normal placental growth. Pediatrics, 39:248, 1967. 21. Winick, M., and Rosso, P.: The effect of severe early malnutrition on cellular growth of human brain. Pediat. Res., 3:181,1969. 22. Zamenhof, S., Van Marthens, E., and Margolis, F. L.: DNA (cell number) and protein in neonatal brain: Alteration by maternal dietary protein restriction. SCience, 160:3825, 1968. New York Hospital 525 East 68th Street New York, New York 10021
Growth may be defined simply as an increase in the size of an animal or an individual organ. This increase in organ size is the result of a continuous accretion of protein and in some cases lipids. Hence the rate of net protein synthesis determines the rate of growth. When protein synthesis and protein degradation reach equilibrium, growth ceases. Since net protein synthesis is linear in all organs in utero and for a varying period postnatally, growth would appear to be a homogeneous process. However this is only one aspect of growth. The protein in any organ is packaged within cells. The same total organ protein may be contained in numerous small cells or in a few larger cells. The constant amount of DNA within the diploid nucleus of any cell in a particular species makes possible accurate determination of the total number of cells by chemical methods. 2 During normal growth, total organ DNA content or cell number increases linearly, then begins to decelerate, and finally reaches a maximum long before the organ size, as determined by net protein accretion, has reached its maximum. 16 As a consequence of this pattern of DNA synthesis, three phases of growth can be described: proportional increase in weight, protein, and DNA content (hyperplasia); an increase in DNA slower than the increase in protein and weight (hyperplasia and concomitant hypertrophy); and finally no further increase in DNA content, with net protein and weight continuing to increase at the same rate (hypertrophy). The growing period, then, is not a homogeneous process when viewed in cellular terms. It is possible that stimuli which retard or accelerate growth may result in different effects depending on whether they influence the rate of cell division (DNA synthesis) or whether they affect the subsequent increase in cell size. *Associate Professor of Pediatrics and Director, Division of Growth and Development, Cornell University Medical College, New York, New York; recipient of a Career Development Award (No. 1-K04·HD-18869-01) from the National Institute of Child Health and Human Development Reported in part at the Eighth Annual Meeting of the Pan American Health Organization Advisory Committee on Medical Research, Washington, June 9-13, 1969.
Pediatric Clinics of North America- Vol. 17, No.1, February, 1970
69
70
MYRON WINICK
In rats it has been shown that early malnutrition will curtail the rate of cell division if it occurs during the period of hyperplasia. In brain, this period is prior to weaning. Not only will neonatal undernutrition curtail the rate of cell division, but it will result in a permanent deficit in brain cell numberY Malnutrition after weaning will prevent the subsequent increase in c;ell size, and this effect can be reversed by later rehabilitation. During the period of rapid cellular proliferation, undernutrition will affect the rate of cell division in any brain area where cells are dividing and in any cell type in the process of division. 14 Figure 1 demonstrates that in animals malnourished from birth, cerebellum, where cells are most rapidly dividing, is affected earliest and most severely. Figure 2 focuses on the cell types involved. In cerebral cortex, neurons are not dividing postnatally and hence are not affected by undernutrition. The number of glia, however, is reduced. In cerebellum, all cell types are reduced. The areas under the lateral ventricle and under the third ventricle are other areas where primitive neuronal division is curtailed. Thus discrete brain areas are disproportionally affected by neonatal undernutrition. The magnitude of the effect is proportional to the rate of ceJ) division in a particular area. In contrast, increased nutrition will accelerate cell division and result in a permanent increase in brain cell number, if it occurs during the proliferati ve phase. 15 ,19 Thus within certain limits the number of cells ultimately present within the adult brain may be programmed during early life, during that DNA MALNUTRITION, PER CENT OF NORMAL
100
80
60
40
20
o
8
10 14 17 21 CEREBRUM
HI PPOCAMPUS
CEREBELLUM
Figure 1. Effect of early malnutrition on DNA content of various parts of rat brain.
71
CELLULAR GROWTH IN INTRAUTERINE MALNUTRITION
_NORMAL ~ MALNUTRITION
Cereb.ellum
Lateral ventricle ,.......-"-.
25
20
~
c;; 15 u
Cerebral cortex
"0 <1>
]i .!!! ~
10
period in postnatal life when brain cells are still dividing. Can this type of programming take place during prenatal life? Certainly all fetal organs are in the proliferative growth phase and therefore would appear susceptible to permanent cellular effects. However, the fetus is protected from the environment by the mother and the placenta. Thus, one may ask, can the mother or the placenta protect the growing fetus from such adverse environmental stimuli as maternal malnutrition, maternal disease, or insufficient uterine vascular supply? In animals it is possible to monitor the effects of various maternal stresses on the cellular growth of both placenta and fetal organs. In the human the fetus is not available for these types of organ analysis and therefore the placenta, which is accessible after birth, is the only organ that can be examined when the fetus is born in a viable condition. Thus if changes in the pattern of cellular growth in placenta may be correlated with cellular growth of fetal organs in animals, perhaps similar changes in human placenta will give us a clue to the effects of similar maternal stresses on the human fetus. . Normal placental growth, both in the rat and in the human, proceeds in the same way as previously described in the organs of the rat. In the rat placenta, DNA synthesis and hence cell division stops at 17 days of a 21-day gestation. I8 In the human placenta, cell division continues until around the 34th to 36th week of gestation. 20 In both cases net protein synthesis continues to term. Thus the same three phases of cellular growth may be described as in other organs of the rat. Stimuli imposed
72
MYRON WINICK
prior to the 17th day or 36th week, respectively, should lead to a permanent reduction in placental cell number, whereas stimuli that are active only after cell division has stopped should affect the size of individual cells but not the total number of cells. Hence examination of the ultimate cellular make-up of the placenta should give us an idea of when, during the course of growth, a stimulus has been active. With these principles in mind, let us examine the effects of a number of precisely timed experimental stimuli on placental and fetal growth in the rat. Then we shall examine the ultimate effects of naturally occurring stimuli on the human placenta. Finally we will attempt to determine the time during pregnancy when these stimuli are active. Maternal protein restriction in rats will retard both placental and fetal growth. In placenta, cell number (DNA content) was reduced by 13 days after conception, cell size (protein/DNA) remained normal, and the RNA/DNA ratio was markedly elevated. Retardation in fetal growth first became apparent at 15 days. After this there was a progressive decrease in cell number in all the organs studied. By term there were only about 85 per cent of the number of brain cells as in control animals (Table 1). These data agree with previous data of Zamenhof, which showed a similar reduction in total brain cell number in term fetuses whose mothers were exposed to a slightly different type of nutritional deprivation. 22 Thus the cellular changes produced by severe prenatal food restriction are reflected in the placenta even earlier than in the fetus, but retardation of cell division in all fetal organs including brain can be clearly demonstrated. By employing radioautography after injecting the mother with tritiated thymidine, cell division can be assessed in various discrete brain regions.! Differential regional sensitivity can be demonstrated in this way by the 16th day of gestation in the brains of fetuses of proteinrestricted mothers (Fig. 3). The cerebral white and gray matter are mildly affected. The area adjacent to the third ventricle and the subiculum are moderately affected, whereas the cerebellum and the area directly adjacent to the lateral ventricle are markedly affected. These data again demonstrate that the magnitude of the effect produced on cell division is directly related to the actual rate of cell division at the time the stimulus is applied. Moreover they demonstrate that the maternal-placental barrier in the rat is not effective in protecting the fetal brain from discrete cellular effects caused by maternal food restriction. The subsequent course of these animals born of protein-restricted mothers can be examined. Chow has reported that even if these animals are raised normally on foster mothers they demonstrate a permanent impairment in their ability to utilize nitrogen. 3 Data from our own laboratory demonstrate that if these animals are nursed on normal foster mothers in normal-sized litters, they will remain with a deficit in total brain cell number at weaning. Thus we can again see early programming of the ultimate number of brain cells. This program, moreover, is written in utero in response to maternal nutrition. These same newborn pups of protein-restricted mothers may be subjected to postnatal nutritional manipulation. If they are raised in
73
CELLULAR GROWTH IN INTRAUTERINE MALNUTRITION
Table 1. Composition of Various Tissues in Newborn Rats Subjected to Maternal Undernutrition PER CENT OF NORMAL CONTROL TISSUE
Weight
Protein
RNA
DNA
Whole animal Brain Heart Lung Liver Kidney
87 91 84 82 82 84
81 85 84 85 80 81
83 82 79 85 85 82
81 84 81 89 85 85
litters of three on normal foster mothers until weaning, the deficit in total number of brain ce1Is may be almost entirely reversed. Although quantitatively the number of cells approaches normal, qualitatively the deficit at birth might very well be made up by an increase in cell number in different areas from those most affected in utero. Thus although it may appear that optimally nourishing pups after exposing them to prenatal undernutrition will reverse the cellular effects, this may not actually be so in specific brain areas. Perhaps most nearly analogous to the situation in humans is exposing, these pups, malnourished in utero, to subsequent postnatal deprivation. One can raise these animals on foster mothers in groups of 18. Animals so reared show a marked reduction in brain cell number by
25
_
NORMAL
~
MATERNAL PROTEIN RESTR I CTI ON
20
5
o
Cerebrum (gray)
Cerebrum (white)
Third ventricle
Subiculum Cerebellum
Lateral ventricle
Figure 3. Effect of prolonged maternal protein restriction on various brain regions in 16-day rat embryo.
74
MYRON WINICK
weaning. This effect is much more pronounced than the effect of either prenatal or postnatal undernutrition alone. Animals subjected to prenatal malnutrition alone as previously described show a 15 per cent reduction in total brain cell number at birth. Those subjected only to postnatal malnutrition show a similar 15 to 20 per cent reduction in cell number at weaning. In contrast, these "doubly deprived" animals demonstrate a 60 per cent reduction in total brain cell number by weaning (Fig. 4). These data demonstrate that malnutrition applied constantly throughout the entire period of brain cell proliferation will result in a profound reduction in brain cell number, greater than the sum of effects produced during various parts of the proliferative phase. It would appear that the duration of malnutrition as well as the severity during this early critical period is extremely important in determining the ultimate cellular make-up of the brain. In rats it has been shown that clamping the uterine artery supplying one horn of the uterus will curtail fetal growthY The cellular events accompanying this fetal growth failure can be examined. Regardless of when the clamping is performed, cell division is curtailed in all fetal organs except brain (Table 2). In contrast, DNA content of placenta is reduced only if the clamping is done before the 17th day; after this time cell size (proteinIDNA) is reduced. The RNAIDNA ratio in placenta increases following uterine artery ligation at any time. 13 The exact significance of this elevated RNA/DNA ratio is not known. It has been noted under a variety of conditions involving tissue "stress." In cardiac hypertrophy secondary to experimental aortic ligation,6 in uterine hypertrophy induced by estrogen,S and in muscles exposed to repeated nerve stimulation/ one of the first changes noted is an increase in the RNA/ DNA ratio. 100
TOTAL BRA I N CELL NUMBER
80
~....
g
60
'0 C QJ
Figure 4. Comparison of the effect on brain cell number of prenatal malnutrition, postnatal malnutrition, and "combined" malnutrition.
..., ~
c...
40 20
o
Prenatal Postnatal Prenatal malnutrition malnutrition & postnatal (birth) (weaning) malnutrition (weaning)
75
CELLULAR GROWTH IN INTRAUTERINE MALNUTRITION
Table 2. Composition of Various Tissues in Newborn Rats After Clamping the Uterine Artery of the Mother at 17 Days of Gestation PER CENT OF NORMAL CONTROL TISSUE
Weight
Protein
Whole animal Brain Heart Lung Liver Kidney
67 91 84 62 62 64
95 84 65.· 70 61
71
RNA
DNA
63 103 79 55 75 82
99 91 59 55 75
71
Organs of animals delivered on the 21st day from ligated horns and then foster nursed to weaning do not attain the expected number of cells by weaning. Thus again a change produced in utero will persist throughout life. In these animals, however, brain was normal at birth and remained normal at weaning. Recently Zamenhof has demonstrated that artificially reducing the number of fetuses in utero will result in offspring with an increased number of brain cells. 10 Certainly then, in rats, manipulation of the maternal environment during gestation.will result in permanent cellular changes in the offspring. The effects of similar naturally occurring environmental stresses during human pregnancies are more difficult to demonstrate. However, certain clear changes can be noted. Placentas from infants with "intrauterine growth failure" show fewer cells and an increased RNA/DNA ratio when compared to controls.12 Fifty per cent of placentas from an indigent population in Chile showed similar findings. Placentas from a malnourished population in Guatemala had fewer cells than normal. 4 In a single case of anorexia nervosa in which a severely emaciated mother carried to term and gave birth to a 2500 gm. infant, the placenta contained less than 50 per cent of the expected number of cells (Fig. 5). Thus both vascular insufficiency and maternal malnutrition will curtail cell division in human placenta. The cellular make-up of the placenta in both of these situations strongly suggests that both stimuli have been active for some time prior to the 34th to 36th week of gestation. The effects of these stimuli on cellular growth of the fetus are more difficult to assess. In both situations fetal growth is retarded and birth weight reduced. 9 Indirect evidence would suggest that cell division in the human fetus might be retarded by maternal undernutrition. If one examines available data on infants who died after exposure to severe postnatal malnutrition, three separate patterns emerge. Breast-fed infants malnourished during the second year have a reduced protein/DNA ratio but a normal brain DNA content. Full-term infants who subsequently died of severe food deprivation during the first year of life had a 15 to 20 per cent reduction in total brain cell number. Infants weighing 2000 gm. or less at birth who subsequently died of severe undernutrition during the first year of life showed a 60
76
MYRON WINICK
....
80
Q)
.J::l
E
::::l
c:
60
Q;
u
'" E .... 0
z
40
'f/<.
20 0
IUGF
Indigent population Malnourished Anorexia in Santiago, Chile population nervosa in Guatemala·
( ) Number of cases .:' Data of Dayton et aI. Figure 5. Comparison of placental DN A content in various types of maternal deprivation.
100
MARASMUS (low bi rth weight)
MARASMUS (normal bi rth weight)
KWASHIORKOR
n:I
E .... o
c:
~
« z
50
a
o Figure 6. Comparison of brain cell number in marasmus, marasmus plus low birth weight, and kwashiorkor.
r
CELLULAR GROWTH IN INTRAUTERINE MALNUTRITION
77
per cent reduction in total brain cell number (Fig. 6).21 It is possible that these children were deprived in utero and represent a clinical counterpart of the "doubly deprived" animal. It is also possible that these were true premature infants and that the premature is much more susceptible to postnatal malnutrition than the full-term infant. Regional growth in the human brain is somewhat different than in the rat brain. Cell division stops at about the same time (8 to 10 months) in cerebellum, cerebrum, and brain stem. Severe malnutrition during this proliferative phase retards cell division in all three of these regions. The rate of cell division is about the same in cerebellum and cerebral cortex, and both are severely affected by malnutrition. In comparison to the rat,5 cell division in human cerebrum is much more rapid during postnatal growth, and human cerebrum is much more affected by postnatal malnutrition. Thus postnatal malnutrition curtails cell division in human brain as it does in rat brain. Prenatal stimuli affect human placenta in much the same way as rat placenta. Although at this time human fetal data are still sketchy, the suggestion that cell division in fetal organs can be curtailed by maternal undernutrition appears to be valid enough to require further studies. In view of the tremendous public health implications of this possibility and in view of the evidence in animals which demonstrates that these changes are permanent, it would seem to me that every effort should be made quickly to confirm or rule out the possibility that undernutrition of the mother may permanently reduce the number of brain cells in her offspring.
REFERENCES 1. Altman, J., and Das, G.: Autoradiographic and histological studies of postnatal neuronogenesis. I. A longitudinal investigation of the kinetics, migration and transformation of cells incorporating tritiated thymidine in infant rats with special reference to postnatal neurogenesis in some brain regions. J. Compo Neurol., 126:337, 1966. 2. Boivin, A., Vendrely, R., and Vendrely, C.: L'Acide desoxyribonucleique du noyau cellulaire, despositaire des caracteres hereditaires; arguments d'ordre analytique. Compt. Rend. Acad. Sci., 226:1061,1948. 3. Chow, B., presented at Gordon Research Conference on Nutrition, New London, New Hampshire, 1967. 4. Dayton, D. H., Filer, L. J., and Canosa, C., reported at 53rd Annual Meeting, FASEB, Atlantic City, N.J., April 13-18, 1969. 5. Fish, I., and Winick, M.: Cellular growth in various regions of the developing rat brain. 6. Gluck, L., Talner, N. J., Stern, H., Gardner, T. H., and Kulovich, M. V.: Experimental cardiac hypertrophy; concentration of RNA in the ventricles. Science, 144:1244, 1964. 7. Logan, J. E., Mannell, W. A., and Rossiter, R.: Chemical studies of peripheral nerve during wallerian degeneration. Biochem. J., 52:482, 1952. 8. Moore, R. J., and Hamilton, T. H.: Estrogen-induced formation of uterine ribosomes. Proc. Nat. Acad. Sci., 52:439, 1964. 9. Smith, C. A.: Effects of maternal undernutrition upon the newborn infant in Holland (1944-45). J. Pediat., 30:229,1947. 10. Van Marthens, E., and Zamenhof, S.: Deoxyribonucleic acid of neonatal rat cerebrum increased by operative restriction of litter size. Exp. Neurol., 23:214, 1969. 11. Wiggleworth, J. S.: Experimental growth retardation in the fetal rat. J. Path. Bact., 88:1,1964. 12. Winick, M.: Cellular growth of human placenta. III. Intrauterine growth failure. J. Pediat., 71 :390-395, 1967. 13. Winick, M.: Diagnosis and Treatment of Fetal Disorders. New York, Springer-Verlag, 1969, pp. 83-101. 14. Winick, M., presented at 53rd Annual Meeting, FASEB, Atlantic City, N.]., April 13-18, 1969. 15. Winick, M., Fish, I., and Rosso, P.: Cellular recovery in rat tissues after a brief period of neonatal malnutrition. J. Nutrition., 95:623, 1968. 16. Winick, M., and Noble, A.: Quantitative changes in DNA, RNA, and protein during prenatal and postnatal growth in the rat. Devel. Biol., 12:451, 1965.
78
MYRON WINICK
17. Winick, M., and Noble, A.: Cellular response in rats during malnutrition at various ages. J. Nutrition, 89:300, 1966. 18. Winick, M., and Noble, A.: Quantitative changes in ribonucleic acids and protein during normal growth of rat placenta Nature, 212:34, 1966. 19. Winick, M., and Noble, A.: Cellular response with increased feeding in neonatal rats. J. Nutrition, 91 :179, 1967. 20. Winick, M., Noble, A., and Coscia, A.: Cellular growth in human placenta I. Normal placental growth. Pediatrics, 39:248, 1967. 21. Winick, M., and Rosso, P.: The effect of severe early malnutrition on cellular growth of human brain. Pediat. Res., 3:181,1969. 22. Zamenhof, S., Van Marthens, E., and Margolis, F. L.: DNA (cell number) and protein in neonatal brain: Alteration by maternal dietary protein restriction. SCience, 160:3825, 1968. New York Hospital 525 East 68th Street New York, New York 10021