THE JOURNAL OF COMPARATIVE NEUROLOGY 310~356364(1991)

Effects of Prenatal Protein Deprivation on Postnatal Development of Granule Cells in the Fascia Dentata s. DIAZ-CINTRA,L. CINTRA, A. G A L V ~ A., AGUILAR, T. KEMPER, AND P.J. MORGANE Departamento de Fisiologia, Instituto de Investigaciones Biomedicas, UNAM, Ciudad Universitaria, Mexico 04510, D.F. (S.D.X., L.C., A.G., A.A.), and Boston City Hospital, Boston 02118 (T.K.) and Worcester Foundation for Experimental Biology, Shrewsbury 01545 (P.J.M.), Massachusetts

ABSTRACT The effect of prenatal protein deprivation on the postnatal development of granule cells in the fascia dentata in the rat was studied at 15, 30, 90, and 220 days of age. The granule cells showed a significant reduction in cell size, decreased number of synaptic spines throughout their dendritic extent, and reduced complexity of dendritic branching in the outer two-thirds of the molecular layer. All of these deficits were present at 15 days and persisted throughout the study (220 days). The least deficits in synaptic spine density occurred at 90 days and in dendritic branching at 30 days. Partial restitution of earlier, more severe deficits was associated primarily with maturational events occurring in the protein deprived rats, whereas later increases in deficits were related primarily to a failure of the protein deprived rats to keep pace with neuronal development occurring in the controls. The present results are similar to those noted in our previous study in this journal of the effect of a low protein diet (8% casein) on these neurons that extended from pregnancy until the time of sacrifice at 30,90, and 220 days of age (Cintra et al., '90; 532:271-277). Taken together, these two studies suggest that the postnatal adaptation of the granule cells to prenatal protein deprivation is primarily due to events that occur during pregnancy and that the site of predilection for the deficit is their dendrites in the outer two-thirds of the molecular layer of the fascia dentata. Key words: hippocampal formation, dentate gyrus, neuronal development, dendritic development, dendritic spines, malnutrition and hippocampal pathology

Several studies have noted a significant effect of nutritional deprivation on the fascia dentata. With undernutrition confined to the period of lactation, Fish and Winick ('69) found a decrease in the total DNA content of the hippocampal formation; Noback and Eisenman ('81) reported a decreased width of the dentate gyrus; and Katz and Davies ('82, '83) noted a decrease in two linear measurements and an area measurement of the hippocampal formation at the level of the habenula. All three measurements included part of the fascia dentata and adjacent hippocampal fields. With undernutrition extending from the period of lactation into the post-weaning period, Ahmed et al. ('87) noted effects on the granule cell to molecular layer synapse ratio and Corder0 et al. ('82) reported a curtailed development of glial cells in the molecular layer of the fascia dentata. Paula-Barbosa ('89), in a study of long-term, post-weaning undernutrition, reported a decreased width of the dentate gyrus and a decreased granule cell packing density at 6, 12, and 18 months. With combined pre- and O

1991 WILEY-LISS, INC.

postnatal nutritional deprivation, Lewis et al. ('79) noted, in an autoradiographic study from postnatal days 1 to 12, that the total cell cycle time was prolonged with the most marked effect on the DNA synthetic phase (S phase). The G1 phase (between mitosis and the synthetic phase) was strikingly shortened. Based on cell cycle time and the labeling index, they concluded that undernutrition markedly diminished the rate of granule cell acquisition. In this study they also measured the width of the granule cell layer and found it significantly decreased in the undernourished rats on postnatal day 12. With combined pre- and postnatal undernutrition, Jordan et al. ('82) reported a decreased density of granule cell in the dentate gyrus at 4 to 5 months of age. Katz et al. ('82) made the same linear measurement used by Katz and Davies ('82, '83) in rats undernourished Accepted May 3,1991. Address reprint requests to Dr. Thomas L. Kemper, Dept. of Neurology, Medical 9, Boston City Hospital, 818 Harrison Ave., Boston, MA 02118.

UNDERNUTRITION AND DENTATE GRANULE CELL MATURATION during lactation, gestation plus the period of lactation as well as the period of lactation plus the post-weaning period. In this study, the width was decreased only when gestation was included in the period of undernutrition. In our previous studies we have used an 8%casein diet initiated 5 weeks prior to conception and continued until the time of sacrifice at 30,90, and 220 days of age. With this paradigm we found that the percentage of the brain that is the hippocampal formation was unaffected by protein deprivation (West and Kemper, '76; Morgane et al., '78). In a recent study of rapid Golgi-impregnated granule cells, we have shown that all parameters of measurement are affected by the nutritional insult (Cintra et al., '90). When this study was compared to our previous rapid Golgi study in these same rats of layer 111111and V pyramidal cells in the visual cortex (Diaz-Cintra et al., '90) and the three types of neurons in the nucleus raphe dorsalis (Diaz-Cintra et al., '81b) and in the nucleus locus coeruleus (Diaz-Cintra et al., '84), it was noted that the granule cells of the fascia dentata were the neurons most affected by undernutrition. The present study was designed to determine the effect of protein deprivation confined to the period of gestation on the postnatal development of the granule cells. This time period was selected because of the evidence provided by Katz et al. ('82) on the importance of prenatal undernutrition on the postnatal development of the hippocampus. These neurons are also of interest in that they are the first link in the hippocampal trisynaptic circuit that has been implicated as a substrate for memory (Teyler and Discenna, '84; Barnes, '88; Eichenbaum and Cohen, '88; Lynch et al., "88). Consistent with this role is the demonstration of defects in spatial learning and memory as a result of hippocampal lesions (O'Keefe and Nadel, '78; Morris et al., '82; Brandeis et al., '89; Morris et al., '90). Changes consistent with involvement of the function of the trisynaptic circuit due to undernutrition include alterations in learning and some aspects of spatial memory (Laughlin et al., '84; Goodlett et al., '86) and in long-term potentiation and kindling (Austin et al., '86; Bronzino et al., '86, '90). An additional feature of the present study is that we explored the effects of prenatal protein deprivation on granule cells generated postnatally, i.e., after the period of the nutritional insult. We have selected for quantitative study neurons in the deep part of the granule cell layer, which are generated exclusively in the postnatal period (Bayer and Altman, '74; Crespo et al., '86). Only 15% of granule cell neurons are generated prior to birth (Bayer and Altman, '74). At the postnatal ages selected for this study, the earlier generated neurons are located in the more superficial part of the granule cell layer.

MATERIALS AND METHODS Twenty-four female and 8 male Sprague-Dawley derived rates were used for breeding. Female rats, weighing 250 to 300 grams on arrival, were housed 3 per colony cage, and males were housed in individual cages. They were maintained in automatically controlled animal rooms at a temperature of 22-24"C., a humidity between 40 and 50%, and 8 hours of light per day. Food and water were provided ad libitum. During the 5 weeks before breeding, 8 female rats were fed a 6%casein diet and 16 female rats a control 25% casein diet. Body weights were recorded every third day. After 5 weeks of adaptation to the diets, 1 male was introduced into each cage containing 3 females at 15:OO

357

hours. Vaginal smears were obtained each morning to detect when mating had occurred (Tonkiss et al., '88). In order to have a record of each female rat, 2 of the 3 were marked on either the left or right side with cresyl violet. One week prior to the expected parturition, all spermpositive females were placed in plastic nesting cages (48 x 27 x 16 cm) with adequate, clean nesting material. The percentage of neonatal deaths was 4% in both diet groups. Litters born the same day to 6%and 25% casein diet rats were weighed and sexed, and 4 males and 4 females were randomly cross-fostered to a 25% casein diet dam. This resulted in 8 litters each of cross-fostered 6% casein diet rats (6125%paradigm) and 25% casein diet rats (25125% paradigm). Body weights of each dam and their pups were taken on the day of delivery and each subsequent day until weaning. They were then weighed every third day until day 30. After weaning the rats were housed in groups of 4 same-sex colony cages. On days 15, 30, 90, and 220, 2 male pups were randomly selected from each litter from both the 6125% and 25125% diet groups for histological study, for a total of 8 rats per diet group at all 4 ages. The rats were anesthetized with pentobarbital, perfused through the heart with 10% neutral-buffered formalin, and the brains removed and weighed the following day. The methods of processing of the tissue and of selection of neurons are the same as those used in our previous study on the effects of concurrent protein deprivation on dentate granule cells (Cintra et al., '90). From each brain a 4-mmwide block, containing the hippocampal formation at the midpoint of its septotemporal extent, was removed and impregnated using the rapid Golgi method, following the modifications of Diaz-Cintra et al. ('81a). The blocks were embedded in low viscosity nitrocellulose and 10 sections serially cut in the frontal plane at a thickness of 90 and 120 pm and mounted in serial order. Each slide was assigned a random number to ensure that observations were blind with respect to diet and age. From each histological section one granule cell was selected at two-thirds the depth of the dorsal leaf of the granule cell layer. This precise specification of location of the perikaryon was necessary since the form of the dendritic tree has been shown to vary with the depth of the cell body in the granule cell layer (Green and Juraska, '85) and the density of synaptic spines was found to be different in neurons in the dorsal and ventral blade of the fascia dentata (Desmond and Levy, '85). Each cell was traced across adjacent sections with a 40X planapochromatic objective lens to obtain camera lucida drawings of the complete dendritic field. From these drawings measurements were made of the number of dendrites crossing concentric rings 36 pm apart according to the method of Sholl('56) (Fig. 1). Measurements of the number of synaptic spines and the diameter of dendrites were made with a lOOX planapochromatic objective lens on complete dendrites within the plane of section. Measurements of the number of synaptic spines and dendritic diameter were made on 50-pm segments of the proximal, middle, and distal dendrites from one complete dendrite parallel to the plane of section. The location of these segments is shown in Figure 1.For each dendritic segment, measurements were made with an ocular reticle, calibrated against a stage micrometer, of cell size, dendritic diameter, and length and diameter of 5 synaptic spines. The latter two measurements were used as correction factors to

S. DIAZ-CINTRA ET AL.

\

H

TABLE 1. Effect of Prenatal Protein Deprivation on the Postnatal Development of Granule Cells in the Dentate Gyrus at 15,30,90,and 220 Days of Age'

F values

Perikalya Major axis Minor axis Number of dendrites Proximal Middle Distal Dendritic intersections at: 36 pm 72 pm 108 pm 144 pm 180 pm 216 pm 252 pm 288 pm 324 pm 360 pm

1

2

3

5

4

6

7

0

9 10

Fig. 1. Camera lucida drawing of a granule cell neuron to show the location of the 10 concentric rings, each 36 km apart. The location of the sample areas for counting synaptic spines is shown on the left of the illustration. D = distal, M = middle, P = proximal.

Diet

Age

Interactions

(df 1,315)

(df3,315)

(df 3,315)

104.94** 18.30**

10.556*** 3.40 NS

37.52*** 3.21 NS

5.81* 16.85** 10.08**

18.93**

52.43*** 72.11*** 69.55*** 5.44 NS 4.96 NS 0.65 NS 16.98** 63.85*** 74.37*** 54.63*** 52.43*** 68.84*** 14.05***

2.05 NW 3.35 NS 3.09 NS 8.80**

62.16*** 118.10*** 176.03*** 121.71*** 39.76*** 3.18 NS

10.16**

5.09* 1.72 NS 1.06 NS 1.26 NS 3.96* 8.86**

10.15*** 7.82** 7.00**

9.94** 1.65 NS

1. Two-way analysis of variance (ANOVA) with nutritional status and age as t h e two factors df = degrees of freedom. *p < 0.05 **p < 0.01 ***p < 0.001.

Statistical Package for Social Sciences (SPSS) version 3.0 (Norusis, '88). Analysis of neuronal measurements were made with 2 between-factors (diet and age) and 5 neurons per subject (N = 8). Each cell was considered different, since the relative variation between animals has been found to be smaller that the variation between neurons (Leuba et al., '89). Significant age effects were further examined with additional comparisons between 15, 30,90, and 220 days of age. Probability values of these comparisons were further adjusted using the Bonferroni method (Ryan, '59).

calculate the total number of dendritic spines (Feldman and Peters, '79; Stirling and Bliss, '81). With these methods the following measurements were made: (1)major and minor axes of the cell body, (2) synaptic spine density per 50-km segment on the proximal part of the dendrite (beginning 10 km from the cell body), on the middle segment, and on the distal dendritic segment (to within 10 pm of the tip), and (3) number of dendritic intersections crossing each of the concentric rings (Fig. 1). RESULTS The mean ? SEM of each of these measurements are graphically shown in Figures 2-4. The results of the ANOVA are shown in Table 1, the Statistical analysis was conducted by using two-way statistical analysis of age-related changes in Table 2, and analysis of variance (ANOVA) from the program in the mean +/- SEM of the various measurements in Figures TABLE 2. Age-Related Changes

F values 15 days vs ~~

30 days

90 days

220 days

90 days

220 days

90 days vs 220 days

(df 1,151)

(df 1,151)

(df 1,151)

(df1,151)

(df1,151)

(df 1,151)

1.20 NS

30.41***

13.36 * * *

17.71***

~

Perikarya Major axis Minor axis Number of dendrites Proximal Middle Distal Dendritic intersections a t 36 pm 72 pm 108 p m 144 p m 180 p m 216 pm 252 pm 288 pm 324 p m 360 um

30 days vs

-2

-

3.27 NS 20.72*** 12.20**

1.72 NS 36.95* * * 19.90***

-

-

19.65***

10.53*** 154.77*** 319.82*** 524.33*** 364.51*** 96.15*** 9.09**

81.56*** 162.20***

226.20*** 205.15*** 40.33*** 0.00 NS

-

-

15.91** 41.40*** 23.43***

0.29 NS 1.08 NS 1.39 NS

-

-

15.66*** 101.57*** 206.33*** 312.71***

819.02*** 89.65*** 4.94'

1.Additional comparisons between ages (n = 16 animals in each group). 2. When no significant age effect was noted on the ANOVA (Table I),comparisons between t h e ages were not made df *p < 0.05 **p < 0.01 ***p < 0.001,

1.60 NS 16.04*** 50.31**" 130.50*** 103.50*** 37.51*** 9.00** =

degrees of freedom.

6.09'

-

2.46 NS -

5.38* 2.34 NS 1.75

7.95** 1.15 NS 0.06 NS

-

-

0.17 NS 4.36* 23.06*** 57.00*** 60.33*** 32.51*** 4.94*

0.96 NS 2.74 NS 1.66 NS 5.26* 2.89 NS 0.59 NS 0.89 NS

-

-

UNDERNUTRITION AND DENTATE GRANULE CELL MATURATION 20

-

MAJOR AXIS

1 8 -

$1 6

g 0

0

-

0

I

MINOR AXIS

--- -- -_

1 8 -

-I

2

:

s14

359

:

0

--

I 1 4

1 2 -

-

1 2 -

**

1 10

#

0

0

0

- 5

1 0

Fig. 2. Major and minor axis of the cell body +/- SEM. The data for the control rats is shown as a solid line and for the prenatally protein-deprived rats as an interrupted line.

1

PROXIMAL SEGMENT

loo

MIDDLE SEGMENT 9 0

$8 0

.

4 0



4 0

153 0

90 AGE IN DAYS

220

1

90

220

DISTAL SEGMENT

-----0



1530

AGE IN DAYS

loo

4 0



1530

--T

d

0

so

220

AGE IN DAYS

Fig. 3. Synaptic spine density on granule cell dendrites +/- SEM. The data for the control rats is shown as a solid line and for the prenatally protein-deprived rats as an interrupted line.

2-4. There were no significant diet effects on brain or body weight. The major axis of the perikaryon showed significant diet and age effects and a significant interaction between diet and age (Table 1).There was a decrease in this measurement in the 6/25%rats at all 4 ages that varied from - 10 to - 18%.Significant age-related changes occurred between 15 and 90 days, 15 and 220 days, 30 and 90 days, and 30 and 220 days (Table 2). The major axis of the perikaryon increased in both diet groups from 15 to 90 days, then decreased between 90 and 220 days in the 6125% rats with little change in this axis in the 25/25% rats (Fig. 2). The

minor axis of the perikaryon showed only a significant diet effect (Table 2), with a decrease in this axis in the 6/25% rats at all ages that varied from -6 to - 15%. Synaptic spines on proximal, middle, and distal dendritic segments all showed significant diet and age effects and a significant interaction between diet and age (Table 1).The number of synaptic spines was decreased in the 6/25% rats at all four ages. On the proximal segments this varied from -5 to -20%, in the middle segments from -5 to -23%, and on the distal segment from -4 to -20%. This decrease was least evident for all 3 segments at 90 days (Fig. 3). Significant age-related changes in proximal dendrites were

S. DIAZ-CINTRA ET AL.

360

A

RING 2

157

RING 1

z W

P

'1 0 4

o i 1530

157

$ 1

9 0 AGE I N DAYS

1530

220

220

90 AGE IN DAYS

1 5 7

RING 3

RING 4

P c

z

2 W 0

W

0

1530

1 5 7

90 AGE IN DAYS

157

a

RING 6

a

W

51

W

I I

t

0

220

In z

0

W

90 AGE IN DAYS

RING 5

2

2

1530

220

u)

f

5i o i

0 '

I-

I

I z W 0

'1

I

I

f

0 1

0'

1530

90

220

AGE IN DAYS

153 0

90

220

AGE IN DAYS

Fig. 4. Number of dendritic intersections for each of the 10 concentric rings +/- SEM shown in Figure 1. The data for the control rats is shown as a solid line and for the prenatally protein-deprived rats as an interrupted line.

noted between 15 and 220 days, 30 and 220 days, and 90 and 220 days (Table 2). This was associated with an increased number of spines between 90 and 220 days in the 25125% rats, whereas the 6125% rats showed little change between these ages (Fig. 3). The middle and distal dendritic segments showed significant age-related changes between 15 and 30 days, 15 and 90 days, and 15 and 220 days. This

was associated with an increased number of spines between these ages in the 6125%rats but not the 25125%rats (Fig. 3). The number of dendritic intersections showed significant diet and age effects and interaction between diet and age for the fourth (144 pm) to the ninth (324 pm) ring and a significant diet effect for the tenth ring (360 pm) (Table 1). In these 7 rings, i.e., rings 4 through 10, the number of

UNDERNUTRITION AND DENTATE GRANULE CELL MATURATION

4

In

1

RING 8 In

iloi

a W c

t-

E

2

W

a

361

'1

E

I

o j r 1 53 0

9 0

AGE IN DAYS

220

1530

1 5 7

90 AGE IN DAYS

220

RING 1 0

RING 9

$1 0

a

W

c

E

1530

9 0 AGE IN DAYS

1530

220

9 0 AGE IN DAYS

220

Figure 4 B

dendrites was decreased at all 4 ages in the 6125% rats. In ring 4 (144 pm) it varied from -1 to -12%, in ring 5 (280 km) from -6 to -36%, in ring 6 (216 km) from -19 to -60%, in ring 7 (151 pm) from - 17 to -94%, in ring 8 (288 pm) from -26 to -63%, in ring 9 (324 pm) from -15 to -91%, and in ring 10 (360 pm) from -15 to -25%. This deficit was least in rings 4 through 8 at 30 days and in rings 9 and 10 at 220 days. Significant age-related changes in rings 4 through 9 were present between 15 and 30 days and 90 and 220 days, and in ring 10 between 15 and 90 and 15 and 220 days (Table 2). In both diet groups this was due to an increased number of dendritic intersections between 15 days and the older ages (Fig. 4). Between 30 and 90 days and 30 and 220 days, there were significant age-related changes in rings 5 through 10 (Table 2). In the 25125% rats, this was associated with an increased number of dendritic intersections between 30 and 90 days and 30 and 220 days for all 4 rings. In the 6125% rats, the number of dendritic intersections in ring 5 was similar to those at 30, 90, and 220 days; in rings 6 through 10, there was an increased number of intersections between 30 and 90 days that was less marked than that noted in the 25125% rats. There was little difference in their number when 30 and 220 days were compared (Fig. 4). Between 90 and 220 days, significant age-related changes were noted only for ring 7. During this time the 25125% rats showed little change in the number of intersections in ring 7, whereas the 6125% rats showed a decreased number.

DISCUSSION These studies show that the granule cells of the fascia dentata in the prenatally protein deprived rats exhibit a deficit at all 4 ages in all measurements of cell size, synaptic spine density, and complexity of dendritic branching of the distal 7 of the 10 concentric rings, with the most marked deficits noted in distal dendritic branching. The least deficit in synaptic spines occurred at 90 days and in number of dendritic intersections at 30 days in 5 of the 7 rings that showed a significant diet effect. These partial restitutions of an earlier, more severe dietary effect were mainly associated with maturational events occurring in the undernourished rats. Between 15 and 90 days, there was a significant increase in synaptic spines on middle and proximal dendrites that was noted only in the prenatally protein deprived rats. Between 15 and 30 days, the undernourished rats showed a greater increase in number of dendritic intersections than that found in the controls. The later increase in the deficits was associated with a failure of the protein-deprived rats to keep pace with maturational events seen in the control animals. This can be seen in the data for the proximal dendritic spines. At this location the control rats, but not the undernourished rats, showed an increased number of spines from 90 to 220 days. The widening deficit in the number of dendritic intersections in the protein deprived rats at 90 and 220 days was associated with significant age-related increases in the number of

362

dendritic intersections in the control rats with a less marked increase, little or no change, or actual decrease in their number during this period in the protein-deprived rats. The latter finding was noted only for the outer 6 dendritic rings. Very few dendrites reached the tenth concentric ring in both diet groups (Fig. 4). The remaining 9 concentric rings can be conveniently divided into thirds, each containing 3 rings. These 3 zones correspond to our sample areas for the number of synaptic spines on proximal, middle, and distal dendritic segments. Synaptic spine deficits showed no marked predilection for any of the 3 sample areas. These effects of prenatal protein deprivation are similar to those noted in our prior study of the effect of an 8% casein diet administered both pre- and postnatally on rapid Golgi-impregnated granule cells in the fascia dentata (Cintra et al., '90). In both dietary paradigms, the female rats were adapted to the low protein diets 5 weeks prior to pregnancy and maintained on the respective diets throughout the pregnancy. In the present study the protein deprivation was more severe (6% casein diet) and confined only to the prenatal period. In our prior study protein deprivation was less severe (8%casein diet) and was continued throughout pregnancy and postnatally until the time of sacrifice at 30,90, and 220 days of age. In both paradigms the granule cells in the protein-deprived rats showed a significant decrease in the major and minor axes of the cell body, a significant decrease in density of dendritic spines, and a significant decrease in complexity of dendritic branching. In both dietary paradigms, the latter effect was most marked on distal dendritic segments. There were differences in the details of the effect. With the concomitant 8%casein diet, there were significant dietary effects on the complexity of the dendritic tree on all but one of the concentric rings but no significant diet effect on the number of synaptic spines on the proximal dendritic segment. In the present paradigm of prenatal undernutrition, there was a significant dietary effect on synaptic spines on all dendritic segments and no significant diet effect on the complexity of the dendritic tree in the inner three rings. Thus in the prenatally proteindeprived rats there was a more widespread effect on synaptic spine density and, in the concomitantly undernourished 8% protein diet rats, a more widespread effect on the complexity of dendritic branching. When the data on the effect of the diets on the number of dendritic intersections and on synaptic spine density on these dendrites are combined, the deficit in total number of synaptic spines on individual granule cells is most marked in both dietary groups in a zone corresponding to the outer two-thirds of the molecular layer. This is the site of termination of the first limb of the hippocampal trisynaptic circuit, a circuit widely implicated as a substrate for learning and memory. This initial limb of the trisynaptic circuit, the perforant pathway, arises from neurons in the entorhinal cortex. To date, the effect of undernutrition on these neurons remains unknown. In agreement with the present study, electrophysiological studies have shown a prominent effect of undernutrition on the function of the hippocampal formation. With the same dietary paradigm used in the current study, Austin et al. ('86) have reported a diminished ability to potentiate the synaptic component of long-term potentiation, and Bronzino et al. ('90) showed alterations in the electrographic and behavioral correlates of induced seizure activity. In the latter study they noted a lowered seizure threshold, longer

S. DIAZ-CINTRA ET AL. after-discharge durations, and inability or delay in obtaining complete kindling. In a prior study, these investigators (Bronzinoet al., '86) reported similar findings on the effects of pre- and postnatal protein deprivation on seizure activity. Both dietary groups showed evidence of curtailed granule cell development, with deficits present at the earliest sample time persisting throughout the period of observation. This accounted for the deficits in cell size and synaptic spine density. Another dietary effect noted in both dietary paradigms was for age-related changes noted in the control rats to be either curtailed or failing to occur in the proteindeprived rats. This latter effect was particularly evident for the deficits in the complexity of distal dendritic branching at 90 and 220 days. These similarities of protein deprivation effects in these 2 dietary paradigms suggest that the postnatal adaptation of the granule cell to nutritional deprivation is primarily due to events that occur during pregnancy. In the present study the dentate cells specifically selected for quantitative study were generated postnatally, after the period of nutritional deprivation, indicating that this nutritional effect on these neurons is not a direct effect of nutritional deprivation on the neurons themselves. The mechanism for this prenatally determined effect on the postnatal development of these neurons is unknown. Taken together, the present study, that of Cintra et al. ('go), and the electrophysiological studies of Austin et al. ('86) and Bronzino et al. ('86, '90) noted above, call attention to the important role of prenatal undernutrition on the postnatal development of the brain and emphasize the need for further research on the effects of nutritional deprivation on the prenatal stages of brain development. The study that most closely approximates the present one is that of Ahmed et al. ('87). These investigators determined the number of molecular layer synapses per granule cell in the fascia dentata of rats undernourished from embryonic day 18 until the time of sacrifice on postnatal days 21,75, and 150. This ratio was unchanged at postnatal day 21, decreased on postnatal day 75, and increased on postnatal day 150. It can be seen in our Figures 3 and 4 that the number of synapses per neuron would most likely be decreased at all ages. At all 4 ages all significant findings indicated deficits in these measurements. These results of Ahmed et al. ('87) are, therefore, different from those found in the present study, indicating that the effect of prenatal undernutrition may be different than that found in predominantly postnatal undernutrition. A similar conclusion can be made from the study of Katz et al. ('821, who, using a variety of different periods of pre- and postnatal undernutrition, noted a decrease in width of the hippocampal formation only in rats in which undernutrition included the period of gestation. These linear measurements were made at the level of the habenula and included parts of the fascia dentata and adjacent hippocampal fields. One feature of the Ahmed et al. ('87) study that closely mirrored the findings of the present study was that the late postnatal age-related changes were different in undernourished and control rats. In control rats the synapse to granule cell ratio increased from postnatal days 2 1 to 75 and then decreased to postnatal day 150. During this time the experimental rats showed a steady increase in this ratio from postnatal days 21 to 150. Two other neuronal populations with prolonged periods of postnatal generation, the granule cells in the cerebellum

UNDERNUTRITION AND DENTATE GRANULE CELL MATURATION and olfactory bulb, have also been shown to be markedly affected by undernutrition. These microneuronal populations are of interest as Altman ('86, '87) has shown that a hypoplasia of these neurons in the cerebellum and hippocampal formation leads to hyperactive behavior and in the hippocampal formation also to learning disabilities. The cerebellar granule cells, which are normally generated up to the time of weaning, show a slight prolongation of their time of generation in undernourished rats (Barnes and Altman, '73a; Sima and Persson, '75). The granule cells of the olfactory bulb and the fascia dentata have no finite time at which their germinal zones stop producing neurons (Altman, '69; Kaplan and Hines, '77). Two studies on the effect of postnatal undernutrition on the DNA content in the hippocampal formation and cerebellum have indicated an earlier deficit in cell number in the cerebellum than in the hippocampal formation (Fish and Winick, '69; Tonkiss et al., '88). Fish and Winick ('69) found the earliest and most striking deficit in DNA content in the cerebellum on postnatal day 6, followed by deficits in DNA in the cerebrum on postnatal day 14, and in the hippocampus on postnatal day 17. In the brainstem there was no deficit in DNA content in the undernourished rats. Tonkiss et al., ('88), in 2 different models of postnatal undernutrition, determined the DNA content of the cerebellum and hippocampal formation on postnatal day 10 and found a deficit in total DNA only in the cerebellum. In agreement with this, they found a significant deficit in the weight of the cerebellum but not the hippocampal formation in these experimental rats. In the cerebellum this deficit appears to be predominately due to decreased numbers of granule cells (Dobbing et al., '71; Barnes and Altman, '73a,b; Clos et al., '77) with no apparent effect on the number of Purkinje cells (Clos et al., '77). Comparable studies are not available for the fascia dentata. In the same model of pre- and postnatal undernutrition used by Cintra et al. ('go), we measured the volume of the olfactory bulb and of its internal granule cell layer at 10,30,and 90 days (Resnicket al., '79). At all 3 ages it was significantly smaller in the protein-deprived rats. When the data were expressed as the percentage of the brain comprising the olfactory bulb and its internal granule cell layer, the only significant finding was a transient deficit in its relative size on postnatal day 10. Comparable data are available on the effect of predominantly postnatal undernutrition on the density of granule cells in the fascia dentata and in the cerebellum. Warren and Bedi ('88) noted that the density of cerebellar granule cells progressively decreased from postnatal day 21 to 75 to 150 in both control and experimental rats. The undernourished rats showed a significantly increased density of granule cells only on postnatal day 21. In the fascia dentata, Ahmed et al. ('87) noted a decreasing density of granule cell neurons in control and experimental rats from postnatal days 21 to 75, then from postnatal days 75 to 150 no change in experimental rats, and an increasing density in the control rats. This resulted in a significant increase in the density of granule cell neurons on postnatal day 21 and a decrease on postnatal day 150. They suggest that this latter decreased density was due to a net loss of neurons in the undernourished rats. This decreased density of granule cells in the fascia dentata of older, undernourished rats has been reported in 2 other studies. Paula-Barbosa et al. ('89) noted it in 6-, 12-, and 16-month-old rats that were nutritionally deprived from the eighth postnatal week until the time of sacrifice. They also found a decreased density of

363

granule cell neurons in the cerebellum at the 2 older ages. Jordan et al. ('82) noted a decreased density of fascia dentata granule cells in rats 4 to 5 months of age that were undernourished during the period of gestation and lactation. Thus the regional DNA studies cited above indicate early deficits in the total number of cells in the cerebellum and hippocampal formation as a result of nutritional deprivation. There is then evidence at older ages for a decreasing density of neurons in cerebellum and fascia dentata, suggesting a later loss of these neurons in the undernourished rats. In agreement with this apparent regressive change in older undernourished rats, it can be seen in our Figures 2-4 that the deficits in morphometric measurement are correspondingly more severe at the older ages. In the dietary paradigm used for our study of the effect of pre- and postnatal protein deprivation on the granule cells of the fascia dentata (Cintra et al., '901, we studied postnatal day 30 effects of protein deprivation on the dendritic tree of cerebellar granule cells (West and Kemper, '76). There was no effect on number of dendrites, whereas their length was decreased by 19%. In this same paradigm we found on postnatal day 90 in Golgi-impregnated olfactory bulb granule cells a deficit in the number of primary apical dendritic branches and in the number of primary basal dendrites. All other measurements of number of dendritic processes, measurements of dendritic length and cell size showed no significant difference from the controls (Resnick et al., '79). In the present study and in our prior study of the granule cells of the fascia dentata (Cintra et al., 'go), all morphometric parameters showed deficits in the protein-deprived rats. Thus the available evidence from the present study and our prior study (Cintra et al., '90) of the granule cells of the fascia dentata indicates that in both prenatal and combined pre- and postnatal protein deprivation there is a more striking effect on postnatal development of dentate granule cells than that noted in our prior studies of cerebellar and olfactory bulb granule cells.

ACKNOWLEDGMENTS This research was supported by NIH grants HD22539-04 and HD-23338,NICHHD.

LITERATURE CITED Ahmed,M.G.E., K.S. Bedi, M.A. Warren, andM.M. Kame1 (1987) Effectsofa lengthy period of undernutrition from birth and subsequent nutritional rehabilitation on the synapse: Granule cell neuron ratio in the rat dentate gyrus. J.Comp. Neurol. 263:146-158. Altman, J.(1987) Morphological and behavioral markers of environmentally induced retardation of brain development: An animal model. Environ. Health Perspectives 74:153-168. Altman, J. (1986) An animal model of minimal brain dysfunction. In M. Lewis (ed): Learning Disabilities and Prenatal Risk. Urbana: Univ. Illinois Press, pp. 241-304. Altman, J. (1969) Autoradiographic and histological studies of postnatal neurogenesis. IV. Cell proliferation and migration in the anterior forebrain, with special reference to persisting neurogenesis in the olfactory bulb. J. Comp. Neurol. 137:433458. Austin, K.B., J.D. Bronzino, and P.J. Morgane (1986) Prenatal protein malnutrition affects synaptic potentiation in the dentate p u s of rats in adulthood. Dev. Brain Res. 29:267-273. Barnes, C.A. (1988) Spatial learning and memory processes: The search for their neurobiologicalmechanisms in the rats. Trends Neurosci. 1I:163169. Barnes, D., and J. Altman (1973a) Effects of different schedules of early undernutrition on the preweaning growth of the rat cerebellum. Expl. Neurol. 38:406-419.

364 Barnes, D., and J. Altman (1973b) Effects of two levels of gestationallactational undernutrition on the postweaning growth of the rat cerebellum. Expl. Neurol. 38:420428. Bayer, S.A., and J. Altman (1974) Hippocampal development in the rat cytogenesis and morphogenesis examined with autoradiography and low level x-irradiation. J. Comp. Neurol. 158:55-80. Brandeis, R., Y. Brandys, and S. Yehuda (1989) The use of the Morris water maze in the study of memory and learning. Int. J. Neurosci. 48:29-69. Bronzino, J.D., R.J. Austin-LaFrance, C.J. Siok, and P.J. Morgane (1986) Effects of protein malnutrition on hippocampal kindling: Electrographic and behavioral measures. Brain Res. 384t348-354. Bronzino, J.D., J. Robert, A. LaFrance, and P.J. Morgane (1990) Effects of prenatal protein malnutrition on perforant path kindling in the rat. Brain Res. 515t45-50. Cintra, L., S. Diaz-Cintra, A. Galvan, T.L. Kemper, and P.J. Morgane (1990) Effect of protein undernutrition on the dentate gyrus in the rat of three age groups. Brain Res. 532:271-277. Clos, J., M. Favre, M. Selme-Matrat, and J. Legrand (1977) Effects of undernutrition on cell formation in the rat brain and specially on cellular composition of the cerebellum. Brain Res. 123:13-26. Cordero, M.E., A. Zvaighft, S. Muzzo, and 0. Brunser (1982) Histological maturation of astroglial cells in the archicortex of early malnourished rats. Pediatr. Res. 16:187-191. Crespo, D., B.B. Stanfield, and W.M. Cowan (1986) Evidence that lategenerated granule cells do not simply replace earlier formed neurons in the rat dentate gyrus. Exp. Brain Res. 62541-548. Desmond, N.L., and W.B. Levy (1985) Granule cell dendritic spine density in the rat hippocampus varies with spine shape and location. Neurosci. Lett. 54.3 19-224. Dib-Cintra, S., L. Cintra, T.L. Kemper, 0. Resnick, and P.J. Morgane (1981a) Nucleus raphe dorsalis: a morphometric Golgi study in rats of three age groups. Brain Res. 207:l-16. Dib-Cintra, S., L. Cintra, T.L. Kemper, 0. Resnick, and P.J. Morgane (1981b) The effect ofprotein deprivation on the nucleus raphe dorsalis: A morphometric Golgi study in rats of three age groups. Brain Res. 211:243-255. DiB-Cintra, S., L. Cintra, T.L. Kemper, 0. Resnick, and P.J. Morgane (1984) The effects of protein deprivation on the locus coeruleus: A morphometric Golgi study in rats of three age groups. Brain Res. 304:243-253. DiB-Cintra, S., L. Cintra, A. Ortega, T.L. Kemper, and P.J. Morgane (1990) Effects of protein deprivation on pyramidal cells of the visual cortex in rats of three age groups. J. Comp. Neurol. 292:117-126. Dobbing, J., J.W. Hopewell, and A. Lynch (1971) Vulnerability of developing brain. VII. Permanent deficit of neurons in cerebral and cerebellar cortex followingearly mild undernutrition. Expl. Neurol. 3 2 ~ 4 3 9 4 4 7 . Eichenbaum, H., and N.J. Cohen (1988) Representation in the hippocampus: What do hippocampal neurons code? Trends Neurosci. 11244-248. Feldman, M.L., and A. Peters (1979) A technique for estimating total spine numbers on Golgi-impregnated dendrites. J. Comp. Neurol. 188t527542. Fish, I., and M. Winick (1969) Effect of malnutrition on regional growth of the developing rat brain. Exp. Neurol. 25t534-540. Goodlett, C.R., M.L. Valentino, P.J. Morgane, and 0. Resnick (1986) Spacial cue utilization in chronically malnourished rats: Task-specific learning deficits. Dev. Psychobiol. 19:l-15. Green, E.J., and J.M. Juraska (1985) The dendritic morphology of hippocampal dentate granule cells varies with their position in the granule cell layer: A quantitative study. Exp. Brain Res. 59:582-586. Jordan, T.C., K.F. Howell, N. McNaughton, and P.L. Heatlie (1982) Effects of early undernutrition on hippocampal development and function. Res. Exp. Med. (Berlin) 180:201-207. Kaplan, M.S., and J.W. Hines (1977) Neurogenesis in the adult rat: Electron microscopicanalysis of light radioautographs. Science 197r1092-1094. Katz, H.B., and C.A. Davies (1982) The effects of early-life undernutrition

S. DIAZ-CINTRA ET AL. and subsequent environment on morphological parameters of the rat brain. Brain Res. 5:53-64. Katz, H.B., and C.A. Davies (1983) The separate and combined effects of early undernutrition and environmental complexity at different ages on cerebral measures in rats. Dev. Psychobiol. 16:47-58. Katz, H.B., C.A. Davies, and J. Dobbing (1982) Effects of undernutrition at different ages early in life and later environmental complexity on parameters of the cerebrum and hippocampus in rats. J. Nutr. 112:13621368. Laughlin, N.K., F. Stanley, and J. Bell (1984) Earlyundernutrition andlater hippocampal damage: effects of spontaneous behaviors and reversal learning. Physiol. Psychol. 11:268-277. Leuba, G., N. Jeanpetre, R. Krafsik, and J.M. Fritschy (1989) Sample size and statistical power in the hierarchical analysis of variance applications in morphometry of the nervous system. J. Neurosci. Meth. 29:231-239. Lewis, P.D., A.J. Patel, and R. Balazs (1979) Effect of undernutrition on cell generation in the rat hippocampus. Brain Res. 168r186-189. Lynch, G., D. Muller, P. Seubert, and L. Larson (1988) Long-term potentiation: persisting problems and recent results. Brain Res. Bull. 21:363372. Morgane, P.J., M. Miller, T.L. Kemper, W. Stern, W. Forbes, R. Hall, J. Bronzino, J. Kissane, E. Hawrylewicz, and 0. Resnick (1978) The effects of protein malnutrition on the developing nervous system in the rat. Neurosci. Biobehav. Rev. 2t137-230. Morris, R.G.M., P. Garrad, J.N.P. Rawlins, and J. O’Keefe (1982) Place navigation impaired in rats with hippocampal lesions. Nature 297t681683. Morris, R.G.M., F. Schenk, F. Tweedie, and L.E. Jarrard (1990) Ibotenate lesions of hippocampus and/or subiculum: dissociating components of allocentric spatial learning. European J. Neurosci. 2:1016-1028. Noback, C.R., and L.M. Eisenman (1981) Some effects of protein-calorie undernutrition on the developing central nervous system of the rat. Anat. Rec. 201:67-73. Norusis, M.J. (1988) SSPS Update Manual and Statistical Guide. O’Keefe, J., and L. Nadel (1978) The Hippocampus as a Cognitive Map. Oxford: Oxford Univ. Press. Paula-Barbosa, M.M., J.P. Andrade, J.L. Castedo, F.P. Azevedo, I. Comoes, B. Vogt, and M.A. Tavares (1989) Cell loss in the cerebellum and hippocampal formation in adult rats after long-term low-protein diet. Exp. Neurol. 103:186-193. Resnick, O., M. Miller, W. Forbes, R. Hall, T. Kemper, J. Bronzino, and P.J. Morgane (1979) Developmental protein malnutrition: influences on the central nervous system of the rat. Neurosci. Biohehav. Rev. 3:233-246. Ryan, T.A. (1959) Multiple comparisons in psychology research. Psychol. Bull. 56%-47. Sholl, D.A. (1956) The Organization of the Neurons in the Cerebral Cortex. NewYork Hafner. Sima, A., and L. Persson (1975) The effect of pre- and postnatal undernutrition in the development of the rat cerebellar cortex. Neurobiol. 523-34. Stirling, R.V., and T.V.P. Bliss (1978) Observations on the commissural projection of the dentate gyrus in the reeler mutant mouse. Brain Res. 150:447465. Teyler, T.J., and P. Discenna (1984) The topographical anatomy of the hippocampus: a clue to its function. Brain Res. Bull. 12311-719. Tonkiss, J., C.A. Cohen, and S.B. Sparber (1988) Different methods for the producing neonatal undernutrition in rats cause different brain changes in the face of equivalent somatic growth parameters. Dev. Neurosci. 10:141-151. Warren, M.A., and K.S. Bedi (1988) The effects of a lengthy period of undernutrition from birth and subsequent nutritional rehabilitation on the granule-to-Purkinje cell ratio in the rat cerebellum. J.Anat. 159:147150. West, C.D., and T.L. Kemper (1976) The effect o f a low protein diet on the anatomical development of the rat brain. Brain Res. 107:221-237.

Effects of prenatal protein deprivation on postnatal development of granule cells in the fascia dentata.

The effect of prenatal protein deprivation on the postnatal development of granule cells in the fascia dentata in the rat was studied at 15, 30, 90, a...
937KB Sizes 0 Downloads 0 Views