T m JOURNAL OF COMPARATrVE NEUROLOGY 292~117-126 (1990)

Effects of Protein Deprivation on Pyramidal Cells of the Visual Cortex in Rats of Three Age Groups S. DiAZ-CMTRA, L. CINTRA, A. ORTEGA, T. KEMPEK, AND P.J. MORGANE Departamento de Fisiologia, Instituto de Investigaciones Biomedicas, UNAM, Ciudad Universitaria, Mexico 04510, D.F. (S.D.-C., L.C., A.O.); Neurological Unit, Boston City Hospital, Boston, Massachusetts 02118 (T.K.); Worcester Foundation for Experimental Biology, Shrewsbury, Massachusetts 01545 (P.J.M.)

ABSTRACT The effect of protein deprivation on rapid Golgi impregnated pyramidal neurons in layers IT/III and V of the rat visual cortex was studied at 30,90, and 220 days of age using morphometric methods. In order to mimic human undernutrition female rats were adapted to either an 8 % or control 25% casein diet 5 weeks prior to conception and maintained on these diets during gestation and lactation. The pups were then weaned and maintained on their respective diets. The undernourished rats showed a significant decrease in brain weight only at 90 days, indicating that the protein deprivation had a mild effect on brain development. Correspondingly, the number of significant histological differences between the two diet groups were least a t 30 and 220 days of age. The effect of the diet was greater on layer V than on layer II/III pyramids. At 30 days of age the effect of the diet was different on the pyramids of these two cell layers, at 90 days there was a mixture of similar and dissimilar effects, and at 220 days the pyramids of these two cell layers showed only minor differences between the two diet groups. Analysis of age-related changes indicated that the effect of the diet was different on layer II/III pyramids compared to layer V pyramidal cells. These different effects apparently accounted for the progression from a dissimilar effect of the diet at 30 days on the pyramids of the two cell layers to only minor differences between them at 220 days. Further analysis of these age-related changes shows that two prominent effects of protein deprivation are for age-related changes to occur in undernourished rats but not in controls and for age-related changes to be out-of-phase with each other in the two diet groups. From these findings, and a review of similar studies in the literature, we propose that these mechanisms are a prominent effect of undernutrition in the post-weaning period and help account for the unexpected increases in morphometric measurements noted in undernourished rats in this and other studies. Key words: cerebral cortex, occipital cortex, dendritic development

Since the classic paper of Sugita ('18), many aspects of the effect of undernutrition on the development of the rodent cerebral cortex have been studied. Despite differences in the methods of producing nutritional deprivation, methods of analysis of the data, differences in sacrifice time and technical considerations (Bedi, '87; Bedi and Warren, '881, some general principles of the effect of undernutrition on cerebral cortical development are emerging. The volume of the cerebral cortex is reduced in proportion to that shown by the whole brain (West and Kemper, '76; Morgane et al., '78;

0 1990 WILEY-LISS, INC.

Thomas et al., '79; Bedi, '87), without evidence of a decrease in the total number of neurons (Sugita, '18; Leuba and Rabinowicz, '79a). This results in increased neuronal cell packing density (Sugit,a, '18; Callison and Spencer, '68; Cragg, '72; Siassi and Siassi, '73; Cordero et al., '76; Leuba and Rabinowicz, '79a; Thomas et al., '79, '80; Bedi et al., '80; Warren and Bedi, '82, '84; Bedi and Warren, '88) and

Accepted September 6,1989.

118

decreased cortical width when compared to controls (Callison and Spencer, '68; Bass et al., '70; Dobbing et al., '71; Cragg, '72; Clark et al., '73; Siassi and Siassi, '73; West and Kemper, '76; Noback and Eiseman, '81). The total number of glial cells has also been reported to be decreased in undernutrition (Leuba and Rabinowicz, '79a). Golgi studies of pyramidal cells in the cerebral cortex have indicated that virt,ually all measurable parameters are affected by undernutrition (Salas et al., '74; West and Kemper, '76; AnguloColmenares et al., '79; Leuba and Rabinowicz, '79b; Noback and Eiseman, '81; Schijnheit, '82, Schonheit and Haensel, '89). The density of dendritic spines is reduced (Salas et al., '74; West and Kemper, '76; Leuba and Rabinowicz, '79b; Nohack and Eiseman, '81; Schonheit, '82, Schonheit and Haensel, '88),width of the dendrites decreased (Salas et al., '74; West and Kemper, '76; Angulo-Colmenares et al., '79; Leuba and Rabinowicz, '79b), and complexity of dendritic branching reduced (Leuba and Rabinowicz, '79b; Schonheit, '82) due to undernutrition. Dendritic extent is reported to be unaffected (West and Kernper, '76; Schonheit, '82) save for apical dendrites (West and Kemper, '76). Quantitative electron microscopic studies have indicated that the density of synapses in the neuropil is little or unaffected by undernutrition (Cragg, '72; Gambetti et al., '74; Warren and Bedi, '84). Medvedev and Babichenko ('88) reported in young and adult mice that undernut.rition significantly alters the ultrastructure of synapses with reduced length of the reactive zone, width of the synaptic cleft and number of cisterns in the spinous apparatus. In a recent study Warren et al. ('89) found a 30'!0 increase in the number of synapses per neuron in the visual cortex of 111-day-old rats undernourished during the previous 29 clays. They attributed this to a delay in t,he normal age-related decline in the pattern of neural connectivity in this area in undernourished rats. In the majority of these morphological studies on the efrects of undernutrition on the brain primary emphasis has been placed on undernutrition during the period of lactation (Sugita. '18; Callison, '68; Bass et al., '70: Dobbing et al., '71; Siassi and Siassi, '73; Salas et al., '74; Cordero et al., '76; 1,euba and Rabinowicz, '79a,b; Thomas et al., '79, '80; Bedi et al., '80; Noback and Eiseman, '81 Schonheit and Haensel, '88). Several studies have also examined the effects of undernutrition during gestation (Gambetti et al., '74; West and Kemper, '76; Morgane et al., '78; Angulo-Colmenares et al., '79; Leuba and Rabinowicz, '79a,b; Warren and Bedi, '82, '84) or continued after weaning (Cragg, '72; West and Kemper, '76; Morgane et al., '78; Schonheit, '82; Warren and Redi, '82, '84). In a few studies undernutrition has been confined only to the period of gestation (Clark et al., '73; Siassi and Siassi. '73). Many of these experiments have included examination of the effects of dietary rehabilitation (Sugita, '18: Bass et al., '70; Dobbing et al., '71; Siassi and Siassi, '73; Cordero et al., '76; Angulo-Colmenares et al., '79; Leuba and Rabinowicz '79a,b; Thomas et al., '79; Bedi et al., '80; Thomas et al., '80). These latter have resulted in reversals of deficits in neuronal cell packing density (Siassi and Siassi, '73; Leuba and Rabinowicz, '79a; Thomas et al., '80; Bedi et al., '80: Warren and Bedi, '84) and in cortical width (Bass et al., '70; Angulo-Colmenares et al., '79) with a variable effect on glial cell density (Leuba and Rabinowicz, '79a). However, some Golgi studies have shown persistent deficits in dendrite branching and synaptic spine density despite prolonged dietary rehabilitation (Leuba and Rabinowicz, '79b). Schiinheit, and Haensel ('88) found deficits in total apical and basal dendritic extent and average spine density in

S. DIAZ-CINTRA ET AL.

layer I11 and V cingulate cortex pyramidal cells a t 20 postnatal days following undernutrition during lactation that was reversed after 40 days of rehabilitation, with many of these measurements at that time actually increased in the experimental rats. In few of the above studies were adult animals that had been exposed to nutritional deprivation throughout their prenatal and postnatal life span examined, which is the paradigm used in the present experiments. The dietary regimen in our studies were chosen to mimic as closely as possible that found in human undernutrition which usually involves protein deprivation for a prolonged period of time, both prenatally and postnatally. The sacrifice times were selected to match our previous studies on the effects of protein deprivation on the nucleus raphe dorsalis and nucleus locus coeruleus (Diaz-Cintra et al., '81, '84) and to complement studies in progress on the effect of a low protein diet on nonpyramidal neurons in the visual cortical area examined in the present study.

rnTHODS The details of the dietary paradigm and breeding procedures can be found in Morgane et al. ('78). In the present experiment, virgin female Charles River C.D. Sprague-Dawley descended rats were fed either a 25% or 8% casein diet 5 weeks prior to conception and mated with normal diet males. They remained on these diets during gestation and lactation following which the pups were weaned and then maintained on their respective diets. At birth, all litters from each diet group were randomized and culled to 8 pups per litter. A t 30, 90, and 220 days, male rats from both groups were weighed, anesthetized with pentobarbital, perfused through the heart with 1000 neutral buffered formalin, and the brains removed and weighed the following day. The extent of the visual cortex was determined in rat brains cut in the frontal plane according to the atlas of Paxinos and Watson ('82). A 4 mm wide block of visual cortex from five brains from both diet groups at all three ages was prepared with the rapid-Golgi technique following the modification of Diaz-Cintra et al. ('81). The blocks were embedded in low viscosity nitrocellulose, serially cut at a thickness of 120-160 ,urn, and five sections from each brain containing visual cortex were mounted in serial order. Each slide was assigned a random number to insure that the observations were blind with respect to age and diet. From each section, well-impregnated pyramidal cells were selected from the lower parts of layers II/III and layer V, yielding a total of 40 neurons from each layer for each age and diet group. All measurements were made by means of a 4Ox planapochromatic or 100x planapochromatic objective with a calibrated ocular reticle. Measurements of basal dendritic complexity were made according to the method of Sholl ('56) using a camera lucida and a projected reticle showing concentric rings the equivalent of 38 pm apart in the histological preparation (Fig. 1).The percentage of dendrites transected by Fig. 1. Camera lucida drawing of pyramidal cells from layers II/III and V of rat visual cortex showing parameters measured. Dendritic density was determined by the number of dendrites crossing 4 concentric rings 38 pm apart. In the largest basal dendrite of each neuron the number of spines was counted between each concentric ring. The small and arrows delineate a 50 K r n segment on the proximal (P), middle (M), terminal (T) parts of the apical dendrite where spines were counted. Dashed lines shows the linear extent (LE) of the apical dendrite.

PROTEIN DEPRIVATION AND VISUAL CORTEX

119

I

I

I LE

I I I i

I

I I

I I

L-

a

!,

V Figure 1

120

S . DfAZ-CINTRA ET AL.

the microtome in the first concentric ring was less than 209; in both control and undernourished animals. With these methods, the following measurements were made: 1) major and minor axis of the cell body, 2) number of intersections of the basal dendrites crossing each concentric ring, 3) number of basal dendritic spines on the largest dendrite within each concentric circle, 4) linear extent of apical dendrite [dendrites in the plane of sections were measured directly and those not in the plane of section were estimated by triangulation according to the method of Bok ('56)], and 5) spine density per 50 pm on proximal, middle, and terminal segments of the apical dendrite. As a part of this study body and brain weights of 25% and 8% casein diet rats were also compared a t each age. Statistical analyses were conducted using the General Linear Model procedures of the SAS data analysis system (SAS Institute Inc., '85). Brain and body weights were analyzed in a two-way analysis of variance (ANOVA) design (diet x age) without the repeated measures program. Additional comparisons of diet effects between both diet groups (n = 5/diet) were made on body and brain weights. Comparisons across the age groups were made on brain weight and, after significant interaction, also on body weight. Neural measures were analyzed with a different program using a repeated-measures design with two between-subjects factors (diet and age) and eight replicates (neurons) for each subject. Significant ANOVA effects of diet, age, or interaction were further examined using three different comparisons. Thus, diet effects were compared at each age between the control and undernourished rats (n = 5/diet), age comparisons was made across the age groups (n = lO/diet), and the effect of diet among age groups were compared after significant interaction (n = 10/diet). Probability values of these comparisons were adjusted upward using the Bonferroni method (Ryan, '59).

between 90 and 220 days (Table 2). Comparisons of body and brain weight between the two diet groups a t each age show a significant decrease in body weight in the undernourished rats at 90 and 220 days, but not at 30 days of age. There was a significant decrease in brain weight only a t 90 days of age (F[1,24] = 11.17).

Layer II/III pyramidal cells Table 3 shows the means (+SEM) uf all parameters measured in control and undernourished rats and the P values for comparisons between the two diet groups a t each age. Table 4 shows the two-way ANOVA results for which there is no significant interaction effect. These are for the number of basal dendritic intersections and on basal and apical dendritic synaptic spine density. Table 5 shows the ANOVA results for data in which there is a significant interaction effect. These are for cell size and length of the apical dendrite. In Table 4 the ANOVA shows no significant diet or age effect on the number of basal dendritic intersections. The number of synaptic spines within the first concentric ring shows a significant diet and age effect and within the second concentric ring a significant diet effect. There are no significant effects within the third concentric ring. Comparisons between the ages shows significant age-related changes in the first concentric ring between 30 and 90 and 30 and 220 days, but not between 90 and 220 days. Inspection of the data in Table 3 reveals that the significant age-related changes in synaptic spine density are primarily due to a decrease in spine density in the control rats between 30 and 90 days of age. Comparison between the two diet groups a t each age shows a significant increase in synaptic spine density in the protein deprived rats at all three ages within the

TAJ3LE 1. Body and Brain Weights in Control and Undernourished Rats of Three Age Groups'

RESULTS Body and brain weights

Age

Table 1 shows the mean (kSEM) for body and brain weights in both diet groups and the P values for comparisons between the two diet groups at each age. The two-way ANOVA is shown in Table 2. The ANOVA reveals significant effects of diet and age on body and brain weight. Body weight shows a significant interaction between diet and age between 30 and 90,90 and 220, and 30 and 220 days, indicating a different age-related diet effect between each of these ages. Comparisons of body and brain weight between the three ages show significant increases in body weight between all three ages and significant increases in brain weight between 30 and 90 days and 30 and 220 days, but not

(davs)

Weights

F Values'

(9)

25 7%

8%

B ~ Y Brain

58.62+ 6.51 I .59 i 0.04

31.31 + 2.22 1.44 * 0.06

30

90 Y

497.10

Brain

2.38

M

220

A

20.01

t

0.01

6a.71 + 9.72 2.45 * 0.05

Body Brain

313.69 h 14.11 2.05

t

0.04

56851 + 34.39 2.28 5 0.13

1.16NS 2.38NS 52.43." 11.17" 22.901"

2.96 NS

'Mean SEM of 5 animals. 'Additional mmpmona between diets at each age on body and brain weight after two-way ANOVAht. **P