THE JOURNAL OF COMPARATIVE NEUROLOGY 282570-580 (1989)

Structure and Histogenesis of the Principal Sensory Nucleus of the Trigeminal Nerve: Effects of Prenatal Exposure to Ethanol

MICHAEL W. MILLER AND SUSAN J. MULLER Department of Anatomy, School of Osteopathic Medicine and Robert Wood Johnson Medical School, University of Medicine and Dentistry of New Jersey, Piscataway, New Jersey 08854

ABSTRACT Clinical and experimental evidence shows that prenatal exposure to etha-

nol causes craniofacial malformations, microcephaly, and abnormal develop- ment of the central nervous system. This study describes the effects of ethanol on the development of the principal sensory nucleus of the trigeminal nerve (PSN). The offspring of two groups of rats were examined. Pregnant females in one group were fed a liquid diet containing 6.7 % (v/v) ethanol (Et) and rats in the other group were fed an isocaloric liquid control diet (Ct). Each preg- nant rat was administered [3H]thymidine on one day during the period from gestational day (G) 10 to G22. After pups grew to 30 days of age, they were killed and their brains were processed by an autoradiographic procedure.

Qualitatively, the PSN of Ct- and Et-treated rats appeared similar; they were composed chiefly of small neurons and a few scattered large neurons. On the other hand, quantitative analyses revealed significant differences between both groups. Although the volume of the PSN of Et-treated rats was not sig- nificantly different (-3.2%) than that for Ct-treated rats, the PSN of Et- treated rats had significantly (P < 0.01) fewer (30.0%) neurons than did the PSN of Ct-treated rats. The number of the small neurons, but not of the large neurons, was affected most by the ethanol exposure. Prenatal exposure to eth- anol also altered the generation of PSN neurons. Most neurons in the PSN of Ct-treated rats were born between G12 and G15, the small neurons being gen- erated before the large neurons. In Et-treated rats, too, small neurons were born before the large neurons; however, the time frame of neuronogenesis was delayed as it occurred between G13 and G16.

Thus, prenatal exposure to ethanol produces profound developmental abnormalities that lead to permanent alterations in the structure of the mature central nervous system.

Key words: birthdates, fetal alcohol syndrome, craniofacial malformations, neuronogenesis, gliogenesis

The number of cells populating a nucleus in the mature central nervous system represents the culmination of four developmental processes: cell proliferation, migration, dif- ferentiation, and death. Neuronal precursors pass through various phases of the cell cycle in the proliferative zones that line the ventricles (e.g., Sauer, '36; Atlas and Bond, '65; Waechter and Jaensch, '72). After exiting this cycle, young neurons migrate to the site of the immature nucleus (Ange- vine and Sidman, '61; Berry and Rogers, '65; Hicks and D'Amato, '68; Al-Ghoul and Miller, '89, in preparation). Once in the nucleus, cell bodies grow, neurons sprout den- drites, and synaptogenesis begins (Clark, '77; Miller, '88a; Al-Ghoul and Miller, '88, in preparation). Neurons which

0 1989 ALAN R. LISS. INC.

are unsuccessful in competing for synaptic sites or trophic factors die (Hamburger and Oppenheim, '82; Clarke, '85).

Alcohol consumption during pregnancy affects these ba- sic developmental processes. Of the structures in the central nervous system, the effects of ethanol on the development of the neocortex have been examined most comprehensively. Cell proliferation is altered by ethanol exposure (Miller, '88b, submitted). Neocortical neurons are derived from cells

Accepted October 19,1988.

571 EFFECTS OF ETHANOL ON NEURONAL GENERATION

in the ventricular zone and from cells in the subventricular zone. Ethanol depresses proliferative activity in the ventric- ular zone and increases proliferative activity in the subven- tricular zone. The migration of young neocortical neurons also is disrupted by ethanol (Miller, '86; '88b), particularly the migration of late-generated neurons. Many of these neu- rons terminate their migration in deep cortex rather than moving to the supragranular laminae. Prenatal exposure to ethanol ultimately affects the postnatal morphogenesis and synaptogenesis of neocortical neurons. The growth of den- drites is increased (Miller et al., '88), dendritic spines are more plentiful and often dysmorphic (e.g., Abel et al., '83; Stoltenburg-Didinger and Spohr, '83; Schapiro et al., '84; Miller et al., '881, and the axonal connections of projection and local circuit neurons are altered (Al-Rabiai and Miller, '87; Miller '87).

The effect of prenatal exposure to ethanol on cellular development underlies the constellation of teratological de- fects evident clinically and in experimental models of fetal alcohol syndrome. These defects include craniofacial mal- formations (e.g., Clarren and Smith, '78; Sulik et al., '81; Edwards and Dow-Edwards, submitted), microcephaly (Parpara-Nicholson and Telford, '57; Tze and Lee, '75; Ran- dall et al., '77; Diaz and Samson, '80; Miller, '87), and mental retardation (Shaywitz et al., '80; Streissguth, '86; Meyer and Riley, '86). In light of these clinical and experimental find- ings, we examined the effects of ethanol on the development of one of the nuclei receiving primary input from receptors in the face, the principal sensory nucleus of the trigeminal nerve.

MATERIALS AND METHODS Animals, breeding, and feeding

Long-Evans hooded rats were obtained from Charles River Breeding Labs (Wilmington, MA). Rats were bred in the following manner. In the evening, one male rat was placed with 4 or 5 females. The following day a t 900 P.M., females were checked for vaginal plugs; if present, that day was designated as gestational day (G) 1 and each pregnant female was placed in a separate cage. From G1 to G4, all pregnant rats were fed a diet of chow and water. On G5, pregnant rats were grouped into weight-matched sets of three. Two rats from each triad were fed protein-enriched liquid diets obtained from Bio-Serv Laboratories (French- town, NJ). One diet contained 6.7% (v/v) ethanol (Et), which represented 37% of the caloric content. Another diet, a control diet (Ct), contained no ethanol, but it was isoca- loric with the ethanol-containing diet. The third rat from each triad was maintained on a diet of chow and water (Ch) throughout the pregnancy.

Ch- and Et-fed animals were both given food ad libitum whereas the Ct-fed animals were pair-fed. According to this system, each Ct-fed rat was given a volume of diet equal to that consumed the previous day by the weight-matched Et- fed female. This ensured that the daily food consumption of the paired Et-fed and Ct-fed rats was of identical caloric value. Et-fed rats were gradually weaned onto the diet over a period of 6 days. From G5 to G6 the Et-fed animals were given a diet containing 2.2% ethanol; from G7 to G9 the diet contained 4.5% ethanol; and from G10 to the day of birth the diet contained 6.7% ethanol.

Animals were fed daily a t 9:OO-1O:OO A.M. (EST), at which time the amount of liquid diet injested was recorded. On alternate days, pregnant rats were weighed. If the weight of an Et-fed rat was greater than 10% less than the weight-

matched Ch- or Ct-fed rats, then it was removed from the study. Data from previous studies (Miller, '87; '88b) showed that the weight gains of Ch- and Ct-fed rats were not signifi- cantly different. Therefore, in order to streamline the quan- titative analyses, the offspring of Ch-fed rats were not examined. After birth all litters were culled to eight and sur- rogate fostered by Ch-fed mothers. This was done to mini- mize the effects of postnatal malnutrition.

Samples of maternal blood were taken a t 1O:OO P.M. on G17 for assays of blood ethanol concentrations. Only one determination was made during the pregnancy because in an earlier study (Miller, '88b) it was shown that the concen- trations of ethanol in the blood of rats fed by the present protocol were stable throughout most of the period during which rats were fed the 6.7 % diet.

Autoradiography The times of origin of cells in the principal sensory

nucleus of the trigeminal nerve (PSN) of Et- and Ct-treated rats were determined by use of standard autoradiographic procedures (e.g., Sidman, '70; Miller, '85). A single pulse injection of [3H]thymidine (New England Nuclear; specific activity 81 Ci/mmole; 5.0 pCi/g body weight) was made on one day during the period of G10 thru G22. Each pregnant rat received only one injection. All injections were placed at 2:OO-3:00 P.M.

Pups were killed on postnatal day 30, which was about four weeks after the completion of neuronal migration (Al- Ghoul and Miller, '88, in preparation). Only one rat from each of ten Ct- and ten Et-treated litters was examined. Anesthetized rats (0.36 mg chloral hydrate/g body weight) were transcardially perfused with a solution of 4.0% para- formaldehyde in 0.10 M phosphate buffer. Fixed brains were removed from their skulls, dehydrated, cleared, and embedded in blocks of paraffin. Each brain was cut into 10 pm coronal sections and a series composed of every tenth section was mounted on slides. Subsequently, the sections were defatted and coated with Kodak NTB2 autoradio- graphic emulsion. The emulsion was exposed for 25-28 days at - 10°C and developed in Kodak developer D19 for 3 min- utes at 11°C. Finally, the slides were rehydrated, stained with cresyl violet, and coverslipped for examination by light microscopy. Only heavily labeled cells were examined (Miller, '85). Presumably these cells incorporated some t3H]thymidine, divided only once, and then migrated to the PSN before making an additional pass(es) through the mitotic cycle.

The amount of shrinkage of the tissue due to the postper- fusion processing was assessed in five Ct- and five Et- treated rats. Two pairs of holes were made in the fixed brainstems by impaling the brainstems with pins that were spaced 1.00 mm apart. One pair of pins was set in a longitu- dinal plane, i.e., the anterior-posterior plane, and the other pair was aligned in the perpendicular transverse plane, i.e., the medial-lateral plane. The distance between these pin holes was measured in the sectioned brains and used as an index of shrinkage.

Morphometry Various morphometric studies were performed. These in-

cluded determination of the volume of the PSN, the number and size of PSN neurons, and the number and size of PSN cells generated on a specific day of gestation.

An estimate of the volume of the PSN on each side of the brainstem was calculated from the sum of the cross-sec-

M.W. MILLER AND S.J. MULLER

7 8 VI IC IV 11 LL LV mcP m5

facial nerve vestibulocochlear nerve abducens nucleus inferior colliculus intratrigeminal nucleus lateral lemniscus nucleus of the lateral lemniscus lateral vestibular nucleus middle cerebral peduncle motor root of the trigeminal nerve

Abbreviations

MV MeV PSN PYr s6 sc sv SVe svo vc

motor nucleus of the trigeminal nerve mesencephalic nucleus of the trigeminal nerve principal sensory nucleus of the trigeminal nerve pyramidal tract tract of the trigeminal nerve superior colliculus supratrigeminal nucleus superior vestibular nucleus subnucleus oralis of the trigeminal nerve ventral cochlear nucleus

573 EFFECTS OF ETHANOL ON NEURONAL GENERATION

tional areas of the seven to nine sections through the nucleus and by multiplying this sum by 100 (tenfold for the thickness of the section and tenfold for the frequency of sec- tions composing the series).

The total number of neurons in the PSN of Ct- and Et- treated rats was assessed as follows. The right brainstems from each of ten Ct- and ten Et-treated rats were used to estimate the total number of PSN neurons. For each animal, the cross-sectional area of neurons in the four middle sec- tions of PSN were determined (Fig. 1). Only neurons which had nuclei were counted. Glia were discriminated from neu- rons on the basis of nuclear size and appearance and somatic size (Vaughan, '84) and they were excluded from these counts. The number of neurons in these sections was esti- mated by use of the Floderus ('44) method:

N, = N,(t)/(t + D - 2k)

where N, was the number of neurons per section, N, was the mean number of neurons counted on each of the middle four sections, t was the section thickness (10 bm), D was the mean maximal diameter of neurons with nuclei, and k was the height of the smallest recognizable somatic cap with cut nucleus. The mean maximal diameter was determined by an iterative calculation described by Smolen et al. ('83). The correction factor, t/(t + D - 2K), was determined indepen- dently for each animal. Two assumptions were made for the purposes of applying this formula: that the cell bodies of all neurons were the same size and that they were spherical. The appropriateness of these assumptions is addressed in the Results.

Once the mean number of neurons per section was deter- mined, the mean total number of neurons in the PSN was calculated. This was achieved by equating the ratio of the mean number of neurons in each of the four sections (N, which was calculated above) by the mean volume of the four sections analyzed (V,) to the ratio of the mean number of neurons in the entire PSN (N,) to the total volume of the PSN Wt);

Solving for the total number of PSN neurons, the formula is

Nt = N,(VJN,.

Implicit in this calculation is the assumption that the cell- packing density was consistent throughout the rostral to caudal extent of the PSN.

From the present data, the duration of the S-phase of the cell cycle of precursors of PSN neurons was estimated. [3H]thymidine is taken up by cells in the S-phase of the cell cycle (Sidman, '70); therefore, the proportion of labeled neurons to the total population was an index of the relative amount of the time spent in the S-phase. This index was cal- culated as the sum of the number of PSN neurons labeled by a series of injections of [3H]thymidine placed during the ges- tation period (i.e., the number of neurons per section labeled by an injection on G12, plus those labeled by an injection on G13, etc.) divided by the total number of neurons per sec- tion. These calculations are based on the assumption that on each day during the period of PSN neuronogenesis a similar proportion of labeled to unlabeled cells die naturally and/or because of the prenatal exposure to ethanol.

All morphometric determinations relied on the Bioquant Image Analysis System (R & M Biometrics, Nashville, TN). Statistical differences were assessed by the Hoetelling's t- test for multiple comparisons.

RESULTS Food consumption

During gestation, the weight gain by Et-fed dams was not significantly different than that of Ch- or Ct-fed rats. From G6 to the day of birth, Et-fed rats gained a mean 31.2% in weight and the Ch- and Ct-fed rats gained 32.2% and 31.9%, respectively. Only two of the 30 Et-fed rats were removed from the study for failing to maintain a weight within 10% of the Ch-fed dams. In the Et-fed rats, the mean blood ethanol concentration (k the standard error of the mean) was 160 k 21 mg/dl (n = 5) on G17. Most Ch- and Ct- treated pups were born on the morning or during the after- noon of G22, whereas Et-treated rats were born on the morning of G23.

Nuclear volume In both Ct- and Et-treated rats, the PSN was a small pon-

tine nucleus that was surrounded by fiber bundles (Figs. 1, 2). The ventral, lateral, and dorsal aspects of the PSN were surrounded by the tract of the trigeminal nerve. Medially the PSN was bordered by the motor root of the trigeminal nerve and by the facial nerve. The rostral limit of the PSN was capped by the lateral lemniscus. The PSN extended from a level rostral to the trigeminal motor nucleus to a level immediately caudal to the motor nucleus. The caudal bor- der of the PSN was shared with the subnucleus oralis; how- ever, the two nuclei were distinguished by the looser packing of cells in the subnucleus oralis and the substantial compo- nent of large neurons which were distributed throughout the subnucleus oralis. In contrast, the PSN was composed of a relatively homogeneous population of densely packed multi- polar neurons. The majority of these neurons were small; however, a few large neurons were dispersed throughout the dorsal region of the PSN (Fig. 2). Prenatal exposure to etha- nol affected neither the position of the PSN in the brain- stem nor its overall cellular appearance.

In three-dimensional reconstructions, the PSN of Et- and Ct-treated rats appeared as an elongate biconcave disk (Fig. 3). In the ten Ct-treated animals examined, no significant difference was determined for the mean volume of the PSN on the two sides of the brainstem. The mean volume for the PSN on the left side was 0.370 k 0.005 mm3 and for that on the right side was 0.375 0.004 mm3. Likewise, in ten ani- mals prenatally exposed to ethanol, there was no difference between sides; the mean volume of the PSN on the right and left sides were 0.362 5 0.002 mm3 and 0.358 f 0.005 mm ', respectively. Since there was no evidence of laterality, the data for the two sides were pooled for comparisons of the dietary groups. Hence, the mean volume of the PSN for Ct- and Et-treated rats was 0.372 +. 0.003 mm and 0.360 * 0.003 mm3, respectively. This 3.2% difference between di- etary groups was not statistically significant.

Shrinkage of the brain produced by the embedding proce- dure was assessed by measuring the distance between two holes placed 1.00 mm apart in the fixed brain. In the embed- ded brainstems of Ct-treated rats, the distmce between the corresponding holes set in the longitudinal plane and those

574 M.W. MILLER AND S.J. MULLER

Fig. 2. Cytoarchitecture of the PSN. In Ct-treated rats (A) and in Et-treated rats (B), the PSN is composed of densely packed small neu-

rons. In addition, a few large neurons are in the dorsal half of the F'SN. Cresyl violet.

set in the transverse plane was reduced to 0.72 -r- 0.03 mm (mean standard error of the mean) and 0.68 2 0.04 mm, respectively. For the processed brainstems of Et-treated rats, the distance between the two holes in the longitudinal plane was 0.69 5 0.05 mm and the distance between the two holes in transverse plane was 0.64 s 0.03 mm. Since there were no statistically significant differences between dietary groups or between planes, shrinkage was not considered to be a confounding for the morphometric studies. Thus, in all of the data presented, the amount of shrinkage was not fac- tored into any of the calculations.

Cellular composition

474 _c 9 and 332 2 6 neurons per section, respectively. In addition, from these data i t was calculated that the correc- ted mean maximal cross-sectional area of Ct- and Et-treated rats was 109 * 2 pm2 and 123 s 1 pm2, respectively. Thus, the mean size of the cell bodies of PSN neurons was signifi- cantly (P < 0.01) larger (30.070) in Et-treated rats than in the pair-fed controls. This difference does not indicate that the size of the neurons was affected by prenatal exposure. Quite the contrary, the difference in mean cell size results from a change in the numbers of small and large neurons. An analysis of the distribution of the size of neurons in the PSN (Fig. 4) showed that the small neurons were most affected by the prenatal exposure. In Et-treated rats, there were

In Ct-treated rats, the corrected mean maximal diameter of a PSN neuron was 11.8 i 0.1 pm. For Et-treated rats, however, the mean diameter of a PSN neuron was 12.5 f 0.2 pm. These data were used in Floderus's ('44) formula, and it was determined that in Ct- and Et-treated rats there were

314 f 13 neurons per section with a corrected mean cross- sectional area of less than 250 pm2, which was 32% less than that in pair-fed controls (452 i 12 neurons). This difference was statistically significant ( P < 0.01). In contrast, the num- ber of neurons with a corrected mean cross-sectional area

CONTROL ETHANOL 7 5

0.372 - 0.003 mm3 0.360 - 0.003 mm3 c

.- - 0 UI

50 0 m

c f 0

r

n 5 z E 2 5 -

TABLE 1. Somatic Eccentricity and Longest Cord Lengths'

- I I I - - i-t *-• Control [{..A

f+ \

i ; '>>I - / : ;

:I

O - - - - O Ethanol

-

'!r ')\I

I f I--a..$xa_

-

Q\r f j ,L

u=.4-..i..jl;.0-. o..b--p-. - I I I

1 0

Ct-treated Et-treated - - X s.e.m. X s.e.m.

Transverse eccentricity 0.784 0.015 0.800 0.010 Longest cord (Sm) 13.3 0.3 13.9 0.2

Longitudinal eccentricity 0.812 0.018 0.789 0.016 Longest cord (rm) 13.2 0.2 14.1 0.3

'Each value is the mean (x i ) of 200 neurons from each of 10 rats. 8.e.m. represents the stan- dard errors of the means, respectively.

greater than 250 fim2 in the PSN of Ct-treated rats (22.4 t 1.4 neurons) was not significantly different than that in Et-treated rats (18.6 t 2.0 neurons).

For the purposes of these morphometric calculations, it was assumed that the cell bodies of all PSN neurons were approximated as spheres. This assumption was tested by determining the mean length of the longest cord of the cross-sectional profile of a cell body and the mean eccen- tricity (the ratio of the length of the shortest cord to the length of the longest cord; a circle has an eccentricity of 1.00) in both transverse and horizontal sections of the PSN. These data are provided in Table 1. Regardless of the prena- tal exposure, the differences in the mean eccentricity of the cell bodies examined in transverse and in horizontal sections was not significant. More important, there were no signifi- cant differences between the Ct- and Et-treated rats. On the other hand, the differences in the mean length of the longest cord in Ct-and Et-treated rats were significant; however, these values were consistently larger in the Et-treated rats than in the controls and by a similar factor. Therefore, the three-dimensional shape of the PSN neurons was not affected by the prenatal exposure to ethanol, and it can be considered spheroidal. Moreover, the orientation of neurons

in the PSN was effectively random and did not bias the results.

The second assumption implicit in applying Floderus's ('44) formula was that all neurons were of similar size. It is recognized that the population of neurons in the rat PSN is not homogeneous in size (e.g., Altman and Bayer, '80; Erzu- rumlu and Killackey, '83). Neurons with a mean cross-sec- tional area greater than 250 pm2, however, composed only 4.73 and 5.60% of the total population in Ct- and Et-treated rats, respectively. Thus, the assumption of homogeneity of size becomes useful.

From the data on the volume of the PSN and the cell- packing density, it was estimated that the mean total num- ber of neurons in the PSN of Ct- and Et-treated rats was 28,000 t 600 and 19,600 400, respectively. This difference was statistically significant (P < 0.01) and showed that the PSN of Et-treated rats had 30.0% fewer neurons than Ct- treated rats.

The mean volume of a neuronal cell body was calculated from the linear data on cell size. Assuming that the cell bod- ies of PSN neurons were spheres, it was determined that the mean volume was 860 17 fim3 in Ct-treated rats and 1,020 10 film3 in Et-treated rats. Taking into account the data on the total number of neurons per PSN and the vol- ume of the PSN, it is estimated that the total volume occu- pied by neuronal cell bodies is 6.65 -r- 0.11% and 5.56 +- 0.04% of the total PSN volume of Ct- and Et-treated rats, respectively.

Time of origin The generation of cells in the PSN was examined from

GI0 through G22. A cell was considered to have been gener- ated on the day of the injection if it was heavily labeled with autoradiographic silver grains (Miller, '85). In Ct-treated rats, no PSN cells were heavily labeled by injections of

576 M.W. MILLER AND S.J. MULLER

Figure 5

EFFECTS OF ETHANOL ON NEURONAL GENERATION 577

30 F' .-. Cantrol o - - - - o Ethanol

1 2 1 4 16 18 2 0 2 2

Time of origin (ges ta t iona l day)

Fig. 6. Number of neurons generated daily. Each value represents the mean of two rats from each of two litters. Notations as in Figure 4.

[3H]thymidine on G10 or G11. Neurons were generated from GI2 through G15. The majority of neurons were generated on G12-G14, and neuronogenesis reached its peak on G13 (Figs. 5, 6). The neurons born on G12, G13, or G14 were small neurons (Figs. 7, 8). In contrast, large neurons were born on G15. Between G16 and G22, mostly glia were born; however, neuronal generation continued a t a low rate during this period. From two to four neurons per PSN were born daily during the period from G16 to G22 (Fig. 9), and these neurons were of mixed sizes.

Although this basic pattern of neuronogenesis was not altered by prenatal exposure to ethanol, the temporal sequence was delayed by about 1 day. As with Ct-treated rats, small neurons in Et-treated rats were generated before large neurons. In Et-treated rats, most small neurons were generated on G13 and G14 (Fig. 5), the peak of neuronal generation occurring on G14 (Fig. 6). Moreover, these neu- rons were significantly smaller than the early generated neurons in the pair-fed control animals (Figs. 7, 8). Although a few large neurons were generated on G15, most were born on G16, i.e., one day later than for the Ct-treated rats. The prenatal diet, however, did not have a significant effect on the size of these late-generated large neurons. A few neurons in Et-treated rats were also heavily labeled by injections of [3H]thymidine placed on 1 day during the period from G16 to G22 (Fig. 9). Interestingly, as with Ct- treated rats, the first heavily labeled glia in Et-treated rats were labeled by an injection of [3H]thymidine on G16. Glial generation continued through the last week of gestation, and the numbers and size of the glia generated daily were not significantly different in Ct- and Et-treated rats. Thus, gliogenesis did not appear to be affected by ethanol expo- sure.

An index of the duration of the S-phase was determined as described in the Materials and Methods. For Ct- and Et-

I I I 1 I 1 5 0

I

I I I I I 1 2 14 16 18 2 0 22

Time of origin (ges ta t iona l day)

Fig. 7. Figure 5.

Mean size of neurons generated daily. Notations as in

treated rats, this index was 0.134 and 0.154, respectively. These indices were derived from data based on a single injection per day; hence, multiplying these indices by 24 hours provided an estimate of the duration of the S-phase. Accordingly, it was estimated that the S-phase of rhomben- cephalic ventricular cells in Ct-treated rats was 3.22 hours and 3.70 hours for Et-treated rats.

DISCUSSION One of the characteristic defects related to prenatal expo-

sure to ethanol is microcephaly (Parpara-Nicholson and Telford, '57; Tze and Lee, '75; Randall et al., '77; Diaz and Samson, '80; Miller, '87). The degree of microcephaly differs within the central nervous system. Gestational exposure to ethanol has a large impact on the volume of the cerebrum; the volume of the cerebrum is 12-17% less in Et-treated rats than in chow or pair-fed controls (Miller, '87). In con- trast, the effect on the brainstem, as exemplified by the PSN, is relatively small (3.2%), albeit significant.

Despite the small effect on the overall volume of the PSN, the cellular composition of the PSN was altered by prenatal exposure to ethanol. The number and size of neurons within the mature PSN in Et-treated rats were an order of magni- tude less than in pair-fed controls. Moreover, the total vol- ume occupied by neuronal cell bodies was different in the two groups. The implication from these data is that the vol- ume of the neuropil, i.e., neuronal and glial processes, and of blood vessels is greater in Et-treated rats than in controls. Such conclusions are in concert with data from neocortex wherein the cell body to neuropil ratio is less in Et-treated rats than in controls (Potempa and Miller, submitted). Moreover, recent evidence has shown that the ethanol- induced increase in neocortical neuropil results from an increase in dendritic and axonal mass rather than in glial or

Fig. 5. Neuronogenesis in the PSN. In Ct-treated rats, small neurons (arrows) were born on G14 (A) and the large neurons (arrows) were born on G15 (C). Note that some large neurons were lightly labeled (arrow- heads) by an injection of ['Hlthymidine on G14. Presumably this label- ing resulted from the incorporation of [3H]thymidine on G14 into the precursors of the large neurons and the radiolabel subsequently was diluted as the precursor cells passed through an additional cycle(s) of the mitotic cycle before migrating to the PSN on G15. In Et-treated rats,

small (B) and large (D) neurons (arrows) were generated on G14 and G16, respectively. Large neurons were only rarely lightly labeled by injections of ['Hlthymidine on G14. Hence, during the 2-day hiatus (be- tween G14 and ClS), the radiolabel was diluted to levels below back- ground while the percursors passed through multiple cycles of the mitotic cycle. Note that no glia in Ct- and Et-treated rats were heavily labeled by injections on G15 or earlier.

M.W. MILLER AND S.J. MULLER 578

12

6

0

12

6 C D U aJ

L a2

u1

(Y

0 12

0

0 0

a 0 - - .- L

5 z 6

0

1 1

6

C

I I I I

G 1 2

- G 1 5

A I I I L 0 2 0 0 400 601

I I I I

G 1 6

I I I I

G 1 7

- G 1 9

k.. - G2 1

J I I 1 0 200 400 600 - C o n t r o l

Area (prn') 0 0 E t h a n o l

Fig. 8. Size distribution of neurons generated daily. Each box describes the size distribution of cells generated on a specific day of ges- tation.

blood vascular mass (Al-Rabiai and Miller, '87; Miller et al., '88; Potempa and Miller, submitted).

In Ct-treated rats, the time of origin of most PSN neurons is G12, G13, G14, or G15. These data on PSN neuronogene- sis agree with those of Nornes and Morita ('79), who exam- ined Ch-treated rats by use of a paradigm similar to that applied in the present study. On the other hand, the present data differ with those of Altman and Bayer ('80) who showed that the onset, peak, and end of neuronal generation in Ch-treated rats are on G13, G14, and G17, respectively, or one to two days later than shown in the present study. This discrepancy may result from the use of different strains of rats. For example, the gestation period for chow-fed Pur- due-Wistar rats was 23 days (Bayer, '80); however, for the Ch-and Ct-fed Long-Evans rats examined in the present study the gestation period was 22 days. Interestingly, the time of origin of PSN neurons was similar in Et-treated rats of the present study and in the Ch-treated rats of Altman and Bayer ('80), and in both of these groups of rats, the ges-

tation period was 23 days. Alternatively, the discrepancy may result from methodological differences. Altman and Bayer injected their rats between 9:00 and 11:OO A.M.; they deduced neuronal times of origin from rats given multiple injections; and any neuron with greater than background levels of silver grains over its nucleus was counted. In the present study, however, rats were injected at 2:OO-3:00 P.M.; each rat was given only one injection; and only heavily labeled cells were counted.

On the other hand, the results of the present study concur with the data of Altman and Bayer ('80) in that most small neurons are generated (between GI2 to G14) before the few large neurons which are scattered in the dorsal PSN (on G15). Such a sequence is atypical for the development of structures in the central nervous system (Jacobson, '78); however, it is similar to some of the patterns in neocortex where, for example, small layer VI neurons are generated before the large neurons in layer V and yet smaller neurons in layer IV are born before the relatively larger neurons in layer II/III (Miller, '85, '88b, c; Peduzzi, '88).

Prenatal exposure to ethanol alters the histogenesis of the PSN. The depression and delay of neuronogenesis may reflect profound changes in the proliferation of PSN precur- sors. PSN neurons are produced in a single germinal zone, the rhombencephalic ventricular zone. Proliferative activity in this zone, as indicated by the proportion of cells labeled by an injection of [3H]thymidine and the thickness of the ventricular zone, appears to be decreased in Et-treated rats (unpublished results). Parallel changes in neuronal genera- tion and cell proliferation have also been described for neo- cortex (Miller, '86, '88b, submitted). Neocortical neurons are derived from two germinal zones, the ventricular and subventricular zones. Proliferative activity is depressed in the ventricular zone. This zone is most prominent early in the generation period, and there is a correlated reduction in the numbers of early generated neurons in ethanol-exposed rats. Thus, these neocortical data are quite similar to those describing the depressive effects of ethanol on the develop- ment of the PSN. The correlation between changes in cell proliferation and neuronal generation is further supported by data on the subventricular zone. In the subventricular zone, proliferative activity is significantly greater in etha- nol-treated rats than in controls. Moreover, this zone is most prominent late in the generation period during an eT- induced surge in neuronal generation. Neurons in the PSN are derivatives of a single proliferative zone, the ventricular zone.

I t was estimated that the duration of the S-phase of the cell cycle of PSN precursors was 15% longer in Et-treated rats. The accuracy of the estimates of the absolute duration of the S-phase, however, depends on the methodological limitations in interpreting autoradiographs (Rogers, '73; Miller, '85; Miller and Nowakowski, '88). For example, the number of neurons counted is based on an analysis of the full 10-pm-section thickness. On the other hand, because the beta emissions from tritium traverse only a limited dis- tance through the tissue, only radiolabeled nuclei within the top 2-3 Nm of the surface can be discriminated as heavily labeled. Hence the estimates of the duration of the S-phase must be multiplied by three to five to account for this con- founding. Regardless, the same criteria were used for both the Ct-treated rats and for the Et-treated rats, and it appears that the duration of the S-phase of rhomben- cephalic ventricular cells is increased by prenatal exposure to ethanol. It cannot be ruled out that the duration of the

EFFECTS OF ETHANOL ON NEURONAL GENERATION 579

Fig. 9. Late-generated neurons. During the period from G16 to G22, a few neurons (arrows) and glia (arrows marked with an “a” or an “0”

designate an astrocyte or an oligodendrocyte, respectively) were born. Ct-treated rats (A and C) and Et-treated rats (B and D).

cell cycle is not affected by the ethanol exposure, or that ethanol reduces the number of young neurons which leave the ventricular zone. A recent study on the effects of ethanol on cycling hepatocytes demonstrates that ethanol alters the duration of the S-phase and other phases of the cell cycle (Higgins, ’87).

The alteration of the cell cycle by ethanol may initiate a cascade of defects in the development of PSN neurons that ultimately leads to an abnormally organized PSN. The increase in the cell cycle may underlie the ethanol-induced decrease in cell proliferation, which in turn leads to a decrease in the numbers of neurons generated, which neces- sitates a compensatory growth in the processes associated with the surviving PSN neurons. Thus, a primary, and cru- cial effect of ethanol on the developing nervous system may

be its effect on the proliferation of neuronal precursors. The numbers of neurons in a structure in the nervous system are simply the reflection of the proliferation in the germinal zone. Factors such as the numbers of cells passing through the cell cycle and the numbers of cells leaving the cycling population and migrating away from the ventricular zone are involved in determining the number of surviving neu- rons. These issues have yet to be addressed.

Prenatal exposure to ethanol also may affect the amount of neuronal death. The difficult issue of assessing the effects of ethanol on neuronal death has been addressed for cere- bellar Purkinje cells (Cragg and Phillips, ’85). Rats prena- tally exposed to ethanol had 63% fewer Purkinje cells on P5 than pair-fed controls. Moreover, during the perinatal pe- riod, the number of cells with pyknotic nuclei was substan-

580

tially higher in ethanol-exposed rats. Comparable effects of ethanol on the PSN would account for the differences in neuronal generation and in the complement of PSN neurons surviving in the adult.

ACKNOWLEDGMENTS We thank Lisa Modarressi for her expert technical assis-

tance. This research has been funded by PHS grants AA 06916, AA 07568, and DE 07734.

M.W. MILLER AND S.J. MULLER

Miller, M.W. (1986) Effects of alcohol on the generation and migration of cerebral cortical neurons. Science 233:1308-1311.

Miller, M.W. (1987) Effect of prenatal exposure to alcohol on the distribution and time of origin of corticospinal neurons in the rat. J. Comp. Neurol. 257:372-382.

Miller, M.W. (1988a) Development of projection and local circuit neurons in rat neocortex. In E.G. Jones and A. Peters (eds): Cerebral Cortex. '401.7. Plenum, New York pp. 133-175.

Miller, M.W. (1988b) Effect of prenatal exposure to ethanol on the develop- ment of cerebral cortex: I. Neuronal generation. Alcohol. Clin. Exg. Res. 12:440-449.

Miller, M.W. (1988~) Maturation of rat visual cortex. IV. Generation, migra- tion, morphogenesis and connectivity of atypically oriented pyramidal neurons. J. Comp. Neurol. 274:387405.

Miller, M.W. (submitted) Effect of prenatal exposure to ethanol on the devel- opment of cerebral cortex. 11. Cell proliferation in the ventricular and subventricular zones of the rat. J. Comp. Neurol.

Miller, M.W., L. Klem, N.L. Chiaia, and R.W. Rhoades (1988) Intracellular study of corticospinal projection neurons in rats prenatally exposed to ethanol. Neurosci. Abs. 14:loOl.

Miller, M.W., and R.S. Nowakowski (1988) Use of bromodeoxyuridine- immunohistochemistry to examine the proliferation, migration, and time of origin of cells in the central nervous system. Dev. Brain Res. 457:44 52.

Nornes, H.O., and M. Morita (1979) Time of origin of the neurons in the cau- dal brain stem of rat. An autoradiographic study. Dev. Neurosci. 2:lOl-- 114.

Parpara-Nicholson, D., and I.R. Telford (1957) Effects of alcohol on repro.. duction and fetal development in the Guinea pig. Anat. Rec. 1%7:438- 439.

Peduzzi, J.D. (1988) Genesis of GABA-immunoreactive neurons in the ferret visual cortex. J. Neurosci. 8:920-931.

Potempa, G., and M.W. Miller (submitted) Number of cells in mature rat somatosensory cortex: Effects of prenatal exposure to ethanol.

Randall, C.L., W.J. Taylor, and D.W. Walker (1977) Ethanol-induced malfor- mations in mice. Alcoholism 1:219-224.

Rogers, A.W. (1973) Techniques of Autoradiography. New York Elsevier. Sauer, F.C. (1936) The interkinetic migration of embryonic epithelial nuclei.

J. Morphoi. 6O:l-11. Schapiro, M.B., N.P. Rosman, and T.L. Kemper (1984) Effects of chronic

ethanol exposure on the developing brain. Neurobehav. Toxicol. Teratol. 6:351-356.

Shaywitz, S.E., D.J. Cohen, and B.A. Shaywitz (1980) Behavior and learning difficulties in children of normal intelligence born to alcoholic mothers. d. Pediatr. 96:97a982.

Sidman, R.L. (1970) Autoradiographic methods and principles for study of the nervous system with thymidine-H3. In W.J.H. Nauta and S.O.E. Ebbesson (eds): Contemporary Research Methods in Neuroanatomy. Springer-Verlag, New York pp. 252-274.

Smolen, A.J., L.L. Wright, and T.J. Cunningham (1983) Neuron numbers in the superior cervical sympathetic ganglion of the rat: A critical compari- son of methods for cell counting. J. Neurocytol. 12739-750.

Stoltenburg-Didinger, G., and H.L. Spohr (1983) Fetal alcohol syndrome and mental retardation: Spine distribution of pyramidal cells in prenatal alco- hol-exposed rat cerebral cortex, a Golgi study. Dev. Brain Res. 11:119- 123.

Streissguth, A.P. (1986) The behavioral teratology of alcohol: Performance, behavioral, and intellectual deficits in prenatally exposed children. tn J.R. West (ed): Alcohol and Brain Development. Oxford, New York pp. 3-44.

Sulik, K.K., M.C. Johnston, and M.A. Webb (1981) Fetal alcohol syndrome: Embryogenesis in a mouse model. Science 214t936-938.

Tze, W.J., and M. Lee (1975) Adverse effects of maternal alcohol consump- tion on pregnancy and fetal growth in rats. Nature 257:479480.

Vaughan, D.W. (1984) The structure of neuroglial cells. In E.G. Jones and A. Peters (eds): Cerebral Cortex. Vol. 2. Plenum, New York pp. 285-329.

Waechter, R.V., and B. Jaensch (1972) Generation times of the matrix cells during embryonic brain development An autoradiographic study in rats. Brain Res. 46t235-250.

LITJ3R.ATURE CITED Abel, E.L., B.T. Sherwin, and S. Jacobson (1983) In utero alcohol exposure:

Functional and structural brain damage. Neurobehav. Toxicol. Teratol. 5,363-366.

Al-Ghoul, W., and M.W. Miller (1989) Transient expression of Alz-50-immu- noreactivity in developing rat neocortex: A marker for naturally occurring neuronal death? Brain Res. 481 :361-367.

Al-Ghoul, W., and M.W. Miller (in preparation) Migration of neurons to the principal sensory nucleus of the trigeminal nerve.

Al-Rabiai, S., and M.W. Miller (1987) Ultrastructural study of layer V in somatosensory cortex of mature rats prenatally exposed to ethanol. SOC. Neurosci. Abs. 13:692.

Altman, J., and S.A. Bayer (1980) Development of the brain stem in the rat. IV. Thymidine-radiographic study of the time of origin of neurons in the pontine region. J. Comp. Neurol. 194:905-929.

Angevine, J.B., Jr., and R.L. Sidman (1961) Autoradiographic study of cell migration during histogenesis of cerebral cortex in the mouse. Nature 192.76CL-768.

Atlas, M., and V.P. Bond (1965) The cell generation cycle of the eleven-day mouse embryo. J. Cell Biol. 26:19-24.

Bayer, S.A. (1980) Quantitative 3H-thymidine radiographic analysis of neuro- genesis in the rat amygdala. J. Comp. Neurol. 194:845-875.

Berry, M., and A.W. Rogers (1965) The migration of neuroblasts in the devel- oping cerebral cortex. J. Anat. 99:691-709.

Clark, H.C. (1977) A Light and Electron Microscopic Analysis of Develop- ment Within the Main Sensory Trigeminal Nucleus of the Rat. Doctoral Dissertation, Virginia Commonwealth Univ., Richmond, VA.

Clarke, P.G.H. (1985) Neuronal death in the development of the vertebrate nervous system. Trends Neurosci. 8:345-349.

Clarren, S.K., and D.W. Smith (1978) The fetal alcohol syndrome. N. Engl. J. Med. 298,1063-106'7.

Cragg, B., and S. Phillips (1985) Natural loss of Purkinje cells during develop- ment and increased loss with alcohol. Brain Res. 325:151-160.

Dim, J., and H.H. Samson (1980) Impaired brain growth in neonatal rats exposed to ethanol. Science 208:751-753.

Edwards, H.G., and Dow-Edwards (submitted) Craniofacial alterations in- duced by prenatal ethanol exposure in the rat. Teratology.

Erzurumlu, R.S., and H.P. Killackey (1983) Development of order in the rat trigeminal system. J. Comp. Neurol. 223:124-137.

Floderus, S. (1944) Untersuchungen uber den Bau der menschlichen Hypo- physe mit besonderer Beruck sichtigung der qualitativen mikromorpho- logischen Verhaltnisse. Acta Pathol. Microbiol. Scand. 53:l-276.

Hamburger, V., and R.W. Oppenheim (1982) Naturally occurring neuronal death in vertehrates. Neurosci. Comment. 1:441363.

Hicks, S.P., and C.J. D'Amato (1968) Cell migration to the isocortex in the rat. Anat. Rec. 160:619434.

Higgins, P.J. (1987) Cell cycle phase-specific perturbation of hepatic tumor cell growth kinetics during short-term in vitro exposure to ethanol. Alco- hol. Clin. Exp. Res. llr550-555.

Jacobson, M. (1978) Developmental Neurobiology. Plenum, New York. Meyer, L.§., and E.P. Riley (1986) Behavioral teratology. In E.P. Riley and

C.V. Voorhees (eds): A Handbook of Behavioral Teratology. Plenum, New York pp. 101-140.

Miller, M.W. (1985) Cogeneration of retrogradely labeled corticocortical pro- jection and GABA-immunoreactive local circuit neurons in cerebral cor- tex. Dev. Brain Res. 23:187-192.