Neurogenesis of the hamster suprachiasmatic nucleus

192

Brain Research, 519 (1990) 192-199 Elsevier

BRES 15565

Neurogenesis of the hamster suprachiasmatic nucleus Fred C. Davis, Richard Boada* and John LeDeaux** Department of Biology, Northeastern University, Boston, MA 02115 (U.S.A.) (Accepted 28 November 1989) Key words: Suprachiasmatic nucleus; Neurogenesis; Circadian rhythm; Hypothalamus; Autoradiography; Hamster

Neurogenesis of the hypothalamic suprachiasmatic nucleus (SCN) was described in the Syrian hamster (Mesocricetus auratus) using tritiated [3H]thymidine autoradiography. Pregnant hamsters were given single intraperitoneal injections of [3H]thymidine at different times during prenatal development, and labeled cells were analyzed in the offspring of 4-5 weeks of age. Cells of the hamster SCN became postmitotic (were 'born') over two and a half days from 10.5 to 13.0 days postfertilization (dpf) with a peak around 11.5 dpf, 4 days before birth. Two gradients in SCN neurogenesis were observed. Posterior cells were produced somewhat earlier than anterior cells and ventrolateral cells were produced before dorsomedial cells. An exception to the second gradient was a small population of ventrolateral cells produced near the end of SCN neurogenesis. The pattern of SCN neurogenesis in the hamster was similar to that described in the rat, including a predominant ventrolateral to dorsomedial gradient and the presence of ventral or ventrolateral cells produced relatively late, contrary to the predominant gradient. INTRODUCTION The suprachiasmatic nucleus (SCN) of the hypothalamus plays a central role in the regulation of vertebrate circadian rhythms. It is a cell-dense area bilaterally represented near the midline of the hypothalamus near the floor of the third ventricle just above the optic chiasm. The SCN appears to be specialized to receive photic information, to generate circadian oscillations, and to regulate rhythmicity in diverse behavioral and neuroendocrine functions 19'24. Explants or slices of the brain that contain the SCN generate circadian rhythms when maintained in vitro 13A5 and transplants of tissue containing the SCN restore rhythmicity in arrhythmic hosts 16. Despite extensive information about the neuroanatomical organization of the SCN 4'27'28, the neural organization or the specific cells that are critical for the generation of rhythmicity are not known. One approach to understanding the relationship between neural organization and circadian function is an analysis of this relationship during development. Development studies have found that much of the organization characteristic of the adult SCN is not required for the generation and entrainment of circadian oscillations. In rodents, the generation of circadian oscillations appears to begin before birth 8A2A7,21 even though in rats, most synaptogenesis, development of neuronal spiking, and appear-

ance of neuropeptide occurs postnatally ~'2°'25. Whether or not the initiation of circadian oscillations is related to a specific aspect of SCN development that occurs during prenatal development is not known. Addressing this question requires detailed information about the formation and early development of the SCN as well as precise determination of the first expression of circadian oscillations. Research on the development of hamster circadian rhythms indicates that circadian oscillations begin in the fetus as early as day 14 of gestation 9. Timed injections of the pineal gland hormone, melatonin during the last 4 days of gestation (days 12-15) set the phase of the pups' postnatal behavioral rhythms 1°. Whether or not the effect of melatonin is direct or indirect, the results indicate that the fetal circadian pacemaker is highly sensitive to an entraining signal early in development. Crossland and Uchwat 6 found that neurogenesis of the hamster SCN occurs between days 10 and 13 of gestation, suggesting that entrainment of a fetal circadian pacemaker occurs during the first 2 or 3 days of postmitotic differentiation of SCN neurons. Because the timing of SCN development is critical for the interpretation of prenatal entrainment experiments the present study was undertaken to confirm the timing of SCN neurogenesis and to determine if the timing is different for different regions within the SCN.

* Present address: Curry School of Education, Univ. of Virginia, Charlottesville, VA 22903, U.S.A. ** Present address: Dept. of Biology, Massachusetts Institute of Technology, Cambridge, MA 02139, U.S.A. Correspondence: EC. Davis, Dept. of Biology, 414 Mugar Life Science Building, Northeastern Univ., Boston, MA 02115, U.S.A. 0006-8993/90/$03.50 (~) 1990 Elsevier Science Publishers B.V. (Biomedical Division)

193

B

V3 50pm 50 pm

II

\

\

)

SS~.

~J

PERCENT LABELEDCELLS

[]

o+

]

I-I0%

]

11-20%

[] No Data

21-30 %

31-40 % >40%

Fig. 1. A: illustration of the grid used to count cells within the suprachiasmatic nucleus (SCN). The grid is superimposed on a generalized drawing of a coronal section through the middle of the SCN on one side of the brain. V3, third ventricle; OC, optic chiasm. B: the key for levels of shading used to indicate the percent labeled cells within each square in Fig. 4. Neurogenesis of the h a m s t e r SCN was examined using tritiated thymidine a u t o r a d i o g r a p h y . Fetuses were exp o s e d to thymidine at different times in d e v e l o p m e n t and allowed to survive b e y o n d weaning before the presence and location of l a b e l e d cells was autoradiographically d e t e r m i n e d . Previous studies of neurogenesis have established that cells labeled by tritiated thymidine at a particular time in d e v e l o p m e n t were, at the time of exposure, neuronal precursors that had not yet started postmitotic differentiation. Once cells begin postmitotic differentiation (are ' b o r n ' ) they no longer incorporate thymidine 26. In the present study, the timing of the transition from being labeled to being unlabeled was described for cells of the h a m s t e r SCN, providing a detailed description o f both the spatial and t e m p o r a l p a t t e r n of SCN neurogenesis. The timing of SCN neurogenesis was, as a whole, similar to that r e p o r t e d earlier 6, and the p a t t e r n of neurogenesis for different subregions was similar to that r e p o r t e d for the rat SCN 3. T h e results provide critical information about the timing

of SCN d e v e l o p m e n t in the h a m s t e r including information about the timing of different and possibly functionally distinct subregions within the SCN. A preliminary r e p o r t of these results was previously published 7. MATERIALS AND METHODS Syrian hamsters (Mesocricetus auratus) from Charles River Laboratories were maintained on a 14:10 light:dark cycle (0600-2000 EST) with food (Purina 5001) and water continuously available. Females were housed with males on the night of ovulation and fertilization was assumed to have occurred at midnight (24.00 h). Each pregnant female received a single intraperitoneal injection of tritiated thymidine (thymidine, [methyl-all], 50-90 Ci/mmole, New England Nuclear) in a dose of 7/~Ci/gram b. wt. Injections were given at 7 different times in gestation (noon or midnight) from 10.5 to 13.5 days posffertilization (dpf). Birth occurred between 15 and 16 dpf and pups were weaned 3 weeks after birth. At 4 to 5 weeks of age pups were deeply anesthetized with sodium pentobarbital and intracardially perfused with 0.9% saline followed by 10% buffered

Anterior

15'

Percent 10' Labeled 5

0

10.511.011.512.012.513.013.5

Middle

20tdtL.__ 151t 15

Percent 10 Labeled

5,

0

10,511.0 11.512.0 12.513.013.5

Posterior

Percent 1 0 Labeled 5 0

10.511.011.512.012.513.013.5

Days Postfertilization Fig. 2. Examples of labeled cells within the SCN. Arrowheads of decreasing size indicate heavily labeled, lightly labeled, and unlabeled cells. Scale bar, 10/~m; V3, third ventricle.

Fig. 3. Percent of SCN cells labeled at different times in prenatal development. The percent of heavily (black bars) and lightly (gray bars) labeled cells are indicated for each age at 3 levels within the SCN along the anterior-posterior axis.

194 formalin. Brains were removed and kept in Bouin's fixative for 24-36 h and then stored in 10% buffered formalin. Thirty-two brains (male and female) from 15 litters were analyzed. Except for 13.5 (one litter) each treatment age was represented by 2 or 3 litters. The brains were prepared for autoradiography by embedding them in paraffin (Paraplast, Fisher Scientific) and cutting serial 10-/~m coronal sections through the anterior-posterior extent of the suprachiasmatic nucleus (SCN). Sections were mounted on albumincoated slides and deparaffinized and rehydrated before dipping in Kodak Autoradiography Emulsion, Type NTB2. After dipping, slides were allowed to dry at room temperature for 24 h and then stored at 4 °C for 5 weeks. Slides were developed for 2 min at 17 °C (Kodak Dektol), stained with thionine, and cover-slipped using Permount (Fisher Scientific). The numbers and locations of autoradiographically labeled cells were quantified by first determining the anterior and posterior limits of the SCN for a particular brain and then selecting 3 representative sections at the anterior and posterior ends and at the middle of the SCN. The anterior and posterior sections were located in from the ends of the SCN approx. 1/8 of the total anterior-posterior extent. The left SCN of each section was analyzed at a magnification of 200x, allowing the entire SCN to be viewed within a single field and at the same time allowing individual cells to be counted. An eye-piece reticule grid of 50 x 50 a m squares was superimposed on the section. The grid matched a grid on paper viewed through a drawing tube. The grid was aligned on the SCN as shown in Fig. 1A and the cells within each square were counted. A 9 × 10 grid was used for anterior and posterior sections and a 9 x 11 grid was used for the middle section. On the drawing paper, the numbers of unlabeled, lightly labeled, and heavily labeled cells were written in each square. The values for each square within each grid were entered into a computer spreadsheet so that the values for each class in each square could be summed within a treatment group. Using these sums the percent cells in each class were calculated for each square. Because each square represented approx, the same anatomical region in different animals it was possible to combine the values for several animals in order to produce a quantitative map based on a group rather than on an individual. In addition to maps of labeled cells, the percentage of each cell class in an entire grid was calculated. 10.5

ANTERIOR

MIDDLE

11.0

11.5

IB

Cells were classified as either unlabeled, lightly labeled, or heavily labeled according to the number of silver grains above the cell. The background number of silver grains was low; areas between cells approx, equal to the area of a cell rarely contained silver grains. Although even two grains over a nucleus appeared to be greater than background, 5-10 grains were used as the criterion for lightly labeled and any number above 10 was considered heavily labeled. Fig. 2 shows examples of lightly and heavily labeled cells. The most heavily labeled cells were likely to have been in the last round of replication at the time of injection but this might also have been true for some lightly labeled cells as well. The amount of label could have been affected by the degree of overlap between replication and availability of labeled thymidine or by the position of the cell nucleus relative to the photographic emulsion layer. Following injections at the earliest age (10.5 dp0 many cells appeared to be labeled above background (2-4 grains) but were counted as unlabeled. To exclude some glia from the cell counts, small, uniformly staining cells were not counted. However, by the histological methods used, it was probably not possible to avoid including astroglia that were equal in size to SCN neurons 27.

RESULTS

Autoradiographic

analysis of [3H]thymidine incorpo-

r a t i o n at d i f f e r e n t t i m e s in d e v e l o p m e n t b y cells o f t h e h a m s t e r S C N s h o w e d t h a t t h e cells b e c a m e p o s t m i t o t i c b e t w e e n 10.5 a n d 13.0 d a y s p o s t f e r t i l i z a t i o n ( d p f ) . T h e d e v e l o p m e n t a l rise a n d fall in n u m b e r s o f l a b e l e d cells w i t h i n t h e S C N is s h o w n in Fig. 3. T h e rise o c c u r r e d e a r l i e r in t h e p o s t e r i o r

SCN

a n d t h e fall w a s m o r e

p r o l o n g e d in t h e a n t e r i o r S C N . T h r o u g h o u t t h e S C N , t h e rise in lightly l a b e l e d

cells p r e c e d e d

t h a t in h e a v i l y

l a b e l e d cells. T h i s r e l a t i o n s h i p is e x p e c t e d if at l e a s t a portion 12.0

o f lightly l a b e l e d 12.5

cells w e r e

13.0

lightly l a b e l e d 13.5

ii1

POSTERIOR I~1 [ [ I II 1/11

Fig. 4. Distributions of SCN cells labeled at different times in development. The days postfertilization when fetuses were exposed to [3H]thymidine are indicated at the top. The distributions of labeled cells are shown on coronal sections through the anterior, middle, and posterior SCN. The percent of cells labeled (heavy and light combined) in each square is indicated by different levels of shading according to the key in Fig. lB.

195 because they were in the next to last rather than the last round of D N A replication at the time of exposure to labeled thymidine. In addition to a posterior to anterior gradient in the genesis of SCN cells there was also a ventrolateral to

dorsomedial gradient. Fig. 4 shows that the first cells to be labeled were in the ventral or ventrolateral SCN while the last cells to be labeled were most c o m m o n in the dorsomedial SCN, close to the third ventricle. Anterior, middle, and posterior sections from a hamster exposed to

Fig. 5. Darkfield and brightfield photomicrographs of anterior (A,B), middle (C,D) and posterior (E,F) coronal sections through the SCN of a hamster exposed to [3H]thymidine at 11.5 days postfertilization. Scale bar, 200/~m; V3, third ventricle; OC, optic chiasm.

196

Fig. 6. Brightfield and darkfield photomicrographsof a coronal section through the mid SCN of a hamster exposed to [~thymidine at 12.5 days postfertilization. Labeled cells can be seen primarily at the dorsomedial and ventrolateral margins. Scale bar, 50/zm; V3, third ventricle; OC, optic chiasm. thymidine at 11.5 dpf are shown in Fig. 5. If it is assumed that the SCN developed with a posterior to anterior gradient in cell production (i.e. the posterior section was the 'oldest'), these sections illustrate the ventrolateral to dorsomedial gradient. In the posterior section, labeled cells were most common in the dorsomedial SCN while in the anterior section they were found throughout the SCN including the ventral portion. In addition to the two gradients just described, a small population of cells at the ventrolateral margin of the SCN were labeled relatively late. In Fig. 4 this group of cells can be seen most clearly in the mid-SCN following thymidine at 12.5 dpf. An example of the SCN from this group is shown in Fig. 6. In addition to labeled cells in the dorsomedial SCN, several cells on the ventrolateral margin were labeled. As noted earlier, it is likely that cell counts included some glia. Using Nissl-stained paraffin sections it is difficult to distinguish between astroglia and neurons in the SCN. It is likely that inclusion of astroglia increased the counts of unlabeled cells since most gliogenesis appears to occur after neurogenesis has ended 3. Some of the cells that were labeled late in SCN neurogenesis and that settled at the ventrolateral margin might have been

glia but most appeared to be neurons having large lightly staining nuclei with single prominent nucleoli. DISCUSSION The pattern of thymidine incorporation by cells of the hamster suprachiasmatic nucleus (SCN) shows that cells of the SCN begin postmitotic differentiation over two and a half days of prenatal development from 10.5 to 13.0 days postfertilization (dpf). This agrees well with an earlier description of hamster SCN neurogenesis by Crossland and Uchwat 6. They found SCN cells to be heavily labeled by thymidine on days 10 to 12.5 of gestation with a peak on day 11.5. Although we observed significant labeling as late as 13.0 dpf, the maximum labeling for the SCN as a whole was also around 11.5 dpf, 4 days before birth. Crossland and Uchwat 6 reported a lateral to medial gradient in neurogenesis of the hamster diencephalon but did not specifically describe gradients within the SCN. In the present study, two gradients of neurogenesis within the SCN were observed. Cells of the posterior SCN were labeled before anterior cells, and cells of the ventrolateral SCN were labeled before those of the dorsomedial SCN. An exception to the second

197 gradient was a small population of cells on the ventrolateral margin that became postmitotic at about the same time as cells at the dorsomedial margin. The pattern of SCN neurogenesis in the hamster was similar to that described for rats by Altman and Bayer2'3. They reported a ventrolateral to dorsomedial gradient with the exception of a population of cells along the ventral margin that were produced at the same time as dorsomedial cells3. They designated the late produced population of ventral cells as the basal suprachiasmatic subnucleus (SUb). They also noted a posterior to anterior gradient in neurogenesis of the dorsomedial SCN, similar to the posterior to anterior gradient we observed in the hamster. In both the rat and the hamster, SCN neurogenesis occurs several days before birth. Altman and Bayer2'3 found neurogenesis to occur between embryonic days 14 and 17 (13.5 to 16.5 dpf) with a peak at 14.5 dpf, 7 days before birth. In both rats and hamsters cells produced near the end of SCN neurogenesis settle along the margin of the SCN in locations contrary to the predominant ventrolateral to dorsomedial gradient. It appears that the 'core' of the SCN is produced first with a ventrolateral to dorsomedial gradient (as well as a posterior to anterior gradient) followed by cells that settle along the margin of the SCN, both the dorsomedial and ventral or ventrolateral margins. As noted by Altman and Bayer3, the ventrolateral to dorsomedial gradient in SCN neurogenesis means that the two major subdivisions of the SCN have distinct developmental histories. In both rats and hamsters, the SCN consists of dorsomedial and ventrolateral regions distinguished by immunohistochemical staining and extranuclear connectivity4'27'2s. The ventrolateral region is characterized by projections from the intergeniculate leaflet of the lateral geniculate nucleus immunoreactive for neuropeptide Y, by dense projections of the retinohypothalamic tract (RHT), by serotonergic projections, and by cells immunoreactive for vasoactive intestinal polypeptide (VIP). The dorsomedial SCN is characterized by cells immunoreactive for vasopressin and somatostatin. The pattern of neurogenesis in rats and hamsters suggests that functionally related cells are produced at the same time. Different times of neurogenesis for functionally distinct cell populations could be achieved in two ways, the sequential production of cells with different fates by the same pool of precursor cells or by the sequential activity of different precursor pools. Altman and Bayer 3 suggested that different but neighboring regions of the germinal neuroepithelium in the floor of the third ventricle give rise to the two subdivisions. They suggested that proliferative activity in the ventralmost neuroepithelium (giving tise to the ventrolateral SCN) precedes activity of slightly more dorsal neuroepithelium. As cells

of the dorsomedial SCN are produced some of the cells migrate to the ventral margin (SUb). Altman and Bayera were able to identify the origin of SCN cells because they examined labeled tissue at different intervals after exposure to [3H]thymidine. Similar studies that follow the fates of cells shortly after the cells are labeled have not been done in the hamster. At present there is no obvious neuroanatomical correlate of the late produced ventral cells (SUb) observed by Altman and Bayer3 or of the ventrolateral cells observed in the hamster. In the rat, the presence of the SUb means that it is uncertain when a particular cell type in the ventral SCN is produced. For example, cells immunoreactive for VIP might be produced at the beginning or at the end of SCN neurogenesis. Double labeling experiments will have to b e performed to determine with certainty when a particular cell type is produced. In the hamster, VIP cells are located in the ventral SCN 4 (personal observations) and do not appear to correlate with the late produced ventrolateral cells, suggesting that VIP cells are produced early in SCN neurogenesis. A general implication of different times of neurogenesis for functionally different cell populations is that the successful culture or transplantation of different cells might depend on the precise age of the fetus when tissue is obtained. In vivo labeling with thymidine followed by culturing of dispersed cells indicates that the cells which survive best are those that completed their terminal division within 24 h of dispersal 1. Research on the comparative anatomy of the mammalian SCN has revealed variation in the organization of different regions defined by the characteristics described above 5. For example, in the gray short-tailed opossum (Monodelphis domestica) cells immunoreactive for vasopressin and for VIP are co-distributed within the dorsomedial SCN. If the separation of cells immunoreactive for VIP and vasopressin in rats and hamsters is related to the ventrolateral to dorsomedial gradient of neurogenesis then such a gradient might not be observed in the opossum. Neurogenesis of the SCN in the gray shorttailed opossum occurs during the first postnatal week (2-3 weeks after fertilization) but whether or not there are gradients of neurogenesis within the SCN has not been reported 22. Whatever the mechanism by which the subregions of the rodent SCN differentiate the information for this appears to be contained within the anterior hypothalamus before neurogenesis of the SCN begins. Roberts et a l . 23 showed that anterior hypothalamus from day 13 rat embryos develop VIP and vasopressin subregions characteristic of the SCN even when transplanted to the anterior chamber of an adult eye. SCN subregions also differentiate when the fetal tissue (day 13-17) is minced prior to transplantation to a host brain 29. This

198 suggests that the segregation of subregions does not strictly d e p e n d on the spatial organization of the neuroepithelium even if specialized regions of the neuroepithelium give rise to the subregions. Analysis of SCN neurogenesis in rats, hamsters, and the gray short-tailed opossum suggests that the SCN is f o r m e d before circadian oscillations are generated by the central nervous system. In each case neurogenesis is c o m p l e t e before the earliest evidence for circadian rhythmicity. This generalization is tenuous since the absence of circadian rhythmicity early in d e v e l o p m e n t might be difficult to detect. In rats and the opossum the evidence for the initiation of circadian rhythmicity is based on m e a s u r e m e n t s of 2-deoxyglucose (2-DG) uptake by the SCN. In the rat, SCN neurogenesis is complete by day 172,3 and the rhythm in 2 - D G u p t a k e is not present until day 1921. In the opossum, SCN neurogenesis is complete by postnatal day 7 and a day/night difference in 2 - D G uptake is not apparent until day 2022. Because the 2 - D G m e t h o d depends on the presence of the SCN it cannot measure rhythmicity before the SCN is f o r m e d and might not be sensitive enough to detect rhythmicity g e n e r a t e d in some other p o p u l a t i o n of cells such as SCN precursors. In hamsters, the evidence for the initiation of circadian oscillations is based on the ability to set the phase of the oscillations during p r e n a t a l development. Evidence suggests that

phase can be set as early as day 14 of gestation 9. A t present, however, there is no compelling evidence that oscillations are or are not g e n e r a t e d earlier. The only evidence that oscillations might be g e n e r a t e d before the SCN is formed comes from experiments in rats in which the phase of corticosterone rhythms m e a s u r e d postnatally seems to have been set before SCN neurogenesis TM. Transplantation studies in rats have shown that circadian rhythms are g e n e r a t e d by an SCN graft that was transplanted during SCN neurogenesis TM. Because the rhythm was not m e a s u r e d until several weeks after transplantation this result shows that the transplanted SCN anlagen had the potential to develop circadian oscillations but does not show that the tissue generates oscillations at the time of transplantation. F u r t h e r work will be required to d e t e r m i n e if the generation of circadian oscillations is a p r o p e r t y specific to the SCN which arises during postmitotic differentiation or is a p r o p e r t y of SCN precursors or o t h e r cells of the developing nervous system.

REFERENCES

11 De Vries, G.J., Buijs, R.M. and Swaab, D,E, Ontogeny of the vasopressinergic neurons of the suprachiasmatic nucleus and their extrahypothalamic projections in the rat brain - - presence of a sex difference in the lateral septum, Brain Research, 218 (1981) 67-78. 12 Deguchi, T., Ontogenesis of a biological clock for serotonin acetyl coenzyme A N-acetyltransferase in pineal gland of rat, Proc. Natl. Acad. Sci. U.S.A., 72 (1975) 2814-2818. 13 Earnest, D.J. and Sladek, C.D., Circadian rhythms of vasopressin release from individual rat suprachiasmatic explants in vitro, Brain Research, 382 (1986) 129-133. 14 Earnest, D.J., Sladek, C.D., Gash, D.M. and Wiegand, S.J., Specificity of circadian function in transplants of the fetal suprachiasmatic nucleus, J. Neurosci., 9 (1989) 2671-2677. 15 Green, D.J. and Gillette, R., Circadian rhythm of firing rate recorded from single cells in the rat suprachiasmatic slice, Brain Research, 245 (1982) 198-200. 16 Lehman, M.N., Silver, R., Gladstone, W.R., Kahn, R.M., Gibson, M. and Bittman, E.L., Circadian rhythmicity restored by neural transplant. Immunocytochemical characterization of the graft and its integration with the host brain, J. Neurosci., 7 (1987) 1626-1638. 17 Hiroshige, T., Honma, K.I. and Watanabe, K., Prenatal onset and maternal modifications of the circadian rhythm of plasma corticosterone in blind infantile rats, J. Physiol. (Lond.), 325 (1982) 521-532. 18 Honma, S., Honma, K., Shirakawa, T. and Hiroshige, T., Maternal phase setting of fetal circadian oscillation underlying the plasma corticosterone rhythm in rats, Endocrinology, 114 (1984) 1791-1796. 19 Moore, R.Y., Organization and function of a central nervous system circadian oscillator: the suprachiasmatic hypothalamic nucleus, Fed. Proc., 42 (1983) 2783-2789.

1 Ahmed, Z. and Fellows, R.E., Determination of the birth date and proliferative state of dissociated cells from the fetal rat brain, Dev. Brain Res., 37 (1987) 77-87. 2 Altman, J. and Bayer, S.A., Development of the diencephalon in the rat. 1, Autoradiographic study of the time of origin and settling patterns of neurons of the hypothalamus, J. Comp. Neurol., 182 (1978) 945-972. 3 Aitman, J. and Bayer, S.A., The development of the rat hypothalamus, Adv. Anat. Embryol. Cell Biol., 100 (1986) 1-178. 4 Card, J.P. and Moore, R.Y., The suprachiasmatic nucleus of the golden hamster: immunohistochemical analysis of cell and fiber distribution, Neuroscience, 13 (1984) 415-431. 5 Cassone, V.M., Speh, J.C., Card, J.E and Moore, R.Y., Comparative anatomy of the mammalian hypothalamic suprachiasmatic nucleus, J. BioL Rhythms, 3 (1988) 71-91. 6 Crossland, W.J. and Uchwat, C.J., Neurogenesis in the central visual pathways of the golden hamster, Dev. Brain Res., 5 (1982) 99-103. 7 Davis, EC., Boada, R. and LeDeaux, J., Neurogenesis of the hamster suprachiasmatic nucleus, Soc. Neurosci. Abstr., 14 (1988) 50. 8 Davis, EC. and Gorski, R.A., Development of hamster circadian rhythms: prenatal entrainment of the pacemaker, J. Biol. Rhythms, 1 (1986) 77-89. 9 Davis, EC. and Gorski, R.A., Development of hamster circadian rhythms: role of the maternal suprachiasmatic nucleus, J. Comp. Physiol. A, 162 (1988) 601-610. 10 Davis, EC. and Mannion, J,, Entrainment of hamster pup circadian rhythms by prenatal melatonin injections to the mother, Am. J. Physiol., 255 (1988) R439-R448.

Acknowledgements. A portion of this work was performed in the Department of Biology at the University of Virginia. The authors thank Dr. Sat Bir S. Khalsa, Dr. Irwin R. Konigsberg, Mark Berryman, and Jane Mannion for technical assistance, and Dr. N. Viswanathan for critically reading the manuscript. The research was supported by NIH Grant HD18686 to EC.D.

199 20 Moore, R.Y. and Berns'tein, M.E., Synaptogenesis in the rat suprachiasmatic nucleus demonstrated by electron microscopy and synapsin 1 immunoreactivity, J. Neurosci., 9 (1989) 21512162. 21 Reppert, S.M. and Schwartz, W.J., The suprachiasmatic nuclei of the fetal rat: characterization of a functional circadian clock using 14C-labeled deoxyglucose, J. Neurosci., 4 (1984) 16771682. 22 Rivkees, S.A., Fox, C.A., Jocobson, C.D. and Reppert, S.M., Anatomic and functional development of the suprachiasmatic nuclei in the gray short-tailed opossum, J. Neurosci., 8 (1988) 4269-4276. 23 Roberts, M.H., Bernstein, M.E and Moore, R.Y., Differentiation of the suprachiasmatic nucleus in fetal rat anterior hypothalamic transplants in oculo, Dev. Brain Res., 32 (1987) 59-66. 24 Rosenwasser, A.M., Behavioral neurobiology of circadian pacemakers: a comparative perspective, Prog. PsychobioL Physiol.

Psychol., 13 (1988) 155-226. 25 Shibata, S. and Moore, R.Y., Development of neuronal activity in the rat suprachiasmatic nucleus, Dev. Brain Res., 34 (1987) 311-315. 26 Sidman, R.L., 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, 1970, pp. 252274. 27 Van den Pol, A.N., The hypothalamic suprachiasmatic nucleus of rat: intrinsic anatomy, J. Comp. Neurol., 191 (1980) 661-702. 28 Van den Pol, A.N. and Tsujimoto, K.L., Neurotransmitters of the hypothalamic suprachiasmatic nucleus: immunocytochemicai analysis of 25 neuronal antigens, Neuroscience, 15 (1985) 1049-1086. 29 Wiegand, S.J. and Gash, D.N., Organization and efferent connections of transplanted suprachiasmatic nuclei, J. Comp. Neurol., 267 (1988) 562-579.