Male zebra finches are most apt to mimic songs heard between posthatch days (PHD) 35 and 65, and this vocal learning depends, in part, on the activation of N-methyl-D-aspartate receptors (NMDAR) within a discrete forebrain circuit that includes the lateral magnocellular nucleus of the anterior neostriatum (lMAN) and area X. Using in situ hybridization, we show that transcripts for both the constitutive NMDAR subunit NR1 and the modulatory subunit NR2B decrease abruptly in the lMAN between PHD20 and 40. This downregulation corresponds to the onset of song learning and a transition from slow to faster NMDAR currents in lMAN neurons. In area X, NR1 mRNA increases as NR2B mRNA decreases during song development. To understand how these changes in NMDAR mRNA might regulate song learning, we next investigated how manipulations that influence song development affect NMDAR mRNA expression. Early isolation from conspecific song (which delays closure of the sensitive period for song learning) selectively increases NR2B, but not NR1 mRNA, within lMAN at PHD60. In contrast, exposure to testosterone beginning at PHD20 (which impairs song development and hastens the developmental transition to faster NMDAR current kinetics within lMAN) accelerates the decline in NR2B mRNA in lMAN, again without affecting NR1 transcript levels. Neither manipulation significantly effects NR1 or NR2B mRNA levels in area X. Our data suggest that developmental changes in the expression of specific NMDAR subunits may regulate periods of neural and behavioral plasticity and that flexibility in the timing of these sensitive periods may be achieved through experience and/or hormone-dependent modulation of NMDAR gene expression.

Juvenile zebra finches (Poephila guttata) learn song in two stages: during sensory acquisition, they memorize the song of an adult tutor, and during sensorimotor learning, they alter their vocalizations to match the stored song model. Like many other forms of neural plasticity and memory formation, vocal learning in zebra finches is impaired by pharmacological blockade of NMDA receptors, but the relevant NMDA receptors have not yet been localized. During song development, one neural region that has been implicated specifically in song learning, the IMAN, exhibits an increased density of NMDA receptors as well as decreased binding affinity for the NMDA antagonist MK-801. To test the hypothesis that sensory acquisition requires activation of NMDA receptors in or near the IMAN we infused the NMDA receptor antagonist amino-5-phosphonopentanoic acid (AP5; 2.5 micrograms 0.1 microliter) directly into the anterior forebrain. Birds receiving AP5 infusions prior to each of 10 tutoring sessions copied significantly less of their tutor's song than did sham-operated birds, saline-infused birds, birds that received AP5 infusions on nontutoring days, or birds that received AP5 infusions into the cerebellum. Furthermore, infusions of AP5 in the anterior forebrain did not impair young birds' ability to discriminate zebra finch from canary song. These findings are consistent with the hypothesis that NMDA receptor activation in the anterior forebrain is necessary for the memorization of song material during avian vocal learning. This is also the first report that song-related regions of the anterior forebrain contribute to sensory acquisition specifically.

The development of estrogen receptors was studied in the preoptic area/anterior hypothalamus (POA/AH) of fetal male and female ferrets. In males this region includes a nucleus (MN-POA/AH), delineated by Nissl stains, which is not discernible in females. The results reveal the distribution of estrogen receptor containing cells during the period when estrogen is known to induce the differentiation of the male ferret's MN-POA/AH. Brains were taken from ferret kits on days 30, 34, 37 and 40 of a 41-42 day gestation, and were processed utilizing the H222 monoclonal antibody to reveal estrogen receptors. At E30 there were numerous H222 immunoreactive (ir) cells in central regions of the POA/AH. From E30 to E40 there was a striking increase in the number of H222ir cells in the POA/AH. A broad sweep of H222ir cells extended from the ventral POA dorsally and laterally into the caudal POA and AH of both males and females. H222ir cells were not restricted to the region of the MN-POA/AH at any fetal age. H222 immunoreaction product at E30 was restricted to nuclear compartments. By E40, H222ir processes extended from some cells with H222ir nuclei in the medial and lateral POA/AH in both males and females. At the older fetal ages immunopositive cell numbers increased in lateral positions. At E34 and E37 (but not E30) selective ventricular zones, and regions between the hypothalamus and amygdala contained H222ir cells, suggesting the presence of estrogen receptors in cells during migration. Although the amygdala contained a few H222ir cells as early as E34, the cortex lacked H222ir cells even as late as E40. The appearance of H222ir cells in positions suggestive of migration is consistent with the hypothesis that estrogen receptors play some role in determining cell positions in certain regions of the developing nervous system.

We examined whether testosterone (T) administered to female ferrets neonatally - a treatment known to enhance masculine coital capacity - induces formation of the sexually dimorphic male nucleus in the dorsal preoptic/anterior hypothalamic area (MN-POA/AH), and/or sensitizes dorsal POA/AH neurons to the stimulatory effect of later androgen treatment on somal dimensions. In males, the MN-POA/AH was present in all subjects, and exposure to androgen following castration at postnatal day 56 (P56) increased both MN-POA/AH volume as well as mean somal areas of MN-POA/AH neurons relative to oil-treated controls. Females given androgen from P5 to P20 and for one month beginning after ovariectomy on P56 failed to develop the MN-POA/AH, but displayed high levels of masculine sexual behavior. Somal areas of dorsal POA/AH neurons in females that received either T or a control neonatally did not increase following androgen treatment at P56. Thus, the correlation that exists between somal enlargement of dorsal POA/AH neurons and masculine sexual behavior in androgen-treated males is not found in behaviorally masculinized females. Masculine coital ability does not appear related to aspects of dorsal POA/AH morphology, supporting data from a previous study in which lesions of the MN-POA/AH caused negligible deficits in masculine sexual behavior of adult male ferrets.

A sexually dimorphic nucleus exists in the dorsal region of the ferret preoptic/anterior hypothalamic area (POA/AH), and is called the male nucleus of the POA/AH (MN-POA/AH) because it is found only in males. Development of the MN-POA/AH was studied in male ferrets, and for comparison a sexually nondimorphic ventral POA/AH nucleus was studied in both sexes. The MN-POA/AH was conspicuous in males as early as embryonic day 37 (E37) of a 41-day gestation, and its volume increased until postnatal day 56 (P56). No nucleus was present in the dorsal POA/AH of females at any age. The densities and average somal areas of cells in the dorsal POA/AH were similar in males and females at E33, before the MN-POA/AH could be visualized. However, at E37 and E41 dorsal cells were greater in density and/or somal area in males than in females, accounting for the appearance of a nucleus in males at these ages. To insure that the dorsal POA/AH nucleus seen in males at E37 and E41 was the presumptive MN-POA/AH present in adult males, pregnant ferrets were given progesterone and either implanted subcutaneously (s.c.) with testosterone (T) or ovariectomized and implanted s.c. with the aromatase inhibitor, 1,4,6-androstatriene-3,17-dione (ATD), on day 30 of gestation. As predicted from previous studies in which subjects were sacrificed in adulthood, formation of a dorsal POA/AH nucleus was promoted in female ferrets by T, and blocked in males by maternal ovariectomy and ATD treatment for animals sacrificed at E41. Much evidence suggests that behavioral sexual differentiation is accomplished in the male ferret between age E28 and P20. The MN-POA/AH is present and potentially functional in males during a considerable portion of this perinatal period.