INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

IN UTERO SWALLOWING contributes importantly to several critical developmental processes including the regulation of amniotic fluid volume and composition, the acquisition and potential recirculation of solutes, and the maturation of the fetal gastrointestinal tract. Human fetal swallowing has been demonstrated as early as 11 wk of gestation (4), with daily swallowing rates near term of 200-500 ml (1, 20). Similar to the human fetus, the near-term ovine fetus swallows 300-1,000 ml/day of amniotic fluid as well as additional pulmonary secretions (8, 30, 34).

Fetal swallowing may be influenced by several factors including fetal neurobehavioral state (9), fluid availability (13, 21), and dipsogenic factors (24, 25). In the near-term ovine fetus [the last 20% of gestation; term = 150 days of gestational age (dGA)], systemic and central osmotic-dipsogenic mechanisms are functional, because swallowing activity and arginine vasopressin (AVP) secretion are stimulated in response to intravenous or intracerebroventricular hypertonic saline (22, 27). However, the preterm ovine fetus (114 ± 1 dGA) does not consistently respond to osmotic stimuli (14). Because swallowing is the major route of amniotic fluid resorption (8, 30, 35), swallowing abnormalities may result in excess (polyhydramnios) or deficient (oligohydramnios) amniotic fluid volume, each of which is associated with significant perinatal morbidity and mortality.

Physiological and neuroanatomic studies in several adult species indicate that osmotic dipsogenic effects are mediated via sites within anterior circumventricular organs (ACVO). Because they have fenestrated capillaries, the ACVO are in a unique position in the central nervous system to function as receptor sites for transduction of blood-borne signals. Specifically, the organum vasculosum of the lamina terminalis (OVLT) and the subfornical organ (SFO), together with the median preoptic nucleus (MPON) (Fig. 1), are intimately involved in the control of extracellular fluid volume (for review, see Refs. 2 and 37). Fos, the protein product of the immediate-early gene c-fos, is an excellent marker of cellular activation in neuroendocrine systems (10, 11) and has been used successfully in adult rats to study ACVO responses to ANG II and carbachol (18, 29).

View larger version (78K): [in this window] [in a new window]   Fig. 1.   Cresyl violet stain of anterior circumventricular organs of a late-gestation fetal sheep showing relationships to other anterior brain structures. A: subfornical organ (SFO). B: median preoptic nucleus (MPON). C: organum vasculosum of lamina terminalis (OVLT). 3V, third ventricle; AC, anterior commissure; OC, optic tract; SON, supraoptic nucleus. Bar, 63.3 µm.

With the last one-third of gestation representing a critical period in the maturation of dipsogenic and AVP secretory mechanisms, we sought to evaluate fetal sheep ACVO and hypothalamic supraoptic (SON) and paraventricular nuclear (PVN) neuronal involvement in response to systemic hypertonicity. We examined Fos-like immunoreactivity after a mannitol infusion to the dam, which, as we have previously demonstrated, raises fetal plasma osmolality from 302 ± 2 to 315 ± 2 mosmol/kg (23).

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Six pregnant sheep (Rambouillet × Columbia) bearing ten fetuses (4 sets of twins) ranging in age from 129-131 dGA were utilized in the present study. Three ewes with a total of six fetuses were randomly assigned to the mannitol protocol, and three ewes (with a total of 4 fetuses) were assigned to the control group. All protocols were approved by the Cornell University Institutional Animal Care and Use Committee.

Instrumentation. Three days before osmotic challenge, indwelling catheters were surgically placed in the left jugular vein of all dams under halothane anesthesia as previously described (16). Catheters were maintained patent by filling with heparin (1,000 IU/ml; Elkins-Sinn, Cherry Hill, NJ). Postsurgically, all ewes were given ampicillin (Amp Equine, SmithKline Beecham, West Chester, PA) twice a day for 2.5 days.

Osmotic challenge. On the day of study, starting at time 0, mannitol (American Regent Laboratories, Shirley, NY) was delivered to the jugular vein of the dam as a 20% solution at a rate of 1 ml · min¹ · kg¹ maternal body weight for 10 min via the indwelling jugular catheter. The control animal received an equivalent volume of isotonic saline. At 75 min, the dam was anesthetized under halothane and a midline abdominal incision was made to gain access to the uterus. A hysterotomy was performed, and the fetal head was gently delivered from the uterus. A catheter was placed in one common carotid artery of the fetus. Immediately thereafter, the fetal head was perfused via the carotid artery with 500 ml normal saline containing heparin (10 IU/ml) and sodium nitrite (2% wt/vol), followed by 500 ml phosphate-buffered acrolein-paraformaldehyde (2.5:4% wt/vol; Sigma, St. Louis, MO). In twin pregnancies, the second twin brain was perfused ~5 min after the first.

Tissue processing. After perfusion, the fetal brains were removed, dehydrated in sucrose (30%), sliced at 30 µm on a Reichert sliding microtome, and stored at 20°C in cryoprotectant (36). Sections were immunohistochemically processed free floating using the avidin-biotin complex method (modified Elite kits, Vector Laboratories, Burlingame, CA) as previously described (16). Final dilutions of the primary antibodies were as follows: 1) Fos 1:200,000 (cat. no. PCO5; Oncogene Science, Uniondale, NY), 2) AVP 1:200,000 (generously donated by Dr. A. Robinson, UCLA, Los Angeles, CA); 3) oxytocin (OT) 1:100,000 (generously donated by Dr. Takashi Higuchi, Fukui, Japan). All tissues for each analysis were immunostained simultaneously.

Cell counts. Cell counts were made on sections at approximately the anterior-posterior midrange of each nucleus from each fetus by an observer who was unaware of the treatment groups. Counts were made at ×400 magnification on a television monitor that yielded a rectangular displayed field of 210 × 160 µm. The number of Fos-immunopositive nuclei were recorded from three selected fields from each nucleus. Fields were chosen randomly using a numbered grid system.

Statistical analysis. Fos counts per field were averaged to obtain a single independent number for each structure from each fetus. Averages for each nucleus were compared between treatment groups using Student's nonpaired t-test with the -level set at 0.05.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

The microscopic anatomy of the ACVO in relation to other anterior brain structures in a late-gestation fetal sheep is presented in Fig. 1. The average number of Fos-like immunoreactive neurons per field in the SFO, OVLT, SON, and PVN was increased (P < 0.05) in fetal sheep of dams given osmotic challenge compared with control fetuses (Figs. 2, 3, and 4). Although there was a tendency for Fos activation of neurons in the MPON to be increased in the fetuses of mannitol-treated dams, there was too much within-group variation in both treated and control fetuses for this result to be significant (Fig. 4). There were no within-group differences between twins and singles or between first- and second-perfused twins with respect to Fos immunostaining.

View larger version (125K): [in this window] [in a new window]   Fig. 2.   Fos-like immunoreactivity in neuronal nuclei of SFO (A and B) and OVLT (C and D) of treated (A and C) vs. control (B and D) fetal sheep brains collected 75 min after intravenous infusion (1 ml · min1 · kg1 for 10 min) of 20% mannitol or normal saline to dam at 130 days of gestation. Bars, 13.4 (A and B) and 33.6 µm (C and D).

View larger version (104K): [in this window] [in a new window]   Fig. 3.   Fos-like immunoreactivity in neuronal nuclei of hypothalamic paraventricular nuclei (PVN; A and B) and SON (C and D) of treated (A and C) vs. control (B and D) fetal sheep brains collected 75 min after intravenous infusion (1 ml · min1 · kg1 for 10 min) of 20% mannitol or normal saline to dam at 130 ± 1 days of gestation.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Average number of Fos-positive neuronal nuclei per field in SFO, MPON, OVLT, SON, and PVN of fetal sheep brains collected 75 min after intravenous infusion (1 mg · min1 · kg1 for 10 min) of 20% mannitol (solid bars; n = 6) or saline (open bars; n = 4) to dam at 130 ± 1 days of gestation. * Significantly different from control (P < 0.05).

Double immmunostaining in the SON and PVN in a subset (n = 3) of the fetuses of mannitol-treated ewes indicated that >80% of AVP neurons expressed Fos in their nuclei, whereas <10% of OT neurons did likewise (see Table 1). Figure 5 presents very typical immunocytochemical data from the SON and the PVN.

View larger version (119K): [in this window] [in a new window]   Fig. 5.   Histological sections of fetal sheep SON (A and B) and PVN (posterior magnocellular area; C and D) double immunostained for Fos and arginine vasopressin (AVP; A and C) or Fos and oxytocin (OT; B and D) 75 min after start of hypertonic challenge (mannitol: 1 ml · min1 · kg1 for 10 min). Note lack of Fos immunostaining in OT- compared with AVP-immunopositive neurons. Bar, 8.2 µm.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Systemically, there are two principle mechanisms for body water regulation. Small increases in extracellular fluid osmolality (~2%) stimulate AVP secretion (with resultant urinary antidiuresis) and thirst. Somewhat larger decreases (~10%) in circulating volume initiate AVP-mediated antidiuresis and likely ANG II-mediated thirst (for reviews of ANG II and thirst in adults, see Refs. 12 and 13). Although these systems are intact in the adult mammal, there have been limited studies examining them in the developing fetus. In the present study, we sought to primarily examine the responsiveness of the osmolality-sensing mechanism in the near-term ovine fetus.

An identical infusion of mannitol to the dam has been previously demonstrated to increase near-term fetal plasma osmolality by 12-13 mosmol/kg (23, 26). Although fetal swallowing responses to this stimulus have not been examined, acute fetal systemic hypertonicity (increase of 14 mosmol/kg) results in fetal swallowing stimulation (27). Similarly, maternal mannitol infusion stimulates fetal AVP secretion (26). In the near-term ovine fetus, AVP secretion is stimulated by lower levels of hypertonicity than that required for swallowing stimulation (22), consistent with most (3), but not all (34), adult studies. Furthermore, intracarotid hypertonic urea stimulation of fetal AVP secretion, but not swallowing, indicates the possibility of separate fetal osmoreceptors for the endocrine and behavioral components of body fluid homeostasis (22). Thus it is likely that both fetal swallowing and AVP secretion were stimulated in the present study. Although it would have been valuable to confirm basal and stimulated fetal plasma osmolality and AVP levels, fetal vascular catheters were not placed during the maternal surgery to avoid nonspecific fetal stress.

The present study is the first, to our knowledge, to demonstrate direct activation of the SFO and OVLT neurons by an osmotic stimulus in fetuses of any species. The activation of the SON and PVN suggests that neuronal pathways from the SFO and OVLT osmoreceptors to the hypothalamus may be intact at this gestational age. Additional input from systemic osmoreceptors associated with the hepatic portal system may also have contributed to the activation of the SON and PVN (2). Finally, there may be functional fetal osmoreceptors within the hypothalamus. The results of the present study agree with experiments in adult rats that used Fos immunoreactivity to demonstrate neuronal activation in ACVO, SON, and PVN via both peripherally and centrally applied stimuli (7, 17, 28, 29). Furthermore, the activation of the near-term (130 dGA) fetal ACVO is consistent with studies demonstrating fetal swallowing stimulation and AVP secretion in response to osmotic challenges (22). It is unknown whether the absence of dipsogenic responses in preterm fetuses (14) is a result of a lack of osmoreceptor sensitivity or function or the lack of development of neural tracts connecting these nuclei to motor neurons.

In contrast to the present study, both the dorsal and ventral portions of the MPON of adult rats express increased Fos immunoreactivity compared with controls in response to hypertonicity (20). The failure in the present study of the fetal MPON to be activated by fetal hypertonicity resulting from maternal mannitol challenge is not entirely unexpected. McKinley et al. (18) studied adult sheep and found that lesions of the OVLT or SFO, either alone or in combination with the MPON, attenuated the increases in drinking and plasma AVP in response to intracarotid hypertonic saline infusion. However, MPON lesion alone could not produce the same result. The lack of increased Fos immunoreactivity in the MPON, accompanied by activation of both the SON and PVN in the mannitol group, suggests the existence of alternative pathways for activation of the SON and PVN by the SFO and OVLT or pathways that simply pass through the MPON without synapsing in the fetal sheep. In addition, stimulatory inputs from additional areas (e.g., nucleus of the solitary tract) also may play a role in Fos activation in the SON and PVN during osmotic challenge as has been described in the adult rat (for review, see Ref. 2).

As shown in Fig. 5 and Table 1, Fos activation in both the SON and PVN occurred exclusively in AVP as opposed to OT neurons. In contrast, intraperitoneal injection of hypertonic saline into adult rats yields strong activation of Fos in OT neurons in SON (70%) and PVN (60%; Ref. 6). Also, in the adult rat, osmotic stimulation induces marked increases of OT neuronal mRNA and peptides (15). The present observation suggests that 1) ovine osmotic stimulation does not alter OT stimulation or 2) differences exist in maturation of the pathways and structures involved at 130 days of gestation in the fetal sheep compared with adults.

In sheep, a dual system of central nervous system osmoreceptors and sodium sensors mediates hypertonicity responses (19). In addition to fetal plasma hyperosmolality, maternal mannitol infusion also results in a loss of fetal plasma volume (due to fetal-to-maternal water flow) and thus may activate fetal renin-ANG II secretion. Although the near-term ovine fetus does not demonstrate swallowing stimulation in response to intravenous ANG II (26), central ANG II does stimulate ovine fetal swallowing (24). Thus central ANG II-mediated responses also may have contributed to the Fos stimulation in the present study. In addition, the blood-brain barrier of the fetal sheep appears to be fully developed by 123 days in utero (5). Given this and the fact that mannitol does not cross the placenta, thus yielding the stated fetal water loss to the ewe, the fetus is in a situation in the present experiment that is analogous to that of a dehydrated postnatal animal. Unfortunately, as interesting and important as it may be, the design of our experiment does not allow us to determine the exact nature of the stimulus for Fos activation of the affected ACVO (i.e., Na⁺ concentration vs. osmolality) and must await further study.

In summary, the present results indicate the responsiveness of near-term ovine fetal neuronal mechanisms to hyperosmolality. Given both the importance of body fluid regulation as a homeostatic process and the potential for imprinting of the in utero environment on adult body fluid homeostatic responsiveness, future studies need to be designed to further understand the mechanisms underlying the development of central sensitivity to body fluid tonicity. These studies need to determine 1) the gestational age at which the ACVO, SON, and PVN neurons first become able to respond to an osmotic challenge with increases in Fos, 2) as stated earlier, the exact nature of the stimulus, 3) the respective roles that each of these anterior brain structures play in neuronal and behavioral (i.e., swallowing) responses to osmotic challenge, and 4) the connections of the ACVO, SON, and PVN to other brain nuclei involved in body fluid homeostasis in fetal sheep.

In conclusion, the SFO, OVLT, SON, and PVN of late-gestation fetal sheep are activated by maternal mannitol-induced fetal plasma hyperosmolality and thus may contribute to the control of fetal fluid balance in late gestation. Unlike the response in adult rats, OT neurons of the fetal sheep SON and PVN do not show Fos activation in response to hypertonic challenge.