Among the many changes in pregnancy that are controlled by central mechanisms (1) are major adaptations in the control of blood volume and osmolality in women and other species, including the rat. By the end of pregnancy blood volume is increased by ca. 55% and plasma volume by ca. 40% (2, 3) while plasma sodium concentration ([Na+]), and osmolarity, are decreased by ca. 4% (2). Clearly, homeostats regulating these variables are reset, and the blood volume expansion involves substantial retention of water and electrolytes, especially NaCl. These changes ensure adequate blood supply to the gravid uterus and other organs in pregnancy. The question addressed here is what are the mechanisms that lead to or maintain this state of physiological hyponatraemic hypervolaemia in pregnancy? Multiple mechanisms regulate blood volume and [Na+] (and osmolarity); roles of aldosterone, angiotensin II (AII), arginine vasopressin (AVP) and atrial natriuretic peptide (ANP) are well-established (4). In some species, a role for oxytocin in promoting natriuresis has been shown (e.g. dog, rat) (5-7). Post-prandial natriuresis may involve oxytocin, ANP and satiety peptides, such as leptin (8). Here we focus on the adaptations in the responsiveness of the vasopressin and oxytocin neurohypophysial systems in late pregnancy, and the roles of the high levels of oestrogens and progesterone and of the pregnancy hormone relaxin in these changes.

Vasopressin

Vasopressin is secreted from the posterior pituitary gland by the axon terminals of magnocellular neurosecretory neurones when action potentials arrive from the cell bodies in the supraoptic (SON) and paraventricular nuclei (PVN) (9). The major stimuli that excite these vasopressin neurones are an increase in extracellular osmolality, generally a result of increased [Na+] (with Cl-), and decreased blood volume and pressure (9, 10). The peripheral actions of vasopressin counteract such changes through stimulating renal water reabsorption (via V2 receptors) and vasoconstriction (via V1a receptors) (11). Consequently, decreased extracellular fluid osmolality and increased blood volume or pressure normally inhibit vasopressin neurones and secretion (10, 12). However, the hyponatraemic hypervolaemia of pregnancy is associated with a reduced threshold (without a change in slope) for hyperosmotic stimulation of vasopressin secretion (13). Accordingly, this shift in threshold indicates that the hyponatraemic hypervolaemia of pregnancy is a simple result of increased water retention caused by vasopressin. Also, the augmentation of vasopressin secretory responses to hyperosmolarity by reduced blood volume persists in pregnancy (3). The role of vasopressin in the hyponatraemic hypervolaemia of pregnancy is complemented by increased water intake (2).

Oxytocin

Oxytocin, like vasopressin, is also secreted from the posterior pituitary gland. Its peripheral roles are primarily to promote uterine contractions during parturition and to effect milk ejection during lactation (14). However, like vasopressin, oxytocin secretion is also stimulated by hyperosmolarity of extracellular fluid, at least in the rat (15-17), and it has peripheral natriuretic actions. These natriuretic actions involve direct effects in the kidney (4) and actions in the right atrium to stimulate ANP secretion (7). Oxytocin receptors are expressed in the heart, and so also is oxytocin (18). Oxytocin is secreted post-prandially (7), and may contribute to post-prandial ANP secretion (19), and hence post-prandial natriuresis.

Salt appetite is increased in pregnancy (2) and total circulating sodium is clearly increased, although water retention reduces plasma [Na+] while increasing volume. In these circumstances a natriuretic response can be expected, through stimulation of ANP secretion by right atrial distension (7), but clearly any such response is either ineffective or suppressed in pregnancy. Indeed, circulating ANP levels decrease in late pregnancy (19-21). Reduced oxytocin content in the atria in pregnancy, and reduced oxytocin receptor mRNA expression near term (19) may contribute to this reduced ANP secretion. Also, in late pregnancy the kidney is refractory to the natriuretic actions of ANP (22), through increased phosphodiesterase-5 activity (23).

The maintenance of the hypervolaemia of pregnancy might thus involve Na+ retention resulting from reduced stimulation of oxytocin secretion due to the hyponatraemia caused by vasopressin action and by increased drinking. Moreover, in contrast with the stimulation of vasopressin secretion by hypovolaemia, oxytocin neurones and secretion are stimulated by hypervolaemia; hence consequent natriuresis will contribute to normalisation of expanded blood volume (4, 24, 25). Thus, it may be that reduced oxytocin responses to the hypervolaemia of pregnancy also contribute to the expanded blood volume of pregnancy. Alternatively, stimulation of oxytocin secretion by further volume increase in late pregnancy may be preserved to prevent volume overload. Hence, it becomes important to consider whether adaptations in the control of oxytocin neurones are involved in maintaining the hyponatraemic hypervolaemia of late pregnancy.

Additional actions of oxytocin that are important in this context are that centrally-released oxytocin, potentially from the dendrites of magnocellular neurones (26) as well as from centrally-projecting parvocellular neurones (27), has salt-appetite suppressing (27, 28, 29) and anorectic actions (30). Hence, reduced central release of oxytocin in pregnancy would favour the increased food and salt intake that is seen especially in late pregnancy (2, 30). Overall, increased vasopressin secretion, and reduced oxytocin secretion (and hence reduced ANP secretion) in pregnancy could underlie the hypervolaemia and hyponatraemia in late pregnancy.


MECHANISMS OF HYPEROSMOTIC STIMULATION
OF VASOPRESSIN AND OXYTOCIN NEURONES
Both magnocellular vasopressin and oxytocin neurones are directly osmosensitive; they respond to increases in extracellular osmolarity by depolarising and increasing their firing-rate (10, 17). Since these neurones control body water and [Na+] they are also osmoreceptors. Depolarisation is the result of shrinkage of the cell-bodies of the neurones, consequent on net water flux out of the cells into the hyperosmotic environment, which activates a stretch-inactivated cationic conductance channel (the osmoreceptor transduction channel, possibly transient receptor potential vanilloid [TRPV]) (10). The firing-rate and secretory responses of vasopressin neurones to graded systemic hyperosmotic challenge are correlated with changes in osmolarity (9, 31). This provides effective negative feedback control of extracellular osmolarity through graded antidiuretic actions of vasopressin (17). The firing-rate and secretory responses of oxytocin neurones show a similar relationship after an intraperitoneal hyperosmotic saline challenge (31, 32), and during intravenous hypertonic saline infusion (a hyperosmotic/ hypervolaemic stimulus) the responses are closely correlated with the amount of salt infused (16).


INPUTS FROM THE LAMINA TERMINALIS AND BRAINSTEM
Lamina terminalis

Although the magnocellular vasopressin and oxytocin neurones are directly osmosensitive, in vivo they rely upon input from lamina terminalis structures to express this property as changes in firing rate. Acute lesion of the ventral lamina terminalis, in the region anterior and ventral to the third ventricle (AV3V: including the organum vasculosum of the lamina terminalis [OVLT] and median preoptic nucleus [nucleus medianus, MnPO], and projections from the subfornical organ [SFO]) induces adipsia, profoundly impairs osmoregulation, and prevents activation of electrical or secretory activity of magnocellular oxytocin and vasopressin neurones (33- 36).

The lamina terminalis input to magnocellular neurones has two roles; firstly, it provides tonic glutamatergic excitatory drive (35), and secondly, the OVLT, SFO and the MnPO contain osmosensitive neurones, which transmit information to the magnocellular neurones (10). The inhibitory transmitter GABA is also important in this pathway. During systemic hyperosmotic stimulation the release in the SON of both glutamate and GABA is increased, and GABA has an important modulatory role to limit the excitation of the magnocellular neurones, although the net effect of electrically stimulating the AV3V region is to excite oxytocin neurone firing rate and oxytocin secretion (16, 35). There is also evidence for roles of both AII and ANP as excitatory transmitters in the lamina terminalis input to the magnocellular neurones (4, 37, 38).

The OVLT and SFO are circumventricular organs, providing access for circulating peptides which act via receptors expressed in these structures to regulate oxytocin and vasopressin neurone activity and thirst. These peptides include AII, which registers low blood volume and pressure, and the pregnancy hormone relaxin (4, 39, 40). AII and relaxin actions via the OVLT or SFO, respectively, regulate magnocellular neurones and thirst (36, 39, 40). In humans, intravenous hypertonic saline infusion induces thirst and activates the lamina terminalis, as visualised with fMRI/BOLD imaging (41). The lamina terminalis has an important role in the regulation of natriuresis, as well as in the control of water balance through the renal actions of vasopressin and through thirst mechanisms; thus an AV3V lesion strongly attenuates the excretion of a systemically infused isotonic or hypertonic Na+ and volume load (7, 42). This has been interpreted as a result of reduced release of ANP by the brain, or perhaps more likely to result from suppressed oxytocin secretion as a result of loss of an ANP input from the lamina terminalis (7). Oxytocin neurones do express ANP, and intriguingly, ANP mRNA expression in the SON is increased near term, while expression in the preoptic area is decreased (43). Since central ANP contributes to mediation of osmotic stimulation of oxytocin neurones (38), reduced central ANP level in late pregnancy might contribute to their reduced response to a hyperosmotic stimulus.

Brainstem inputs

A wide spectrum of information from the periphery is conveyed by projections from the brainstem to the magnocellular oxytocin and vasopressin neurones in the hypothalamus. These are predominantly from the noradrenergic neurones in the A1 group in the ventrolateral medulla and the A2 group in the nucleus tractus solitarii (9). The inputs include neural signals from arterial baroreceptors and atrial volume receptors, though these project predominantly to vasopressin neurones (44), so the route for cardiovascular inputs to oxytocin neurones (25) is unclear. Peptide hormones from the gastro-intestinal tract signal information about food intake and digestion (e.g. cholecystokinin, CCK) (45, 46) and cytokines signalling infection (47) also act via relays to the hypothalamus in the brainstem. Since food intake involves a Na+ load activation of magnocellular oxytocin neurones, and hence oxytocin secretion, by CCK after a meal would contribute to Na+ excretion.


OXYTOCIN NEURONE RESPONSES TO HYPEROSMOTIC STIMULATION IN PREGNANCY
Intraperitoneal injection of hypertonic saline (1.5M, 4 ml/kg) reliably increases circulating oxytocin concentrations in urethane-anaesthetised virgin female rats, but in late pregnant rats the increase in oxytocin secretion is reduced (Fig. 1), despite a similar increase in plasma [Na+] (48). The first question from this finding was whether oxytocin neurones are excited less by activation of the lamina terminalis input. However, this was found not to be the case as the substantial oxytocin secretory response to electrical stimulation of the AV3V region is not different between late pregnant and virgin rats (48). This result indicated that either the hyperosmotic stimulus used in late pregnancy was insufficient if the threshold for stimulation was not reduced pari passu with the physiological hyponatraemia of pregnancy, or that the oxytocin neurones or the lamina terminalis input, is actively inhibited in late pregnancy.

Fig. 1. Oxytocin responses to different stimuli in late pregnancy. Female rats were given naloxone (2-5 mg/kg i.v.) to antagonise actions of endogenous opioids before exposure to the stimuli. Changes in plasma oxytocin concentration were measured in blood samples collected from urethane-anaesthetised virgin and pregnant (day 21) rats after (i) electrical stimulation of the AV3V region (0.5 mA, 1 ms pulses, 10s/10s off at 25 Hz for 2 min), (ii) hypertonic saline (1.5 M NaCl i.p.), (iii) cholecystokinin (CCK; 20 µg/kg i.v.); or from conscious rats following (iv) interleukin-1ß (IL-1ß; 500 ng/kg i.v.). After naloxone both AV3V stimulation and i.p. hypertonic saline less effectively stimulate oxytocin secretion in late pregnant rats compared to virgin rats, indicating an opioid-independent reduction in the drive by lamina terminalis inputs in late pregnancy. Conversely, oxytocin secretory responses to CCK and IL-1ß are greater in late pregnant rats than in virgin rats after naloxone, indicating that without endogenous opioid inhibition the brainstem input to the oxytocin neurones is more effective in late pregnancy. *p<0.05 vs within treatment (One-way ANOVA). See (48, 51, 52, 53).

In studies of hyponatraemia induced by V2 agonist (DDAVP) treatment and liquid diet feeding, the threshold for hyperosmotic stimulation of oxytocin secretion does not shift to the left (31). Consequently, the curve describing the relationship between plasma [Na+] and oxytocin secretion is non-linear, rising sharply for increases just above the normal plasma [Na+] (31). Hence, the decrease in plasma [Na+] in late pregnancy can be predicted to mean that a larger increase in plasma [Na+] is needed to produce a measurable oxytocin secretory response. Indeed, increasing the hypertonic saline stimulus by giving late pregnant rats an i.p. injection of 5 ml/kg 2M NaCl produces a similarly large stimulation of oxytocin secretion to that produced by 4 ml/kg 1.5M NaCl in virgin rats at 40-50 minutes post-injection (15). Furthermore, the oxytocin secretory response to combined hypovolaemia and hyperosmotic stimulation in pregnancy, induced by subcutaneous polyethylene glycol injection, is retained (49).

Local mechanisms inhibiting oxytocin neurones

Although the explanation for the reduced oxytocin responses to a hyperosmotic stimulus in late pregnancy seems simple, there are several mechanisms that inhibit oxytocin neurones which might be involved. These include nitric oxide (NO), endogenous opioids, endocannabinoids and allopregnanolone.

Nitric oxide

Oxytocin neurones strongly express neuronal nitric oxide synthase (nNOS), which generates (NO) (15). However, the expression of nNOS mRNA in the supraoptic nucleus is decreased towards the end of pregnancy, and whilst a nNOS inhibitor (L-NNA) enhances the oxytocin secretory response to an i.p. hypertonic saline stimulus in virgin rats, it is without effect near the end of pregnancy (15). Hence inhibitory (NO) mechanisms in oxytocin neurones are evidently inactive by the end of pregnancy.

Endogenous opioids

In virgin rats endogenous opioids inhibit oxytocin secretion only at the level of the posterior pituitary (50). In vitro studies show that this inhibitory -opioid-mediated mechanism is down-regulated as pregnancy nears term (50). Indeed, treatment of non-pregnant and late pregnant female rats with the opioid antagonist naloxone before challenge with an i.p. hypertonic saline stimulus is much less effective in enhancing the oxytocin secretory response in the pregnant rats (Fig. 1) (48, 51). This finding confirms the reduced response of oxytocin neurones to the hyperosmotic stimulus. Moreover, electrical stimulation of the AV3V region after naloxone is also less effective in stimulating oxytocin secretion in late pregnant compared with non-pregnant rats (Fig. 1).

In striking contrast, the oxytocin response to systemic CCK is greatly potentiated after naloxone in late pregnant rats compared with non-pregnant rats (Fig. 1) (52). CCK acts via the brainstem, and this effect of naloxone in late pregnancy is typical of other stimuli that act via this route, including systemic interleukin-1ß (Fig. 1) (53, 54) and intravenous hypertonic saline infusion, which acts as a hypernatraemic and hypervolemic stimulus (42, 51). The brainstem input to the oxytocin neurones (specifically the A2 noradrenergic projection mediating IL-1ß signalling) is less effective in late pregnancy due to preterminal opioid inhibition of noradrenaline release (47). The source of this opioid is likely to be the A2 neurones themselves, from up-regulated pro-enkephalin-A gene expression (47). Hypersecretion of oxytocin in response to interleukin-1ß after naloxone pre-treatment in late pregnancy compared with non-pregnant rats is evidently due to increased efficiency of stimulus-secretion coupling in the posterior pituitary, probably consequent on the enlarged oxytocin store (55). Similar analysis has not been performed for hypervolaemic stimulation of oxytocin neurones in late pregnancy, but it would seem advantageous for effectiveness of this input to be restrained to maintain the physiological hypervolaemia of pregnancy.

Endocannabinoids

Stimulated oxytocin neurones generate endocannabinoids, which act presynaptically to retrogradely inhibit synaptic inputs (56). For example, endocannabinoids inhibit the activation of oxytocin neurones by electrical stimulation of the OVLT (57). Whether any altered production or actions of endocannabinoids in pregnancy is involved in the attenuation of oxytocin neurone responses to osmotic stimulation has not been investigated.

Allopregnanolone

Allopregnanolone is a neuroactive steroid metabolite of progesterone which allosterically modulates GABAA receptors, including those on oxytocin neurones. When the receptors are occupied by GABA, allopregnanolone prolongs the opening time of the receptor channel, and consequently the neurones are more potently inhibited (58). Whether the high levels of allopregnanolone in pregnancy, consequent on metabolism of high progesterone levels (59), are involved in the reduced oxytocin neurone responses to a hyperosmotic stimulus has not been investigated. However, allopregnanolone does seem to be responsible for upregulating expression of pro-enkephalin A in the NTS in pregnancy (54), hence inhibiting the effectiveness of excitatory brainstem inputs to oxytocin neurones. So, in the absence of evidence for up-regulated inhibition of oxytocin neurone responses to hyperosmotic stimulation in pregnancy, the question of whether responsiveness is reduced because of the hyponatraemia of pregnancy remains.


HORMONES OF PREGNANCY AND OSMORESPONSIVENESS
OF OXYTOCIN NEURONES
Sex steroids: oestrogen and progesterone

Circulating levels of 17ß-oestradiol and, to a larger extent, progesterone, are greatly increased in late pregnancy (1). There are several sites where oestrogen and progesterone might act to modulate central osmoreceptor mechanisms, though resulting actions on oxytocin and vasopressin secretion in this context are not clear. Classical oestrogen receptor (ER)- is expressed in neurones in the SFO, OVLT and MnPO, whereas only ER-ß is expressed in the magnocellular vasopressin and oxytocin neurones, predominantly in the former (60-62). The SFO, OVLT and MnPO neurones that express ER- respond to hyperosmotic stimulation (63) and project to the magnocellular neurones (60). Magnocellular PVN and SON neurones strongly express G protein-coupled receptor 30 (GPR 30), regarded as the non-genomic ER, mediating rapid membrane actions (64), such as on electrical activity of oxytocin neurones (65). Progesterone receptor is not expressed in the magnocellular neurones, and few neurones in the lamina terminalis express progesterone receptor (66). However, allopregnanolone, the neuroactive steroid metabolite of progesterone which is abundant in late pregnancy, modulates oxytocin neurones via allosteric enhancement of GABAA receptor function (58).

We manipulated sex steroid levels in late pregnant and virgin rats, aiming to reverse or mimic, respectively, the reduced oxytocin response to hyperosmotic stimulation (67). Firstly, oxytocin responses to i.p. hypertonic saline were measured in pregnant rats on day 21, after bilateral ovariectomy 24 h previously, which reduced plasma 17ß-oestradiol and progesterone levels by ca. 67%, and increased [Na+] by ca 2 nmol/l. However, ovariectomy did not alter the suppressed oxytocin secretory response to hypertonic saline (67). Secondly, virgin rats were exposed to pregnancy levels of 17ß-oestradiol and progesterone for 17 days, via s.c. implanted capsules [as in (68)]. In these rats, the oxytocin response to the i.p. hypertonic saline stimulus was not different from that in control virgin rats, even after i.v. naloxone (5 mg/kg) (67). Hence ovarian steroids alone seem not to be responsible for the reduced oxytocin responses to the hyperosmotic stimulus in pregnancy. Similarly, chronic 17ß-oestradiol and progesterone treatment does not reduce the threshold for hyperosmotic stimulation of vasopressin secretion (69).

Ovariectomy removes the source of relaxin as well as 17ß-oestradiol and progesterone, so the experiment described above indicated that relaxin does not seem important in maintaining suppressed oxytocin responses to hyperosmotic stimulation in just the last day of pregnancy. We next evaluated a role for more prolonged actions of relaxin.

Relaxin

Relaxin is an ovarian insulin-like peptide hormone, produced by the corpora lutea in pregnancy (70). In the rat, relaxin is detectable in blood around day 10 of gestation and levels increase throughout pregnancy (71). Relaxin relaxes the interpubic ligament, inhibits uterine contractions, contributes to softening of the uterine cervix and promotes nipple development (72). There is strong evidence that relaxin is also the driver for the major changes in blood volume and osmolality changes in pregnancy. In late pregnancy passive immunization with relaxin antibody reduces water intake (73, 74), and relaxin knockout mice do not have reduced plasma osmolality near the end of pregnancy, unlike wild-type mice (72).

Systemic relaxin infusion activates neurones in the SFO, OVLT, MnPO, and magnocellular neurones in the SON and PVN, as shown by Fos immunocytochemistry (40). Direct actions have been shown in the SFO by electrophysiological recording, and most relaxin-responsive neurones also respond to AII (40). AII is also a powerful dipsogen and stimulates vasopressin secretion (39), and may partly mediate the central actions of relaxin (40, 75). Relaxin stimulates drinking via the SFO since actions on drinking are abolished by SFO lesion, while activation of the OVLT, and SON and PVN magnocellular neurones is unaffected (40, 76). OVLT lesion reduces relaxin-induced Fos expression in the SON and magnocellular PVN without affecting the drinking response to relaxin (40), indicating that stimulation of vasopressin and oxytocin secretion by relaxin is mediated by the OVLT. Given acutely, by i.v. injection, relaxin excites both vasopressin and oxytocin neurones in non-pregnant rats (77), but importantly, the stimulatory action on oxytocin (but not vasopressin) secretion is lost in late pregnancy (78). Hence, in late pregnancy central actions of relaxin will favour water and Na+ retention through sustained stimulation of vasopressin secretion without stimulation of oxytocin secretion (Fig.2).

Fig. 2. Neuropeptide mechanisms in the hyponatraemic hypervolaemia of late pregnancy. Hypernatraemia (high [Na+]) stimulates magnocellular oxytocin and vasopressin neurones in the paraventricular (PVN) and supraoptic (SON) nuclei directly, and via osmoreceptive neurones in the organum vasculosum of the lamina terminalis (OVLT), subfornical organ (SFO) and nucleus medianus (NM; median preoptic nucleus). The projection from the lamina terminalis involves glutamate (GLU; excitatory) and -aminobutyric acid (GABA; inhibitory) transmission, atrial natriuretic peptide (ANP) and angiotensin II (AII). The SFO also mediates drinking responses to high [Na+] and to relaxin. The high level of relaxin from the corpora lutea in late pregnancy stimulates magnocellular vasopressin neurones via relaxin receptors (RXFP1) in the OVLT and SFO. Outwith pregnancy and in early but not late pregnancy, relaxin similarly stimulates oxytocin neurones. AII in the pathway from the lamina terminalis partly mediates relaxin actions on magnocellular neurones. In late pregnancy relaxin-stimulated vasopressin secretion, consequent renal water reabsorption, and drinking lead to the hyponatraemic hypervolaemia of pregnancy. Vasopressin neurone responsiveness to a hyperosmotic stimulus is maintained, despite the increased blood volume and [Na+]. In contrast the threshold for osmotic stimulation of oxytocin secretion is not decreased, and together with reduced ANP secretion and action, Na+ retention is favoured. These effects of relaxin may require interaction with actions of oestrogen on the SFO, NM and OVLT, and progesterone actions, perhaps through its neuroactive steroid metabolite, allopregnanolone (AP) at GABA receptors.

The action of relaxin via the OVLT on drinking (79) and vasopressin secretion (which is increased 3.5-fold) (40, 80) is maintained over 3 weeks. Consequently, chronic central relaxin treatment, or relaxin expression in brain induced via a viral vector, reduces plasma osmolality by ca. 3% (80). This effect of relaxin is a result of a reduced threshold for hyperosmotic stimulation of vasopressin secretion, and increased drinking (81).

Radiolabelled relaxin binds strongly in the SFO, and there is weaker binding in the OVLT, PVN and SON (82). Of the family of four G-protein coupled relaxin receptors, LGR7 (or RXFP1) (70) is the primary receptor for relaxin; this receptor activates adenylate cyclase, and interacts with nitric oxide signalling mechanisms (70). RXFP1 (LGR7) mRNA and protein are similarly strongly expressed in the SFO, the magnocellular PVN and SON, but weakly in the OVLT (83).

We have tested the hypothesis that chronic relaxin exposure is responsible for the reduced responses of oxytocin neurones to hyperosmotic stimulus in late pregnancy. Contrary to expectation, removing the source of endogenous relaxin in pregnancy by bilateral ovariectomy on day 15 (with 17b-oestradiol and progesterone replacement to maintain pregnancy) did not alter the reduced oxytocin response to the hyperosmotic stimulus tested 5 days later. The oxytocin response was indistinguishable from that in sham-operated pregnant rats, and significantly less than the response in virgin rats, even after naloxone administration (Brunton PJ, Bull PM, Russell JA, unpubl.). While this experiment did not support a continuing role for relaxin on oxytocin neurone responsiveness in the last few days of pregnancy, we next tested the action of relaxin infused centrally for 7 days in virgin rats with or without 17ß-oestradiol and progesterone implants to simulate pregnancy levels. As expected, i.c.v. relaxin infusion for seven days increased water intake (Fig. 3a), and reduced plasma [Na+] (Fig. 3b). Importantly, the combination of i.c.v. relaxin infusion and systemic pregnancy levels of 17ß-oestradiol and progesterone significantly reduced the oxytocin secretory response to the acute i.p. hyperosmotic stimulus, provided that opioid inhibition, presumably at the posterior pituitary gland (50) was antagonised by naloxone treatment. Hence, these results indicate that a chronic action of relaxin with 17ß-oestradiol and progesterone in late pregnancy may cause the reduced oxytocin response to hyperosmotic stimulation. The mechanisms and sites at which relaxin and the female sex steroids interact to modulate oxytocin neurone responsiveness need further study.

Fig. 3. Oxytocin response to acute hyperosmotic stimulus in virgin rats after chronic central relaxin infusion. Virgin rats were fitted, under halothane anaesthesia, with an intracerebroventricular (i.c.v.) cannula aimed at the right lateral ventricle by a dorsal approach, secured to the skull with dental acrylic and connected to a s.c. osmotic minipump (Alzet 2001, infusion rate 1ml/h). The minipump infused for 7 days either artificial cerebrospinal fluid (aCSF; n=5) or synthetic human recombinant relaxin (RLX) at a rate of 250ng/h. Rats were also implanted with either empty capsules (RLX; n=5) or oestradiol-17ß (E) and progesterone (P) filled silastic capsules implanted s.c. (RLX+E/P; n=5) see (68). An additional group were untreated controls (n=6). Post-surgery the rats were housed singly, and daily water intake was monitored by weighing the drinking bottle. After 7 days the rats were anaesthetised with urethane (1.25 g/kg i.p.), and the femoral artery and vein were cannulated for blood sampling and drug administration, respectively. Two basal blood samples were collected 10 min apart before i.p. injection of 1.5M NaCl (4ml/kg). Further samples were withdrawn 35 and 45 minutes later, then naloxone was injected (5 mg/kg i.v.) to antagonise opioid inhibition of oxytocin secretion, and a final sample was drawn 10 minutes later. Plasma oxytocin concentration was determined by radioimmunoassay [as in (68)], and plasma [Na+] by flame photometry. Data are presented as group means ± sem. Data were analysed by ANOVA (water intake), or RM ANOVA (oxytocin data), with significance level at p<0.05. Post-hoc tests were by the Newman-Keuls method. Relaxin infusion (a) increased mean daily water intake (across days 3-7; *p<0.004 vs non-relaxin-infused groups) and (b) significantly reduced plasma [Na+] (*p<0.05 vs basal levels in same group; +p<0.01 vs RLX+E/P group; ++p<0.02 vs relaxin-infused groups); hence effectiveness of the i.c.v. relaxin infusion was confirmed (see 79, 80). Hypertonic saline given i.p. (c) increased plasma oxytocin concentration similarly across the 4 groups (*p<0.01) at 35 min (and at 45 min, data not shown). Naloxone further increased oxytocin secretion in each group (#p<0.001), however the increase from basal was significantly attenuated in the RLX+E/P group versus the other groups (15-fold increase in the RLX+E/P group, vs 62-fold, 53-fold and 50-fold increases in the control, aCSF and RLX only groups, respectively; **p<0.05). Basal plasma oxytocin concentration was greater in the RLX+E/P group than in the other groups (p<0.05). All experiments were conducted in accordance with the UK Animals (Scientific Procedures) Act 1986.

OTHER ADAPTATIONS IN PREGNANCY:
APPETITE PEPTIDES AND OXYTOCIN NEURONES
CCK and neuropeptide Y (NPY) are peptides that regulate appetite but also stimulate oxytocin neurones (45, 84, 85), indicating additional roles in regulating post-prandial natriuresis. Magnocellular oxytocin neurones release oxytocin from their dendrites when stimulated, which may contribute to suppression of salt and food appetite (30), complementing the natriuretic actions of oxytocin. Changes in pregnancy in release of central oxytocin in response to CCK, NPY, or indeed hyperosmotic stimulation, have not been reported. In late pregnant rats, systemic CCK is less effective in stimulating oxytocin secretion compared with virgin rats (86).

Moreover, in late pregnancy the oxytocin response to exogenous NPY, given centrally (NPY is produced by arcuate nucleus neurones and oxytocin neurones express NPY receptors) (87), is lost (85). In contrast, the eating response to NPY is maintained in late pregnancy (85). This lack of a stimulatory effect of NPY on oxytocin neurones in pregnancy is expected to contribute to both increased food and Na+ intake and the retention of Na+, supporting the hypervolaemia of pregnancy. We have sought similar changes in oxytocin neurone responses to leptin, an adipocyte cytokine-like peptide that signals fat deposition and acts centrally to suppress appetite and regulate metabolism, predominantly via the arcuate nucleus, but also via the nucleus tractus solitarii (88). Furthermore, leptin can enter the brain via the choroid plexus and oxytocin neurones express leptin receptors (89). Leptin has natriuretic actions on the kidney, particularly in conditions of mild sodium and volume expansion (8), so its central actions on oxytocin neurones and any changes in pregnancy become of interest. The paradoxical increase in plasma leptin levels during the latter half of pregnancy, when food consumption is markedly increased (90) indicates that pregnancy is a physiological state of leptin resistance (91). In a preliminary study, we have investigated acute actions of intravenous leptin administration in virgin and late pregnant rats by recording extracellularly the firing of individual identified SON oxytocin neurones in urethane-anaesthetised rats (Arunachalam S, Leng G, Russell JA, unpubl.). We found that rat recombinant leptin (i.v. 100 µg/rat) excited oxytocin neurones similarly in ad libitum fed virgin and late pregnant rats within a few minutes, indicating oxytocin neurones do not become leptin resistant in late pregnancy. We also studied oxytocin neurone responses to leptin after overnight fasting, predicting that the decreased circulating leptin level (92) might increase oxytocin neurone responsiveness to exogenous leptin in late pregnant rats. Instead, leptin did not excite oxytocin neurones in fasted pregnant rats but still excited oxytocin neurones in fasted virgin rats. This loss of leptin action on oxytocin neurones with fasting in late pregnancy is unexplained, but might contribute to maintenance of hyponatraemic hypervolaemia, as well as favouring increased food intake.


CONCLUSION
Overall, the reduced oxytocin response in late pregnancy to the hypertonic saline challenge used in our studies seems not to be a result of up-regulation of any of the several inhibitory mechanisms investigated. Rather, the reduced response can be interpreted as a result of the stimulation by relaxin of vasopressin secretion and drinking, with a consequent reduction in extracellular [Na+] that simply shifts the operational range on the osmotic stimulus-oxytocin response curve to the left. Since oxytocin is natriuretic, this shift will support the hypervolaemia of pregnancy. Other adaptations that restrain oxytocin neurone responsiveness in pregnancy will contribute to maintaining this state.

Acknowledgements: PJB and JAR are supported by the BBSRC. SA is supported by a PhD Studentship from the College of Medicine and Veterinary Medicine, University of Edinburgh and an ORS award. Dr PM Bull provided technical assistance. Human recombinant relaxin was kindly donated by Professor GW Tregear, Howard Florey Institute, University of Melbourne, Australia.

Conflicts of interest statement: None declared.


REFERENCES

1. Brunton PJ, Russell JA. The expectant brain: adapting for motherhood. Nat Rev Neurosci 2008; 9: 11-25.
2. Atherton JC, Dark JM, Garland HO, Morgan MR, Pidgeon J, Soni S. Changes in water and electrolyte balance, plasma volume and composition during pregnancy in the rat. J Physiol 1982; 330: 81-93.
3. Barron WM, Stamoutsos BA, Lindheimer MD. Role of volume in the regulation of vasopressin secretion during pregnancy in the rat. J Clin Invest 1984; 73: 923-932.
4. Antunes-Rodrigues J, de Castro M, Elias LL, Valenca MM, McCann SM. Neuroendocrine control of body fluid metabolism. Physiol Rev 2004; 84: 169–208.
5. Brooks FP, Pickford M. The effect of posterior pituitary hormones on the excretion of electrolytes, in dogs. J Physiol 1958; 142: 468-493.
6. Verbalis JG, Mangione MP, Stricker EM. Oxytocin produces natriuresis in rats at physiological plasma concentrations. Endocrinology 1991; 128: 1317-1322.
7. Haanwinckel MA Elias LK, Favaretto AL, Gutkowska J, McCann SM, Antunes-Rodrigues J. Oxytocin mediates atrial natriuretic peptide release and natriuresis after volume expansion in the rat. Proc Natl Acad Sci USA 1995; 92: 7902-7906.
8. Villarreal D, Reams G, Freeman R, Spear R, Tchoukina I, Samar H. Leptin blockade attenuates sodium excretion in saline-loaded normotensive rats. Mol Cell Biochem 2006; 283(1-2): 153-157.
9. Leng G, Brown CH, Russell JA. Physiological pathways regulating the activity of magnocellular neurosecretory cells. Prog Neurobiol 1999; 57: 625-655.
10. Bourque CW. Central mechanisms of osmosensation and systemic osmoregulation. Nat Rev Neurosci 2008; 9: 519-531.
11. Manning M, Stoev S, Chini B, Durroux T, Mouillac B, Guillon G. Peptide and non-peptide agonists and antagonists for the vasopressin and oxytocin V1a, V1b, V2 and OT receptors: research tools and potential therapeutic agents. Prog Brain Res 2008; 170: 473-512.
12. Renaud LP. CNS pathways mediating cardiovascular regulation of vasopressin. Clin Exp Pharmacol Physiol 1996; 23: 157-160.
13. Durr JA, Stamoutsos B, Lindheimer MD. Osmoregulation during pregnancy in the rat. Evidence for resetting of the threshold for vasopressin secretion during gestation. J Clin Invest 1981; 68: 337-346.
14. Burbach PJ, Young L, Russell JA. Oxytocin: Synthesis, secretion, and reproductive function. In: Knobil and Neill’s Physiology of Reproduction, J D Neill (ed). Third Edition, San Diego, Elsevier, 2006, pp. 3055-3128.
15. Srisawat R, Ludwig M, Bull PM, Douglas AJ, Russell JA, Leng G. Nitric oxide and the oxytocin system in pregnancy. J Neurosci 2000; 20: 6721-6727.
16. Leng G, Brown CH, Bull PM, et al. Responses of magnocellular neurons to osmotic stimulation involves coactivation of excitatory and inhibitory input: an experimental and theoretical analysis. J Neurosci 2001; 21: 6967–6977.
17. Bourque CW, Ciura S, Trudel E, Stachniak TJ, Sharif-Naeini R. Neurophysiological characterization of mammalian osmosensitive neurones. Exp Physiol 2007; 92: 499-505.
18. Mukaddam-Daher S, Jankowski M, Wang D, Menaouar A, Gutkowska J. Regulation of cardiac oxytocin system and natriuretic peptide during rat gestation and postpartum. J Endocrinol 2002; 175: 211-216.
19. McCann SM, Gutkowska J, Antunes-Rodrigues J. Neuroendocrine control of body fluid homeostasis. Braz J Med Biol Res 2003; 36: 165-181.
20. Jansakul C, King RG, Boura AL, Brennecke SP, Handberg GM. Plasma levels of atrial natriuretic peptides during pregnancy and post partum in the rat. J Endocrinol 1989; 120: 113-117.
21. Nadel AS, Ballermann BJ, Anderson S, Brenner BM. Interrelationships among atrial peptides, renin, and blood volume in pregnant rats. Am J Physiol 1988; 254: R793-R800.
22. Masilamani S, Castro L, Baylis C. Pregnant rats are refractory to the natriuretic actions of atrial natriuretic peptide. Am J Physiol 1994; 267: R1611-R1616.
23. Knight S, Snellen H, Humphreys M, Baylis C. Increased renal phosphodiesterase-5 activity mediates the blunted natriuretic response to ANP in the pregnant rat. Am J Physiol Renal Physiol 2007; 292: F655-F659.
24. Randolph RR, Li Q, Curtis KS, Sullivan MJ, Cunningham JT. Fos expression following isotonic volume expansion of the unanesthetized male rat. Am J Physiol 1998; 274: R1345-R1352.
25. Godino A, Giusti-Paiva A, Antunes-Rodrigues J, Vivas L. Neurochemical brain groups activated after an isotonic blood volume expansion in rats. Neuroscience 2005; 133: 493-505.
26. Ludwig M, Leng G. Dendritic peptide release and peptide-dependent behaviours. Nat Rev Neurosci 2006; 7: 126-136.
27. Geerling JC, Loewy AD. Central regulation of sodium appetite. Exp Physiol 2008; 93: 177-209.
28. Stricker EM, Verbalis JG. Central inhibition of salt appetite by oxytocin in rats. Regul Pept 1996; 66: 83-85.
29. Puryear R, Rigatto KV, Amico JA, Morris M. Enhanced salt intake in oxytocin deficient mice. Exp Neurol 2001; 171: 323-328.
30. Douglas AJ, Johnstone LE, Leng G. Neuroendocrine mechanisms of change in food intake during pregnancy: a potential role for brain oxytocin. Physiol Behav 2007; 91: 352-365.
31. Verbalis JG, Dohanics J. Vasopressin and oxytocin secretion in chronically hyposmolar rats. Am J Physiol 1991; 261: R1028-R1038.
32. Brimble MJ, Dyball RE. Characterization of the responses of oxytocin- and vasopressin-secreting neurones in the supraoptic nucleus to osmotic stimulation. J Physiol 1977; 271: 253-271.
33. Johnson AK, Buggy J. Periventricular preoptic-hypothalamus is vital for thirst and normal water economy. Am J Physiol 1978; 234: R122-R129.
34. Gruber KA, Wilkin LD, Johnson AK. Neurohypophyseal hormone release and biosynthesis in rats with lesions of the anteroventral third ventricle (AV3V) region. Brain Res 1986; 378: 115-119.
35. Leng G, Blackburn RE, Dyball REJ, Russell JA. Role of anterior peri-third ventricular structures in the regulation of supraoptic neuronal activity and neurohypophysial hormone secretion in the rat. J Neuroendocrinol 1989; 1: 35-46.
36. McKinley MJ, Mathai ML, McAllen RM, et al. Vasopressin secretion: osmotic and hormonal regulation by the lamina terminalis. J Neuroendocrinol 2004; 16: 340-347.
37. Geddes BJ, Parry LJ, Summerlee AJ. Brain angiotensin-II partially mediates the effects of relaxin on vasopressin and oxytocin release in anesthetized rats. Endocrinology 1994; 134: 1188-1192.
38. Chriguer RS, Antunes-Rodrigues J, Franci CR. Atrial natriuretic peptide mediates oxytocin secretion induced by osmotic stimulus. Brain Res Bull 2003; 59: 505-511.
39. Fitzsimons JT. Angiotensin, thirst, and sodium appetite. Physiol Rev 1998; 78: 583-686.
40. Sunn N, Egli M, Burazin TC, et al. Circulating relaxin acts on subfornical organ neurons to stimulate water drinking in the rat. Proc Natl Acad Sci USA 2002; 99: 1701-1706.
41. Egan G, Silk T, Zamarripa F, et al. Neural correlates of the emergence of consciousness of thirst. Proc Natl Acad Sci USA 2003; 100: 15241-15246.
42. Antunes-Rodrigues J, Ramalho MJ, Reis LC, et al. Lesions of the hypothalamus and pituitary inhibit volume-expansion-induced release of atrial natriuretic peptide. Proc Natl Acad Sci USA 1991; 88: 2956-2960.
43. Jankowski M, Marcinkiewicz M, Bouanga JC, Gutkowska J. Changes of atrial natriuretic peptide in rat supraoptic neurones during pregnancy. J Neuroendocrinol 2004; 16: 441-449.
44. Cunningham JT, Bruno SB, Grindstaff RR, et al. Cardiovascular regulation of supraoptic vasopressin neurons. Prog Brain Res 2002; 139: 257-273.
45. McCann MJ, Verbalis JG, Stricker EM. LiCl and CCK inhibit gastric emptying and feeding and stimulate OT secretion in rats. Am J Physiol 1989; 256: R463-R468.
46. Buller KM, Day TA. Involvement of medullary catecholamine cells in neuroendocrine responses to systemic cholecystokinin. J Neuroendocrinol 1996; 8: 819-824.
47. Brunton PJ, Meddle SL, Ma S, Ochedalski T, Douglas AJ, Russell JA. Endogenous opioids and attenuated hypothalamic-pituitary-adrenal axis responses to immune challenge in pregnant rats. J Neurosci 2005; 25: 5117-5126.
48. Bull PM, Douglas AJ, Russell JA. Opioids and coupling of the anterior peri-third ventricular input to oxytocin neurones in anaesthetized pregnant rats. J Neuroendocrinol 1994; 6: 267-274.
49. Koehler EM, McLemore GL, Martel JK, Summy-Long JY. Response of the magnocellular system in rats to hypovolemia and cholecystokinin during pregnancy and lactation. Am J Physiol 1994; 266: R1327-R1337.
50. Douglas AJ, Dye S, Leng G, Russell JA, Bicknell RJ. Endogenous opioid regulation of oxytocin secretion through pregnancy in the rat. J Neuroendocrinol 1993; 5: 307-314.
51. Russell JA, Douglas AJ, Bull PM, Pumford KM, Bicknell RJ, Leng G. Pregnancy and opioid interactions with the anterior perithird ventricular input to magnocellular oxytocin neurones. Prog Brain Res 1992; 91: 41-53.
52. Douglas AJ, Neumann I, Meeren HK, et al. Central endogenous opioid inhibition of supraoptic oxytocin neurons in pregnant rats. J Neurosci 1995; 15: 5049-5057.
53. Brunton PJ, Sabatier N, Leng G, Russell JA. Suppressed oxytocin neuron responses to immune challenge in late pregnant rats: a role for endogenous opioids. Eur J Neurosci 2006; 23: 1241-1247.
54. Brunton PJ, Russell JA. Keeping oxytocin neurons under control during stress in pregnancy. Prog Brain Res 2008; 170: 365-377.
55. Russell JA, Leng G, Douglas AJ. The magnocellular oxytocin system, the fount of maternity: adaptations in pregnancy. Front Neuroendocrinol 2003; 24: 27-61.
56. Di S, Malcher-Lopes R, Marcheselli VL, Bazan NG, Tasker JG. Rapid glucocorticoid-mediated endocannabinoid release and opposing regulation of glutamate and gamma-aminobutyric acid inputs to hypothalamic magnocellular neurons. Endocrinology 2005; 146: 4292-4301.
57. Sabatier N, Leng G. Presynaptic actions of endocannabinoids mediate alpha-MSH-induced inhibition of oxytocin cells. Am J Physiol Regul Integr Comp Physiol 2006; 290: R577-R584.
58. Koksma JJ, van Kesteren RE, Rosahl TW, et al. Oxytocin regulates neurosteroid modulation of GABA(A) receptors in supraoptic nucleus around parturition. J Neurosci 2003; 23: 788-797.
59. Concas A, Mostallino MC, Porcu P, et al. Role of brain allopregnanolone in the plasticity of gamma-aminobutyric acid type A receptor in rat brain during pregnancy and after delivery. Proc Natl Acad Sci USA 1998; 95: 13284–13289.
60. Voisin DL, Simonian SX, Herbison AE. Identification of estrogen receptor-containing neurons projecting to the rat supraoptic nucleus. Neuroscience 1997; 78: 215-228.
61. Somponpun SJ, Holmes MC, Seckl JR, Russell JA. Modulation of oestrogen receptor-beta mRNA expression in rat paraventricular and supraoptic nucleus neurones following adrenal steroid manipulation and hyperosmotic stimulation. J Neuroendocrinol 2004; 16: 472-482.
62. Sladek CD, Somponpun SJ. Estrogen receptors: their roles in regulation of vasopressin release for maintenance of fluid and electrolyte homeostasis. Front Neuroendocrinol 2008; 29: 114-127.
63. Somponpun SJ, Johnson AK, Beltz T, Sladek CD. Estrogen receptor-alpha expression in osmosensitive elements of the lamina terminalis: regulation by hypertonicity. Am J Physiol Regul Integr Comp Physiol 2004b; 287: R661-R669.
64. Brailoiu E, Dun SL, Brailoiu GC, et al. Distribution and characterization of estrogen receptor G protein-coupled receptor 30 in the rat central nervous system. J Endocrinol 2007; 193: 311-321.
65. Brown CH, Brunton PJ, Russell JA. Rapid estradiol-17beta modulation of opioid actions on the electrical and secretory activity of rat oxytocin neurons in vivo. Neurochem Res 2008; 33: 614-623.
66. Francis K Meddle SL, Bishop VR, Russell JA. Progesterone receptor expression in the pregnant and parturient rat hypothalamus and brainstem. Brain Res 2002; 927: 18-26.
67. Bull PM, Brown LA, Russell JA. Ovarian hormones and reduced osmoresponsiveness of oxytocin neurones in urethane-anaesthetized pregnant rats. J Physiol 1996; 495.P: P113-P114.
68. Douglas AJ, Johnstone H, Brunton P, Russell JA. Sex-steroid induction of endogenous opioid inhibition on oxytocin secretory responses to stress. J Neuroendocrinol 2000; 12: 343-350.
69. Barron WM, Schreiber J, Lindheimer MD. Effect of ovarian sex steroids on osmoregulation and vasopressin secretion in the rat. Am J Physiol 1986; 250: E352-E361.
70. Bathgate RA, Ivell R, Sherwood OD, Summers RJ. International Union of Pharmacology LVII: Recommendations for the Nomenclature of Receptors for Relaxin Family Peptides. Pharmacol Rev 2006; 58: 7-31.
71. Sherwood OD, Crnekovic VE, Gordon WL, Rutherford JE. Radioimmunoassay of relaxin throughout pregnancy and during parturition in the rat. Endocrinology 1980; 107: 691-698.
72. Zhao L, Roche PJ, Gunnersen JM, et al. Mice without a functional relaxin gene are unable to deliver milk to their pups. Endocrinology 1999; 140: 445-453.
73. Zhao S, Malmgren CH, Shanks RD, Sherwood OD. Monoclonal antibodies specific for rat relaxin. VIII. Passive immunization with monoclonal antibodies throughout the second half of pregnancy reduces water consumption in rats. Endocrinology 1995; 136: 1892-1897.
74. Hornsby DJ, Wilson BC, Summerlee AJ. Relaxin and drinking in pregnant rats. Prog Brain Res 2001; 133: 229-240.
75. Sinnayah P, Burns P, Wade JD, Weisinger RS, McKinley MJ. Water drinking in rats resulting from intravenous relaxin and its modification by other dipsogenic factors. Endocrinology 1999; 140: 5082-5086.
76. Thornton SN, Fitzsimons JT. The effects of centrally administered porcine relaxin on drinking behaviour in male and female rats. J Neuroendocrinol 1995; 7: 165-169.
77. Way SA, Leng G. Relaxin increases the firing rate of supraoptic neurones and increases oxytocin secretion in the rat. J Endocrinol 1992; 132: 149-158.
78. Parry LJ, Poterski RS, Summerlee AJ. Effects of relaxin on blood pressure and the release of vasopressin and oxytocin in anesthetized rats during pregnancy and lactation. Biol Reprod 1994; 50: 622-628.
79. Summerlee AJ, Hornsby DJ, Ramsey DG. The dipsogenic effects of rat relaxin: The effect of photoperiod and the potential role of relaxin on drinking in pregnancy. Endocrinology 1998; 139: 2322-2328.
80. Silvertown JD, Fraser R, Poterski RS, Geddes B, Summerlee AJ. Central effects of long-term relaxin expression in the rat. Ann N Y Acad Sci 2005; 1041: 216-222.
81. Weisinger RS, Burns P, Eddie LW, Wintour EM. Relaxin alters the plasma osmolality-arginine vasopressin relationship in the rat. J Endocrinol 1993; 137: 505-510.
82. Osheroff PL, Phillips HS. Autoradiographic localization of relaxin binding sites in rat brain. Proc Natl Acad Sci USA 1991; 88: 6413-6417.
83. Ma S, Shen PJ, Burazin TC, Tregear GW, Gundlach AL. Comparative localization of leucine-rich repeat-containing G-protein-coupled receptor-7 (RXFP1) mRNA and [33P]-relaxin binding sites in rat brain: restricted somatic co-expression a clue to relaxin action? Neuroscience 2006; 141: 329-344.
84. Neumann I, Landgraf R, Takahashi Y, Pittman QJ, Russell JA. Stimulation of oxytocin release within the supraoptic nucleus and into blood by CCK-8. Am J Physiol 1994; 267: R1626-R1631.
85. Brunton PJ, Bales J, Russell JA. Neuroendocrine stress but not feeding responses to centrally administered neuropeptide Y are suppressed in pregnant rats. Endocrinology 2006; 147: 3737-3745.
86. Leng G, Dye S, Bicknell RJ. Kappa-opioid restraint of oxytocin secretion: plasticity through pregnancy. Neuroendocrinology 1997; 66: 378-383.
87. Wolak ML, DeJoseph MR, Cator AD, Mokashi AS, Brownfield MS, Urban JH. Comparative distribution of neuropeptide Y Y1 and Y5 receptors in the rat brain by using immunohistochemistry. J Comp Neurol 2003; 464: 285-311.
88. Ellacott KL, Halatchev IG, Cone RD. Characterization of leptin-responsive neurons in the caudal brainstem. Endocrinology 2006; 147: 3190-3195.
89. Hakansson ML, Brown H, Ghilardi N, Skoda RC, Meister B. Leptin receptor immunoreactivity in chemically defined target neurons of the hypothalamus. J Neurosci 1998; 18: 559-572.
90. Kawai M, Yamaguchi M, Murakami T, Shima K, Murata Y, Kishi K. The placenta is not the main source of leptin production in pregnant rat: gestational profile of leptin in plasma and adipose tissues. Biochem Biophys Res Commun 1997; 240: 798-802.
91. Grattan DR, Ladyman SR, Augustine RA. Hormonal induction of leptin resistance during pregnancy. Physiol Behav 2007; 91: 366-374.
92. Ladyman SR, Grattan DR. Region-specific reduction in leptin-induced phosphorylation of signal transducer and activator of transcription-3 (STAT3) in the rat hypothalamus is associated with leptin resistance during pregnancy. Endocrinology 2004; 145: 3704-3711.