Developmental Brain Research 143 (2003) 119–128www.elsevier.com/ locate/devbrainres

Research report

S uckling and genital stroking induces Fos expression in hypothalamicoxytocinergic neurons of rabbit pups

a , a,b c,d*Mario Caba , Maria J. Rovirosa , Rae Silvera ´ ´Lab. Biologıa de la Reproduccion, IIB, Universidad Veracruzana, A.P. 114, Xalapa, Ver. 91000,Mexico

b ´ ´Centro de Investigacion en Reproduccion Animal, CINVESTAV-UAT, Tlaxcala, MexicocDepartment of Psychology, Barnard College and Columbia University, New York, NY 10027,USA

dDepartment of Anatomy and Cell Biology, College of Physicians and Surgeons, Columbia University, New York, NY 10032,USA

Accepted 25 February 2003

Abstract

Maternal behaviour in the rabbit is unusual among mammals because the doe visits her litter to nurse once every 24 h. In the presentstudy we examined the consequences of milk intake on oxytocinergic (OT) and vasopressinergic (AVP) neurons of the supraoptic (SON)and paraventricular (PVN) nuclei of 7-day-old pups before suckling, after suckling and following anogenital stroking in un-nursed pups.To determine neuronal activation we assessed the expression of the Fos protein combined with antibodies against OT and AVP at twolevels in the SON (supraoptic rostral, SOr, and supraoptic retrochiasmatic, SOrch), and three levels in the PVN (anterior, PVab; medialPVm and caudal, PVc). Daily nursing bouts lasted only 22866 s throughout the observed 7 days, and pups ingested up to 34.9569.0% oftheir body weight in milk on day 7, the day of perfusion. Suckling induced a significant increase in the number of double-labeled Fos/OTcells in both subdivisions of the SON (P,0.01) and in PVab and PVm (P,0.01). The effect in the SON was related to suckling, as it wasnot seen in stroked, un-nursed pups, which showed Fos increases only in PVab and PVm. All regions in the SON and PVN showedsignificant increases in the number of Fos/AVP neurons after suckling or stroking but, contrary to OT, the number of double-labeledFos/AVP cells was very low. In conclusion, our results show that the oxytocinergic system of the SON and PVN is differentially activatedby suckling of milk and anogenital stroking, and that the vagal-hypothalamic axis is mature in 7-day-old rabbits. 2003 Elsevier B.V. All rights reserved.

Keywords: Food-intake; Vasopressin; Gastric distension; Maternal behavior; Supraoptic nucleus; Paraventricular nucleus

1 . Introduction [9]. Chronobiological studies demonstrate that, like theirmother, pups have a daily anticipatory behavioural arousal

Behavioral studies in the rabbit, both in the field[4] and which is established between days 1 and 5 of life and lastsunder laboratory conditions[11,15,41] indicate that this for 1–3 h prior to the daily suckling episode[16,17]. Inspecies has an unusual pattern of maternal care. Shortly spite of the short duration of the nursing bouts, pups canafter parturition the doe leaves the litter in the nest, ingest as much as 30% of their body weight in milk[8,41],previously constructed, and about 24 h later she returns to resulting in abrupt gastric distension. The neurobiologicalnurse for a brief period of 3–5 min. Even if the mother has consequences of daily suckling have been little explored.continuous access to the nest box she exhibits this nursing Using the product of the gene c-fos as a marker ofrhythm, both in light–dark[16] and continuous light[17] neuronal activation it was found that daily sucklingconditions. Rabbit pups are born altricial (i.e., no fur, with induces Fos expression in the supraoptic nucleus (SON)eyes and outer ears sealed;[15]), nevertheless they readily and in the paraventricular nucleus (PVN) of 7-day-oldlocate and grasp the mother’s nipples to suck milk guided rabbits[1]. Because oxytocinergic neurons are activated inby the emission of a nipple pheromone from her ventrum adult rats by gastric distension with a balloon[27] and by

ingestion of a satiating meal[37] we investigated whetheroxytocin (OT) neurons become active following nursing in*Corresponding author. Tel.:152-228-812-5757; fax:152-228-812-newborn rabbits. Additionally, we explored whether simi-5746.

E-mail address: mcaba@uv.mx(M. Caba). lar changes occurred in vasopressinergic (AVP) neurons, as

0165-3806/03/$ – see front matter 2003 Elsevier B.V. All rights reserved.doi:10.1016/S0165-3806(03)00064-6

120 M. Caba et al. / Developmental Brain Research 143 (2003) 119–128

a result of the sudden fluid intake following their usual in phosphate buffer (PB, pH 7.4). The brains were24-h fluid deprivation. removed immediately after perfusion, cryoprotected suc-

In the present study we used Fos to determine the cessively in 10, 20 and 30% sucrose in PB, and were thenproportion of activated OT- and AVP-immunoreactive (-ir) sectioned coronally at 50mm with a cryostat (Micron).neurons before and after a suckling episode. We selected Serial sections were collected in PB from the level of the7-day-old rabbit pups as subjects as they are well entrained opening of the organum vasculosum of lamina terminalisto the scheduled feeding regime at this age[16,17]. to the mammillary bodies. Every two of four sections were

used for double-labeling of Fos and OT or Fos and AVP asdescribed below, following protocols previously estab-

2 . Methods lished for Fos protein[7] and OT- and AVP-ir[6] in rabbitbrain. Tissue was washed in PB four times, 5 min each, to

2 .1. Subjects and housing remove excess aldehydes and then exposed for 10 min in0.5% hydrogen peroxide solution to eliminate endogenous

New Zealand white female rabbits bred in our colony in peroxidase activity. Nonspecific tissue antibody reactions´Xalapa, Mexico, were maintained under controlled light were blocked by placing the sections in 3% normal horse

(13 h light:11 h dark) and temperature (2362 8C) con- serum (Vector Labs, Burlingame, CA, USA) for 1 h atditions and were provided with rabbit pellets (Purina) and room temperature. Sections were then incubated for 48 h atwater ad libitum. Females were mated and housed in- 48C in the polyclonal Fos antibody SC-52 G diluted atdividually in wire mesh cages inside the rabbit colony, and 1:5000 (Santa Cruz Biotechnology, Santa Cruz, CA) in 3%from day 28 of pregnancy until delivery were monitored normal horse serum with 0.3% Triton X-100 (Sigma, St.daily. Each female was provided a wooden box (50 cm Louis, MO, USA). Tissue was placed in biotinylated horselong330 cm wide332 cm high) with a round (24 cm anti-goat serum (1:200, Vector Labs) for 1 h and afterdiameter) opening on one side for the construction of the washes in PB, it was incubated in avidin–biotin–HRPmaternal nest. At parturition pups were separated from the complex (1:250, Vector Labs). Fos antibody–peroxidasedoe, litter size was adjusted to four to five and they were complex was stained with a solution of 0.05% diaminoben-kept warm in an acrylic box with material from the nest. zidine (Polysciences, Warrington, PA) in the presence of

nickel sulfate (10 mg/ml, Fisher Scientific, Pittsburgh, PA,2 .2. Experimental groups USA), cobalt chloride (10 mg/ml, Fisher Scientific) and

0.01% hydrogen peroxide, which produced a black–purpleFrom days 1 to 6 pups were maintained away from the precipitate. After 10 min the tissue was transferred to PB

maternal nest and returned to it once every day for nursing to stop the reaction. Following three washes in PB,at 10:00 h. As previously established[10,11,15], in order sections were sequentially incubated in antibodies againstto track their daily milk intake, pups were stimulated to OT or AVP. Tissue was incubated for 24 h at 48C in theurinate by gently stroking the anogenital area and weighed monoclonal antibody MAB AI-28 to OT or MAB IIID-7 tobefore and after being placed in the maternal nest. The AVP (gift of A.-J. Silverman), diluted at 1:1000 in 3%duration of nursing bouts was quantified by determining normal horse serum with 0.3% Triton X-100 (Sigma). Thethe time of entrance into and exit from the nest box for MAB AI-28 and MAB IID-7 antibodies have been welleach mother. On day 7, the following experimental groups characterized and tested in rabbit[6,14]. Sections were(n54 pups per group) were established: Group 1, Before then incubated sequentially in biotinylated horse anti-Suckling (not-nursed and sacrificed at 10:00 h). Group 2, mouse serum IgG (1:200, Vector Laboratories) and avidin–After suckling (sacrificed at 11:30 h, 90 min after nursing). biotin–HRP complex (1:250, Vector Laboratories). HRPGroup 3, Anogenital Stroking (not-nursed pups, that label was demonstrated by using 0.05% diaminobenzidinereceived anogenital stroking to induce urination and were and 0.01% hydrogen peroxide which produced a brownsacrificed at 11:30 h, 90 min after their scheduled time of cytoplasmic precipitate. Sections were mounted ontonursing). Subjects of Group 3 were identical to nursed gelatin-subbed slides, dehydrated and cleared in Hemo-Depups in Group 2, but they were not nursed, thereby serving (Fisher Scientific) and then coverslipped with Permount. Inas controls for the possible effect of stroking/urination. all cases, tissue sections from two of the three groups wereRabbit pups normally urinate by themselves and the incubated together. Control sections were processed asmother does not brood them, cleans them little if at all and above but the primary antibody was omitted.does not retrieve pups that stray from the nest[15].

2 .4. Analysis of tissue2 .3. Perfusion and immunocytochemistry

Sections were coded, so that the observer was blind toRabbits were anesthetized with an overdose of sodium the physiological status of the animal, and were examined

pentobarbital i.p. and were perfused transcardially with under bright-field illumination on an Olympus BX41saline solution (0.9%), followed by 4% paraformaldehyde microscope at3100 and3200 magnification. OT- and

M. Caba et al. / Developmental Brain Research 143 (2003) 119–128 121

AVP-ir was identified as a brown precipitate in the 3 . Resultsperikarya, whereas Fos-ir was identified by black nuclearstaining; double-labeled neurons had brown cytoplasm and3 .1. Milk-intake from days 1 to 7black nuclei. Immunostaining was completely absent intissue processed without primary antibodies. The mean daily duration of nursing was 22866 s

The anatomical identification of OT- and AVP-ir neu- (Mean6S.E.) over the observed 7 days. Despite the brevityrons was made on the basis of our previous experience of these episodes pups increased in body weight fromwith the rabbit hypothalamus[5,6], and they were counted 62.462.9 g on day 1 to 98.965.1 g on day 6bilaterally in the supraoptic and paraventricular nucleus. In (Mean6S.E.). This amount of milk represented an increas-the SON we distinguished two regions, the principal rostral ingly higher percentage of the pups’ body weight, such that(SOr) and the retrochiasmatic (SOrch) subdivision, as on day 7 they ingested on average 34.9569.0%previously reported in the rabbit by our laboratory[6] and (Mean6S.E.) of their body weight (Fig. 1).others[31] and counts were performed at levels A3 (SOr)and A2 (SOrch) of the rabbit atlas[29]. In order to 3 .2. Distribution of hypothalamic OT and AVP neuronssystematically sample the PVN we selected three regionsfor analysis: anterior (anterobasal extension, PVab), medial OT- and AVP-ir perikarya were concentrated in the SON(PVm) and caudal (PVc) which corresponds to levels A2, and PVN in the three groups studied, as previouslyA1 and AP0, respectively, in the rabbit atlas[29]. OT- and mentioned. In the SON we identified two different popula-AVP-ir neurons were counted on both sides in the supraop- tions, a rostral one (SOr) and a retrochiasmatic onetic and paraventricular nucleus. In every subject we (SOrch). The rostral portion lies along the base of thecounted single-labeled OT and AVP neurons and double- forebrain, dorsal and lateral to the optic chiasm, and thelabeled OT/Fos and AVP/Fos neurons in matched sections retrochiasmatic subdivision is found medial to the opticat the above-mentioned levels. Analysis of variance tract along the base of the brain; both portions are joined(ANOVA) was used to determine whether there were by a few scattered immunoreactive perikarya. In the rostraldifferences in cell number among the conditions studied. subdivision OT- and AVP-ir neurons were intermingled butThis was followed by a post hoc analysis using the Tukey- OT neurons predominate in the dorsal part of the nucleusKramer test (GB-STAT vers. 6.0). Probability levels of while most AVP neurons formed dense clusters against theP,0.05 were considered significant. Values given are ventral portion, close to the optic chiasm. In the retroch-means6S.E. iasmatic portion both OT- and AVP-ir neurons were

Fig. 1. Milk intake over the study period expressed in relation to daily body weight. On day 7 pups ingested on average 34.9569.0% of their body weightin milk.

122 M. Caba et al. / Developmental Brain Research 143 (2003) 119–128

T able 1Mean number (6S.E.) of single OT-IR neurons, double-labeled Fos/OT and percent double-labeled Fos/OT neurons across brain regions and experimentalconditions

Nucleus Condition Total neurons Fos/OT % Double(mean6S.E.) labeled

SOr Before Suckling 56.064.33 0.5060.25 0.89After Suckling 51.564.44 40.5062.25 78.6Anogenital Stroking 52.2561.88 8.2565.10 15.7

SOrch Before Suckling 50.563.09 0.7560.25 1.4After Suckling 35.2566.08 32.2565.94 91.4Anogenital Stroking 38.565.83 5.2563.35 13.6

PVab Before Suckling 40.7561.37 2.561.65 6.1After Suckling 56.25610.33 30.561.5 54.2Anogenital Stroking 41.065.11 17.2564.47 42.6

PVm Before Suckling 73.2569.11 6.560.28 8.9After Suckling 60.0610.17 5562.04 91.6Anogenital Stroking 87.562.10 40.560.28 46.2

PVc Before Suckling 51.75610.76 24.066.81 46.3After Suckling 39.5611.44 27.7562.87 70.1Anogenital Stroking 42.063.71 19.2565.45 45.8

Abbreviations same as inFig. 2.

interspersed along the rostrocaudal extension of this sub- double Fos/OT and Fos/AVP, depending on the physiolog-division and AVP-ir neurons did not form dense aggre- ical condition of the animal.gates, as they did in the rostral portion. The PVN had aprominent rostral anterobasal extension dorsal to thesuprachiasmatic nucleus. At the main and caudal levels the3 .3. Changes in the number of Fos /OT neuronsneurons formed clusters that clearly delimited the nucleus.OT- and AVP-ir neurons were found in these three levels. The number of double-labeled Fos/OT neurons variedFurthermore double-labeled Fos/OT and Fos/AVP cells among experimental conditions, both in the SOr andwere seen throughout the nucleus without any apparent SOrch. In the 10:00 h not-nursed subjects, only 0.89% ofregional localization. The total number of OT and AVP-ir OT neurons in the SOr were transcriptionally activated, ascells did not differ in any region of both SON and PVN of indicated by Fos protein. In the 11:30 h nursed subjectsthe three conditions studied (Tables 1 and 2). In both Fos/OT neurons increased to 78.6% (P,0.01) versus thenuclei, there were significant changes in the number of 10:00 h not nursed and versus the 11:30 h not-nursed

T able 2Mean number (6S.E.) of single AVP-IR neurons, double-labeled Fos/AVP and percent double-labeled Fos/AVP neurons across brain regions andexperimental conditions

Nucleus Condition Total neurons Fos/AVP % Double(mean6S.E.) labeled

SOr Before Suckling 40.2564.26 0 0.0After Suckling 31.7564.64 1.7560.48 5.5Anogenital Stroking 49.7563.94 4.560.86 9.0

SOrch Before Suckling 49.061.41 0 0.0After Suckling 54.061.47 6.560.96 12.0Anogenital Stroking 45.062.79 4.560.96 8.8

PVab Before Suckling 29.7564.76 0 0.0After Suckling 26.062.27 1.560.29 5.7Anogenital Stroking 27.562.84 2.560.29 9.0

PVm Before Suckling 45.065.90 0 0.0After Suckling 41.563.06 5.2560.48 12.8Anogenital Stroking 48.2568.98 1.060.58 5.1

PVc Before Suckling 20.2563.75 0 0.0After Suckling 31.564.48 2.060.41 6.3Anogenital Stroking 28.566.13 0.560.5 1.7

Abbreviations same as inFig. 2.

M. Caba et al. / Developmental Brain Research 143 (2003) 119–128 123

groups (Table 1, Figs. 2 and 3). The same trend was found the 11:30 h not-nursed group (9.0%;P,0.01) relative toin the SOrch where the percent of Fos/OT double-labeled the 10:00 h not-nursed group. In the SOrch the proportioncells increased to 91.4% at 11:30 h nursed pups (P,0.01 of Fos/AVP cells increased in the 11:30 h nursed groupversus 10:00 h not-nursed and 11:30 h not-nursed pups; (12.0%;P,0.01) and in the 11:30 h not-nursed groupTable 1, Figs. 2 and 3). In the PVab there was a significant (8.8%;P,0.01) relative to the 10:00 h group. A similarincrease in the number of Fos/OT-ir neurons both in the effect was observed in the PVab (10:00 h not-nursed,11:30 h nursed (P,0.01) and 11:30 h not-nursed groups P,0.01, versus 11:30 h nursed and 11:30 h not-nursed). In(P,0.05) versus 10:00 h not-nursed group (Table 1, Fig. the PVm and PVc the proportion of double-labeled neurons2). A similar trend was found in the PVm (P,0.01 10:00 h significantly increased only in the 11:30 h nursed subjectsnot nursed versus 11:30 h nursed and versus 11:30 h (P,0.01) versus 10:00 h not-nursed and versus 11:30 hnot-nursed groups;Table 1, Figs. 2 and 3). The distribution not-nursed pups.of double-labeled Fos/OT cells was throughout the nu-cleus as shown inFigs. 3 and 4.There were no significantchanges in the PVc (Table 1, Fig. 2). 4 . Discussion

3 .4. Changes in the number of Fos /AVP neurons The results of the present experiments provide insightinto the neural consequences of suckling, gastric distension

There were no Fos/AVP-ir neurons detected in the 10:00 and anogenital stimulation in the young rabbit. Here weh not-nursed group and few double-labeled cells in the report that ingestion of more than 30% of body weight inother two conditions (Table 2, Fig. 5). In the SOr there milk on nursing day 7, induced Fos nuclear immuno-was an increase in the percentage of double-labeled reactivity in OT-ir cells in the supraoptic and paraventricu-neurons in the 11:30 h nursed group (5.5%;P,0.05) and lar nuclei of the rabbit pup hypothalamus. Moreover,

Fig. 2. Difference in the mean number of Fos/OT double-labeled cells before and after suckling and after anogenital stroking. Note the increase ofFos/OT in both regions of the SON after suckling but not after anogenital stroking. In the PVab and PVm there is an increase in the number ofdouble-labelled Fos/OT cells both after suckling and after anogenital stroking. There is not change in the PVc. *Before suckling versus After suckling;1 & & , 11,&&After suckling versus Anogenital stroking; Before suckling versus Anogenital stroking.P,0.05;** P,0.01. Abbreviations: SOr, supraopticrostral; SOrch, supraoptic retrochiasmatic; PVab, paraventricular nucleus anterobasal extension; PVm, paraventricular nucleus medial part; PVc,paraventricular nucleus caudal part.

124 M. Caba et al. / Developmental Brain Research 143 (2003) 119–128

Fig. 3. Photomicrographs of sections immunostained for Fos (black nuclei) and OT (brown cytoplasm) in the PVm (A–C) and SOr (D–F), before suckling(A,D), after suckling (B,E) and after anogenital stroking (C,F). There is no expression of Fos in OT cells (open arrow) before suckling, but after sucklingmost cells are double-labeled Fos/OT (arrows), both in the PVm and SOr. Anogenital stroking also induced Fos in OT cells in the PVm, but not in the SOr.Magnification bar550 mm. Abbreviations same as inFig. 2.

anogenital stroking also induced Fos in OT-ir cells in the tonomic nervous system[33]. In the rat oxytocinergicPVN but not in the SON. neurons in the anterior medial and lateral parvocellular part

Previous studies indicate that gastric distension in the rat of the PVN project to the nucleus of the solitary tract andinduces an abrupt increase in firing frequency in putative dorsal motor nucleus of the vagus[30,33]. However,OT neurons of the SON[27] and PVN[34] as well as Fos detailed morphometric studies[31] demonstrate that theexpression in the PVN[26]. The activation of the oxy- cytoarchitecture of the PVN in the rabbit differs con-tocinergic system in our subjects can potentially lead to an siderably from that of the rat where eight differentarray of autonomic and neuroendocrine effects. magnocellular and parvocellular cell groups have been

Neurons of both the rostral and retrochiasmatic portions recognized[33]. In the rabbit such subdivisions are notof the SON project primarily to the neural lobe[32]. The obvious and neurons of various sizes are interspersedneurons of the PVN project to the neural lobe, median throughout the nucleus[31]. Moreover, there are no tracteminence and to several sites associated with the au- tracing studies in this species aimed at identifying projec-

M. Caba et al. / Developmental Brain Research 143 (2003) 119–128 125

peptide, it is well known that central OT is concomitantlyreleased by gastric distension or anorectic treatments[38].The significant elevation in the number of double-labeledFos/OT neurons in the SON and PVN also suggests that inour 7-day-old pups the vagal-hypothalamic axis is maturesimilar as has been found in 10-day-old[23] but not in2-day-old [28] rats. It has been proposed that plasma OTrelease is strongly correlated with inhibition of food-intakein adult rats, i.e., a process promoting satiety[25,38]. Themechanism for this effect is not clear as peripheraladministration of OT has little or no effect on food intake[38]. In addition to an increase in the number of double-labeled Fos/OT neurons in the SON following suckling, asimilar increase in the PVN also occurred. This suggests apossible intracerebral release of OT by parvocellularneurons. Although we did not recognize particular mag-nocellular or parvocellular regions in the PVN, it ispossible that, as in rats, the PVm and PVc regions of oursubjects contain parvocellular neurons projecting to theabove-mentioned brainstem regions. As shown (Fig. 4) inthe PVm the double-labeled Fos/OT neurons are foundthroughout the extent of the nucleus where, according tomorphometric analysis in adult rabbits, a mixture ofparvocellular and magnocellular neurons is observed[31].Therefore, we may speculate that the observed increase inthe number of Fos/OT neurons in the putative parvocellu-lar region of the PVm and PVc can be related to theoxytocinergic modulation of vagal digestive motor func-tions proposed for rats[19].

Our study indicates another possible function of OT inthe PVN, related to the neural consequences of anogenitalstroking. As shown inFig. 2, anterior and medial regionsof the PVN showed significant increases in the proportionof Fos/OT cells following either anogenital stroking ornursing, but in contrast to nursed subjects, anogenitalstroking alone did not activate oxytocinergic neurons in theFig. 4. Camera lucida drawings of a representative subject after suckling

(A) and after anogenital stroking (B) mapping the distribution of single SON. It is possible that the increase in the number ofOT-IR (empty circles) and double-labeled Fos/OT-IR (filled circles) cells Fos/OT in the PVN was due to the stress of manipulation.in the PVm at the level where the analysis was made. Each symbol There is growing evidence that central OT is involved inrepresents one cell. Note that the distribution of double Fos/OT cells is

stress responses and is released by some forms of stress,throughout the whole nucleus. The enclosed rectangle is the area shownsuch as immobilization[21], shaking [24], and forcedin the photomicrograph ofFig. 3B,C. IIIv, third ventricle.swimming[39], all of which include a long exposure to thestressor. However, opposite effects have also been found.

tions from specific PVN populations to the pituitary or It has been reported in the rat that massage-like stroking ofother nervous system regions. the abdomen in conscious animals and brushing of the leg

As we observed a large increase in double-labeled Fos/ under anesthesia is followed by increased oxytocin plasmaOT neurons both in the SON and PVN it is possible that levels and lowered blood pressure[35]. In neonatal ratsmilk ingestion in our subjects produced an elevation of stroking of the anogenital region to stimulate urination, asoxytocin in plasma as occurs in 10-day-old rats following in our subjects, did not elevate ACTH but even reversedgastric distension[23]. This proposed peripheral release of the stress-induced ACTH response[36]. From these dataOT seems to be related to milk intake in both subdivisions we may speculate that the gentle anogenital stroking in ourof the SON as the percentage of Fos/OT neurons did not subjects, as well as the social contact during suckling,increase in not-nursed anogenital stroked animals (Fig. 2, activates the oxytocinergic system in the PVN, whichTable 1). Although on the basis of morphological evidence could, in turn, be related to the proposed positive socialin the rat the magnocellular OT neurons of the SON would benefits and bonding associated with oxytocin in otherbe more involved in the proposed peripheral release of the species[35,40], and also growth-promoting effects[3].

126 M. Caba et al. / Developmental Brain Research 143 (2003) 119–128

Fig. 5. Differences in the mean number of Fos/AVP double-labeled cells before and after suckling and after perigenital stroking. After suckling there wasan increase in the number of Fos/AVP cells in all regions of the SON and PVN, whereas perigenital stroked animals showed increases in both regions of

1 &the SON and the PVab. *Before suckling versus After suckling, After suckling versus Anogenital stroking, Before suckling versus Anogenital stroking.& ,11,&&**P,0.05; P,0.01. Abbreviations: Same as inFig. 2.

Experiments in the rabbit demonstrate that the physical lowing suckling or anogenital stroking, however, a small,contact between mother and young across the first 7 days but significant increase in the number of Fos/AVP neuronspostpartum plays a major role in maintaining maternal was found in both the SON and the PVN. It is unlikely thatresponsiveness[12]. Although the role of oxytocin on mechanical gastric distension per se activates AVP neuronsmaternal behavior in the rabbit has not been specifically as stomach distension provoked by inflating a balloonanalyzed, we reported that the number and size of OT-ir intragastrically produces an increase in the firing frequencyneurons in the SON, PVN and lateral hypothalamic area of OT but not of AVP neurons[27]. Moreover, theincrease during the last days of pregnancy and is even ingestion of large quantities of milk is more likely relatedhigher on day 1 postpartum[6]. It is possible that these to an inhibition of the magnocellular vasopressinergicchanges mediate the facilitation/maintenance of maternal system as this has been demonstrated following intake ofbehavior in the rabbit as hormonal treatments that induce large quantities of water[2]. However, the ingestion ofmaternal nest building provoke similar increases in OT-ir milk could indeed activate hypothalamic vasopressin neu-neurons as those seen in late pregnant rabbits[5]. Nothing rons in our subjects. In the rat the ingestion of carbohy-is known in the rabbit pup about the participation of OT in drate-sweetened, but not isotonic, solutions, induced c-fosthe establishment of affiliative bonding with their mother. protein in vasopressin and oxytocin neurons of the SONStudies in the rat demonstrate that oxytocin mediates and PVN, through an effect on putative oropharyngealacquisition of maternal-associated odor preference when osmorreceptors[20]. The authors reached this conclusioninfused centrally in young pups, an effect blocked by as intragastric feeding with a cannula did not induce theoxytocin antagonists[22]. In the rabbit the emission of the afore-mentioned effect. From these results we may specu-nipple-search pheromone by the mother is essential for late that the activation of both AVP and OT neuronssurvival, as anosmic pups are unable to suckle milk and following suckling can be due, at least partially, to thetherefore starve[9]. Possibly the detection of the phero- osmotic stimuli from the milk ingested.mone triggers the oxytocinergic system in the pup which Osmotic stimuli produced by milk ingestion, however,may be related to the acquisition of maternal-associated does not explain the increases in Fos in AVP neurons in theodor preference. SON and PVab following anogenital stroking in the non-

In contrast to Fos/OT neurons, Fos/AVP neurons were suckled animals (Group 3). It is well known that in the ratnot detected in any of the regions analyzed in the ex- thirst produced by saline injection induces Fos in AVPperimental group which was not-nursed at 10:00 h. Fol- neurons of the SON and PVN a finding that may explain

M. Caba et al. / Developmental Brain Research 143 (2003) 119–128 127

Rosenblatt, Maternal behavior in New Zealand white rabbits:Fos/AVP neurons in subjects of Group 3, which may startquantification of somatic events, motor patterns, and steroid plasmato experience thirst as they missed their only opportunitylevels, Physiol. Behav. 55 (1994) 1081–1089.

to drink fluids in 24 h. It is also possible that activation of ´ ´[12] G . Gonzalez-Mariscal, A.I. Melo, R. Chirino, P. Jimenez, C. Beyer,OT and AVP neurons could have resulted from perioral J.S. Rosenblatt, Importance of mother /young contact at parturitionstimulation associated with suckling. In the rat, injection of and across lactation for the expression of maternal behavior in

rabbits, Dev. Psychobiol. 32 (1998) 101–111.lidocaine into the vibrissal pads, or cutting of the infraorbi-[13] M .A. Hofer, A. Fisher, H. Shair, Effects of infraorbital nerve sectiontal nerve, abolishes or severely impairs nipple attachment

on survival, growth, and suckling behaviors of developing rats, J.[13,18]. Furthermore, oral stimulation provided by sham Comp. Physiol. Psychol. 1 (1981) 123–133.feeding in rats induces increased c-fos mRNA in the SON [14] A . Hou-Yu, A.T. Lamme, E.A. Zimmerman, A.-J. Silverman,and magnocellular division of the PVN, compared to Comparative distribution of vasopressin and oxytocin neurons in the

rat brain using a double-label procedure, Neuroendocrinology 44gastric feeding [20]. To distinguish the influence of(1986) 235–246.suckling from that of milk ingestion on the OT and AVP

[15] R . Hudson, H. Distel, Temporal pattern of suckling in rabbit pups: aneurons in rabbits, it will be necessary to explore the model of circadian synchrony between mother and young, in: S.M.effects of direct intragastric infusion of solutions, as well Reppert (Ed.), Development of circadian rhythmicity and photo-as the effect of suckling without obtaining any milk (by periodism in mammals, Perinatology Press, Ithaca, NY, 1989, pp.

83–102.mammary gland-duct ligation of the doe) in anogenitally[16] B . Jilge, The ontogeny of circadian rhythms in the rabbit, J. Biol.stimulated and control pups.

Rhythms 8 (1993) 247–260.[17] B . Jilge, Ontogeny of the rabbit’s circadian rhythms without an

external zeitgeber, Physiol. Behav. 58 (1995) 131–140.[18] C .A. Kenyon, S. Keeble, P. Cronin, The role of perioral sensation inA cknowledgements

nipple attachment by weanling rat pups, Dev. Psychobiol. 15 (1982)409–421.

This work was supported by a grant from CONACYT, [19] M .J. McCann, R.C. Rogers, Oxytocin excites gastric-related neuronsRef. [ 34652-N (to M.C.) in rat dorsal vagal complex, J. Physiol. (London) 428 (1990)

95–108.[20] N . Naimi, S. Rivest, I. Racotta, D. Richard, Neuronal activation of

the hypothalamic magnocellular system in response to oropharynge-R eferences al stimuli in the rat, J. Neuroendocrinol. 9 (1997) 329–340.

[21] T . Nakashima, T. Noguchi, T. Furukawa, M. Yamasaki, S. Makino,S. Miyata, T. Kiyohara, Brain oxytocin augments stress-induced[1] K . Allingham, C. von Saldern, P.A. Brennan, H. Distel, R. Hudson,long-lasting plasma adrenocorticotropic hormone elevation in rats,Endogenous expression of c-Fos in hypothalamic nuclei of neonatalNeurosci. Lett. 321 (2002) 161–164.rabbits coincides with their circadian pattern of suckling-associated

[22] E . Nelson, J. Panksepp, Oxytocin mediates acquisition of maternallyarousal, Brain Res. 783 (1998) 210–218.associated odor preferences in preweanling rat pups, Behav. Neuro-[2] B .H. Appelgren, T.N. Thrasher, L.C. Keil, D.J. Ramsay, Mechanismsci. 110 (1996) 583–592.of drinking-induced inhibition of vasopressin secretion in dehy-

[23] E .E. Nelson, J.R. Alberts, Y. Tian, J.G.Verbalis, Oxytocin is elevateddrated dogs, Am. J. Physiol. 261 (1991) R1226–R1233.in plasma of 10-day-old rats following gastric distension, Dev. Brain[3] E . Bjorkstrand, A.-L. Hulting, K. Uvnas-Moberg, Evidence for aRes. 111 (1998) 301–303.dual function of oxytocin in the control of growth hormone secretion

[24] T . Nishioka, J.A. Anselmo-Franci, M.F. Callahan, M. Morris, Stressin rats, Regul. Peptides 69 (1997) 1–5.increases oxytocin release within the hypothalamic paraventricular[4] S . Broekhuizen, J.L. Mulder, Differences and similarities in nursingnucleus, Brain Res. 781 (1998) 57–61.behaviour of hares and rabbits, Acta Zool. Fenn. 174 (1983) 61–63.

[25] B .R. Olson, M.D. Drutatosky, E.M. Stricker, J.G. Verbalis, Brain´[5] M . Caba, G. Gonzalez-mariscal, C. Beyer, Hormonal paradigmsoxytocin receptor antagonism blunts the effects of anorexigenicactivating maternal nest-building in ovariectomized rabbits modifytreatments in rats: evidence for central oxytocin inhibition of foodhypothalamic oxytocin immunoreactivity, in: Soc. Neurosci. 26thintake, Endocrinology 129 (1991) 785–790.Annual Meeting, Washington, DC, 1996, Abstr. 704.15.

[26] B .R. Olson, M. Freilino, G.E. Hoffman, E.M. Stricker, A.F. Sved,´ ´[6] M . Caba, R. Silver, G. Gonzalez-Mariscal, A. Jimenez, C. Beyer,J.G. Verbalis, C-Fos expression in rat brain and brainstem nuclei inOxytocin and vasopressin immunoreactivity in rabbit hypothalamusresponse to treatments that alter food intake and gastric motility,during estrus, late pregnancy, and postpartum, Brain Res. 720Mol. Cell. Neurosci. 4 (1993) 93–106.(1996) 7–16.

[27] L .P. Renaud, M. Tang, M.J. McCann, E.M. Stricker, J.G. Verbalis,[7] M . Caba, J. Bao, K.Y.F. Pau, C. Beyer, A. Gonzalez, R. Silver, H.G.Cholecystokinin and gastric distension activate oxytocinergic cellsSpies, Coitus-induced activation of c-fos and gonadotropin-releasingin rat hypothalamus, Am. J. Physiol. 253 (1987) R661–R665.hormone in hypothalamic neurons in female rabbits, Mol. Brain Res.

[28] L . Rinaman, G.E. Hoffman, E.M. Stricker, J.G. Verbalis, Exogenous78 (2000) 69–79.cholecystokinin activates cFos expression in medullary but not[8] V . Denenberg, R. Huff, S. Ross, P. Sawin, M. Zarrow, Maternalhypothalamic neurons in neonatal rats, Dev. Brain Res. 77 (1994)behavior in the rabbit: quantification of nest building, Anim. Behav.140–145.2 (1963) 494–499.

[29] C .H. Sawyer, J.W. Everett, J.D. Green, The rabbit diencephalons in[9] H . Distel, R. Hudson, Nipple-search performance by rabbit pups:stereotaxic coordinates, J. Comp. Neurol. 101 (1954) 801–824.changes with age and time of day, Anim. Behav. 32 (1984) 501–

[30] P .E. Sawchenko, L.W. Swanson, Immunohistochemical identification507.´ ´ of neurons in the paraventricular nucleus of the hypothalamus that[10] C . Escobar, R. Hudson, M. Martınez-Gomez, R. Aguilar-Roblero,

project to the medulla or to the spinal cord in the rat, J. Comp.Metabolic correlates of the circadian pattern of suckling-associatedNeurol. 205 (1982) 260–272.arousal in young rabbits, J. Comp. Physiol. 186 (2000) 33–38.

[31] S . Schimchowitsch, S.C. Moureau, F. Laurent, M.E. Stoeckel,[11] G . Gonzalez-Mariscal, V. Diaz-Sanchez, A.I. Melo, C. Beyer, J.S.

128 M. Caba et al. / Developmental Brain Research 143 (2003) 119–128

Distribution and morphometric characteristics of oxytocin- and [37] J .G. Verbalis, M.J. McCann, C.M. Mchale, E.M. Stricker, Oxytocinvasopressin-immunoreactive neurons in the rabbit hypothalamus, J. secretion in response to cholecystokinin and food: differentiation ofComp. Neurol. 285 (1989) 304–324. nausea from satiety, Science 232 (1986) 1417–1419.

[32] D .A. Sherlock, P.M. Field, G. Raisman, Retrograde transport of [38] J .G. Verbalis, R.E. Blackburn, G.E. Hoffman, E.M. Stricker, Estab-horseradish peroxidase in the magnocellular neurosecretory system lishing behavioral and physiological functions of central oxytocin:of the rat, Brain Res. 88 (1975) 403–413. insights from studies of oxytocin and ingestive behaviors, Adv. Exp.

[33] L .W. Swanson, P.E. Sawchenko, Paraventricular nucleus: a site for Med. Biol. 395 (1995) 209–225.the integration of neuroendocrine and autonomic mechanisms, [39] C .T. Wotjak, J. Ganster, G. Kohl, F. Holsboer, R. Landgraf, M.Neuroendocrinology 31 (1980) 410–417. Engelmann, Dissociated central and peripheral release of vasopres-

[34] Y . Ueta, H. Kannan, H. Yamashita, Gastric afferents to the paraven- sin, but not oxytocin, in response to repeated swim stress: newtricular nucleus in the rat, Exp. Brain Res. 84 (1991) 487–494. insights into the secretory capacities of peptidergic neurons, Neuro-

[35] K . Uvnas-Moberg, Oxytocin may mediate the benefits of positive science 85 (1998) 1209–1222.social interaction and emotions, Psychoneuroendocrinology 23 [40] L .J. Young, Z. Wang, T.R. Insel, Neuroendocrine bases of mono-(1998) 819–835. gamy, Trends Neurosci. 21 (1998) 71–75.

[36] H .J.J. van Oers, E.R. de Kloet, T. Whelan, S. Levine, Maternal [41] M .X. Zarrow, V.H. Denenberg, C.O. Anderson, Frequency ofdeprivation effect on the infant’s neural stress markers is reversed suckling in the pup, Science 150 (1965) 1835–1836.by tactile stimulation and feeding but not by suppressing corticos-terone, J. Neurosci. 18 (1998) 10171–10179.