Brain Research 925 (2002) 67–75www.elsevier.com/ locate /bres

Research report

Dynamical study of tyrosine hydroxylase expression and its correlationwith vasopressin turnover in the magnocellular neurons of the

supraoptico-posthypophysial system under long-term salt loading ofadult rats

a,b b b c´Marina Abramova , Florence Marsais , Andre Calas , Jean Thibault ,a,d ,*Michael Ugrumov

aLaboratory of Hormonal Regulations, Institute of Developmental Biology, Russian Academy of Sciences, 26 Vavilov str., Moscow 117334, Russiab ´ ´Departement de Neurobiologie des Signaux Intercellulaires, Institut des Neurosciences, CNRS UMR 7624, Universite P. et M. Curie, 7 quai St.

Bernard, F-75252 Paris Cedex 05, Francec ´ ´ ´Laboratoire de Neurobiologie Moleculaire, Faculte des Sciences et Technologie, Universite Paris 12, Val de Marne,

´ ´ ´61 Avenue de General de Gaulle, 94010 Creteil, FrancedLaboratory of Neurohistology, Institute of Normal Physiology, Russian Academy of Medical Sciences, 8 Baltiiskaya str, Moscow, Russia

Accepted 30 October 2001

Abstract

Using immunocytochemistry, in situ hybridization and image analysis, we attempted to compare the dynamical expression of tyrosinehydroxylase (TH) and vasopressin (VP) mRNAs and proteins in the magnocellular neurons of the supraoptic nucleus in rats drinking 2%NaCl for 1, 2 and 3 weeks. Three stages in the reaction of VPergic neurons have been distinguished. The initial stage (first week) showeda synchronous activation of TH and VP mRNAs and protein expression as well as an increased number of TH-immunoreactive neurons.The next stage (second week) was characterized by a further increase in the number of TH-immunoreactive neurons. The number ofVPergic neurons also increased significantly. Although the TH and VP mRNAs levels fell during the second week of osmotic stimulation,the TH content increased significantly, and the VP content remained at the same level. During the last stage (third week),TH-immunoreactive neurons increased in number and were as numerous as VP-immunoreactive neurons in intact rats. These data suggestthat, finally, all the VPergic neurons begin to synthesize TH. The concentrations of VP and TH mRNAs did not change during the thirdweek of osmotic stimulation, while the VP and TH contents increased. Thus, our study shows that there is a correlation between THexpression and VP expression and suggests similar mechanisms for the regulation of VP and TH gene expression and synthesis duringlong-term osmotic stimulation. 2002 Elsevier Science B.V. All rights reserved.

Theme: Endocrine and autonomic regulation

Topic: Osmotic and thermal regulation

Keywords: Hypothalamo-posthypophysial system; Immunocytochemistry; In situ hybridization; Osmotic stimulation

1. Introduction and in situ hybridization, it has repeatedly been demon-strated that a number of non-catecholaminergic, mainly

Since the initial application of double immunolabeling peptidergic, neurons co-express tyrosine hydroxylase (TH),on sections or the combination of immunocytochemistry the first rate-limiting enzyme of catecholamine synthesis

[4,13,22,26,30]. Some neurons co-express TH under nor-mal conditions [20], while others only after stimulation

*Corresponding author. Laboratory of Hormonal Regulations, Institute [16,20,33], perturbation of the afferents [16] or in geneticof Developmental Biology, Russian Academy of Sciences, 26 Vavilov

pathology [11,16].str., Moscow 117334, Russia. Tel.: 17-095-135-8842; fax: 17-095-135-VP neurons are generally considered as a promising cell8012.

E-mail address: michael-ugrumov@mail.ru (M. Ugrumov). model to study this phenomenon [5]. They are involved in

0006-8993/02/$ – see front matter 2002 Elsevier Science B.V. All rights reserved.PI I : S0006-8993( 01 )03260-7

68 M. Abramova et al. / Brain Research 925 (2002) 67 –75

the regulation of water–mineral metabolism [3] and are technique using successive incubations with: (a) 1% goatactivated in salt-loaded animals. Only the magnocellular normal serum for 30 min at room temperature, (b) rabbitVP neurons in the supraoptic nucleus (SON) and in the polyclonal antibodies to TH (1:2000) or VP (1:4000)paraventricular nucleus are capable of expressing TH in (Chemicon, Temecula, USA) overnight at 48C, (c) goatresponse to osmotic stimulation [19,20]. TH expression anti-rabbit-biotinylated antibodies (1:200) (Vector, Burl-under osmotic stimulation is reversible and is switched off ingame, USA) for 2 h at room temperature, and (d)after rehydration of the animals [32]. We focused our study avidin–biotin complex (1:1:100) (Vector) for 40 min aton the SON since its neuronal population is homogenous room temperature. The specificity of the antiserum to THand projects the axons to the pituitary posterior lobe where was checked earlier [2]. The controls in this study con-VP is released into the general circulation [8]. sisted of either the omission of the primary antibodies or

A number of studies have been devoted to the evaluation their preabsorption with antigens. Normal serum, primaryof the functional significance of TH in VPergic neurons and secondary antibodies were diluted in PBS with 0.1%using different experimental models of TH expression in Triton X-100 (Sigma, St. Louis, MO, USA). The sectionsthese neurons [11,16,19,20,32,33]. The present study were rinsed with PBS after each incubation except the first.attempted to evaluate the dynamics of TH expression and All incubations were performed in a humid chamber. Theits possible correlation with VP expression in the neurons sections were rinsed with 0.05 M Tris–HCl buffer (pHof the SON during long-term osmotic stimulation. Quan- 7.6), and peroxidase of the avidin–biotin complex wastitative and semi-quantitative immunocytochemistry for VP revealed in the same buffer containing 0.05% 3,39-and TH as well as in situ hybridization for VP and TH diaminobenzidine tetrahydrochloride (Sigma) and 0.01%mRNAs were used to this aim. H O . The developing time was constant for each antigen.2 2

Finally, the sections were dehydrated, mounted in Per-mount, coverslipped, and examined using a light micro-

2. Materials and methods scope.Carto , software developed by IMSTAR (France), was

2.1. Animals and processing of the materials used to carry out a quantitative and semi-quantitativeexamination of VP- and TH-immunoreactive cell bodies in

Adult male Wistar rats, weighting 200–300 g, were used the SON in intact and salt-loaded rats. Moreover, thein this study. They were maintained at 21–238C in a semi-quantitative analysis was used to evaluate the opticallight–dark cycle (light period between 08.00 and 20.00). density of the VP-immunoreactive material in randomlyControl (intact) rats had free access to food and tap water selected areas of the posterior lobe. According to this(15 animals), whereas the experimental rats were given 2% approach, the concentrations of VP and TH could besaline instead of tap water for 1 (15 animals), 2 (six accurately assessed, and the results obtained at differentanimals), and 3 weeks (six animals). The rats were periods of osmotic stimulation could be easily comparedanaesthetized with Pentobarbital (50 mg/kg body weight), between experimental animals and with those in controlperfused via the heart first with saline for 10 min at 378C animals. Indeed, all preparative procedures of the sectionsand then with 4% paraformaldehyde in 0.1 M sodium were performed under standardized conditions that allowedphosphate buffer (pH 7.3) for 15 min at 48C. Thereafter, us to apply meaningful quantitative analysis [27]. Eachthe hypothalami and pituitaries were postfixed by immer- fifth section in the whole series of the SON (10 sectionssion in the same fixative for 2 h at 48C, rinsed in 0.02 M per animal) and the posterior lobe was reviewed under aphosphate buffer with 0.9% NaCl (PBS, pH 7.2–7.4) for 1 light microscope with a 253 objective, followed byh at 48C, immersed in 15% sucrose PBS at 48C overnight, transfer of the image via a video camera (CCD) to theand frozen in isopentane at 2458C. computer monitor (IBM PC Pentium 60) and recording.

The cells were generally considered as immunoreactive if2.2. Quantitative and semi-quantitative their staining intensity exceeded that of the background,immunocytochemistry and this difference disappeared due to the omission of the

primary antibodies or the preliminary preabsorption ofMaterials from three control and three experimental rats primary antibodies with pure antigen.

at each period of salt loading were used for the immuno- Relative concentrations of VP- and TH-immunoreactivecytochemical study. Serial 14-mm-thick coronal sections of materials were measured as the ‘grey level’ (GL), relatedthe SON and pituitaries were cut with a cryostat mi- to the optical density (OD) of the specimen [27]:crotome at 2208C, and successive sections were mounted

OD 2 OD 5 log(GL )Specimen Background Backgroundon different slides. Each slide contained eight sections2 log(GL )(two sections per animal) of the SON or of the pituitary of Specimen

intact animals as well as of the animals salt-loaded for 1, 2,and 3 weeks. The sections were further processed for VP or The GL of the background was measured in the vicinityTH immunocytochemistry according to the avidin–biotin of the SON or posterior lobe corresponding to nervous

M. Abramova et al. / Brain Research 925 (2002) 67 –75 69

tissue with no specific immunostaining. The following the average optical density for five sections. Final ex-average parameters of the SON and posterior lobe were perimental results were expressed as a percentage of theestimated in each experimental group: (a) optical density of optical density in the controls for VP and in 1-weekthe VP- or TH-immunoreactive materials in cell bodies in salt-loaded rats for TH considered as 100%.the SON and of the VP-immunoreactive material in areasof the posterior lobe; (b) the number of VP- and TH- 2.4. Statisticsimmunoreactive cell bodies per section; (c) area of in-dividual VP- or TH-immunoreactive cell bodies; (d) the Statistical analysis was performed using descriptiverelative amounts of VP- or TH-immunoreactive materials statistics and a non-parametric Mann–Whitney test forper cell body (area3optical density). The profiles of independent samples.individual cell bodies or of randomly selected areas of theposterior lobe were outlined with a light pen to obtain amorphometric mask. The optical densities of VP- and

3. ResultsTH-immunoreactive materials were only measured in cellbodies sectioned through the nuclei. The amounts of VP

3.1. Intact ratsand TH per cell were presented as a percentage of controls,considered to be 100%.

According to the immunocytochemical study, numerousVP-immunoreactive neurons (Fig. 1A) were distributed2.3. Quantitative and semi-quantitative in situthroughout the SON showing a dorsoventral gradient inhybridizationtheir concentration that was particularly clear at the levelof the optic tracts (caudal chiasma). The majority ofMaterials from three control and three experimental ratsVP-immunoreactive cell bodies were characterized byat each period of salt loading were used for the in situstrong homogenous immunostaining.hybridization study. Each fourth and fifth coronal cryostat

In addition to VP-immunoreactive neurons, occasionalsections of the SON of intact and salt-loaded rats wereTH-immunoreactive neurons of the same size were scat-mounted on slides arranged in two sets, which weretered in the SON (Fig. 1B). They showed rather weak andprocessed for in situ hybridization of VP mRNA and THhomogenous immunostaining.mRNA, respectively. One set consisted of five slides, each

The posterior lobe was occupied with an extensivecovered with eight sections (two sections per animal).network of VP-immunoreactive fibers. Fiber varicositiesAfter storage at 2808C, the cryostat sections were warmedcontained a large amount of VP-immunoreactive materialto room temperature and rinsed in sterile PBS.(Fig. 2A). The distribution of VP-immunoreactive fibers inTwo oligonucleotide probes were used for in situthe posterior lobe was not homogenous. The concentrationhybridization: (a) VP-41, a 41-mer oligonucleotide probeappeared to be higher along the border with the inter-complementary to the mRNA sequence coding for 115–mediate lobe and in the proximal region, next to the128 amino acid residues of preproVP [28]; and (b) TH-64,pituitary stalk, compared to the rest of the posterior lobe.a 25-mer probe complementary to exon 2 of the TH

According to the in situ hybridization study, radioactivemRNA sequence [18]. The VP-41 sequence has no coun-labeling of the SON was high for the VP probe andterpart with oxytocin mRNA. The specificity of the probesundetectable for the TH probe (Fig. 3).VP-41 and TH-64 has been demonstrated [18,28]. The

probes were radiolabeled, and in situ hybridization wasprocessed as described earlier [19]. 3.2. Rats salt-loaded for 1 week

The sections were exposed to beta-Max X-ray film(Amersham, Buckinghamshire, UK) at 48C. The optimal According to the immunocytochemical study, no changeexposure time was chosen using control slides. Computer in the number of VP-immunoreactive neurons was ob-image analysis of the radioactive signal (see below) served for the first week of salt loading (Fig. 4). The areafollowing the different exposure periods allowed us to of the VP cell bodies increased more than 50% (Fig. 5),avoid overexposure of the whole materials. while the optical density diminished considerably (Fig. 6).

The X-ray films were examined using a microcomputer The relative amount of VP-immunoreactive material perinterfaced to an image analysing system (IMSTAR, neuron remained at the same level as in intact animalsFrance). Autoradiographic images on the X-ray films were (Fig. 7).converted into video images with a video camera (Lhesia The same treatment of the animals resulted in a signifi-LH4015). To adjust the possible defects in the illumination cant increase of: the number of TH-immunoreactive neu-of the optical pathways, all images were corrected using rons (Fig. 4), the average area of cells bodies (Fig. 5), thethe same background image. optical density (Fig. 6) and the relative amount of TH-

The exposure time was overnight for the VP probe and immunoreactive material per neuron (Fig. 7).13 days for the TH probe. The results were estimated as In the posterior lobe, the optical density of the VP-

70 M. Abramova et al. / Brain Research 925 (2002) 67 –75

Fig. 1. Vasopressin-immunoreactive (A, D) and tyrosine hydroxylase-immunoreactive (B, C, E) neuronal cell bodies and fibers on adjacent cryostatsections of the supraoptic nucleus. (A, B, C) Control (intact) rats. (D, E) Rats salt-loaded for 2 weeks. OT, optic tracts. Scale bar 50 mm for (A, B, D, E);12 mm for (C).

immunoreactive material decreased during the first week of 3.3. Rats salt-loaded for 2 weekssalt loading (Fig. 8).

According to the in situ hybridization study, radio- According to the immunocytochemical study, the num-labeling (optical density) of the SON with VP and TH ber and size (Figs. 1C, 4 and 5) of VP-immunoreactive cellradioactive probes increased considerably (Figs. 3 and 9). bodies increased significantly during the second week of

Fig. 2. Vasopressin-immunoreactive fibers, ordinary swellings and Herring bodies on cryostat sections of the posterior lobe. (A) Control (intact) rats. (B)Rats salt-loaded for 2 weeks. PL, posterior lobe; IL, intermediate lobe. Scale bar 50 mm.

M. Abramova et al. / Brain Research 925 (2002) 67 –75 71

Fig. 3. Vasopressin (A–D) and tyrosine hydroxylase (E–H) mRNAs in the supraoptic nucleus of rats. Supraoptic nucleus radioactively labeled by35vasopressin (A–D) and tyrosine hydroxylase (E–H) [ S] probes in control rats (A, E) and rats salt-loaded for 1 (B, F), 2 (C, G) and 3 (D, H) weeks. Scale

bar 2000 mm.

72 M. Abramova et al. / Brain Research 925 (2002) 67 –75

Fig. 4. Number of immunoreactive neurons in the supraoptic nucleus perFig. 7. Relative amounts of vasopressin (VP) and tyrosine hydroxylasesection. Vasopressin-immunoreactive (VP-IR) and tyrosine hydroxylase-(TH) per neuron in the supraoptic nucleus and VP in the posterior lobeimmunoreactive (TH-IR) neurons in control (intact) rats (c) and in rats(PL). Control (c) rats and rats salt-loaded for 1, 2 and 3 weeks (w). Datasalt-loaded (2% NaCl) for 1, 2 and 3 weeks (w). *P,0.01 compared toare presented as a percentage of the control level, considered as 100%.control. **P,0.01, comparison between selected groups. NS, non-signifi-*P,0.01 compared with the previous group in line: control, 1, 2 and 3cant.weeks of salt loading.

Fig. 5. Area of immunoreactive neurons in the supraoptic nucleus.Vasopressin-immunoreactive (VP-IR) and tyrosine hydroxylase-immuno-reactive (TH-IR) cell bodies in control (intact) rats (c) and in rats

Fig. 8. Optical density of vasopressin-immunoreactive (VP-IR) materialsalt-loaded (2% NaCl) for 1, 2 and 3 weeks (w). *P,0.01 compared to

in the posterior lobe. Control (intact) rats (c) and rats salt-loaded (2%control. **P,0.01, comparison between selected groups.

NaCl) for 1, 2 and 3 weeks (w). *P,0.01 compared to control. **P,

0.01, comparison between selected groups. NS, non-significant.

Fig. 9. Relative concentrations of vasopressin and tyrosine hydroxylasemRNAs in the supraoptic nucleus. Optical density of the supraopticnucleus radiolabeled by vasopressin (VP) and tyrosine hydroxylase (TH)

35Fig. 6. Optical density of immunoreactive neurons in the supraoptic [ S] probes in control rats (c) and rats salt-loaded for 1, 2 and 3 weeksnucleus. Vasopressin-immunoreactive (VP-IR) and tyrosine hydroxylase- (w). The data represent: for VP, the percentage of the control level, beingimmunoreactive (TH-IR) cell bodies in control (intact) rats (c) and in rats 100%; for TH, the percentage of the 1-week salt-loading level, beingsalt-loaded (2% NaCl) for 1, 2 and 3 weeks (w). *P,0.01 compared to 100%. NS, non-significant. *P,0.01 compared to control. **P,0.01,control. **P,0.01, comparison between selected groups. comparison between selected groups.

M. Abramova et al. / Brain Research 925 (2002) 67 –75 73

salt loading. No difference in the optical density of these dynamics of TH and VP gene expression and syntheses incells and the relative amount of VP-immunoreactive the magnocellular neurons of the SON during long-termmaterial per neuron was observed compared to salt loading osmotic stimulation using immunocytochemistry, in situfor 1 week (Figs. 6 and 7). hybridization, and image analysis.

The number of TH-immunoreactive neurons increased When evaluating the rate of preproVP and TH, it shoulddramatically (by 3.5-fold) over the second week of salt be taken into account that these proteins differ considera-loading (Figs. 1D and 4), although they remained less bly in some specific properties. Thus, TH is a non-releas-numerous compared to VP-immunoreactive neurons. An able cytosolic protein that is degraded by cytosolic peptid-increase of the area, optical density and the relative amount ases [10]. Conversely, preproVP is a releasable proteinof TH-immunoreactive material per neuron was also a stored in secretory granules and protected from enzymaticcharacteristic of this experimental group (Figs. 5–7). As degradation by the granule membrane [8]. From the abovestated above, TH-immunoreactive neurons did not differ in comments it follows that the average concentration (opticalsize from VP-immunoreactive neurons. TH-immuno- density) or content (optical density3area of the TH-im-reactive and VP-immunoreactive neurons with clear cyto- munoreactive cell body) of TH-immunoreactive materialplasmic vacuoles first appeared in the second week of salt per neuron is proportional to the difference betweenloading. synthesized TH and enzymatically degraded TH. In turn,

The optical density of the VP-immunoreactive material the average concentration (optical density) or contentin the posterior lobe decreased during the second week of (optical density3area of the VP-immunoreactive cellsalt loading (Figs. 2B and 8). body) of VP-immunoreactive material per neuron is pro-

According to the in situ hybridization study, radio- portional to the difference between synthesized VP andlabeling (optical density) of the SON with VP and TH released VP. The relative content of either protein perradioactive probes decreased significantly during the sec- neuron appears to be a more informative index of theond week of salt loading (Figs. 3 and 9). synthetic activity of peptidergic neurons than the con-

centrations of these materials if neurons change in size in3.4. Rats salt-loaded for 3 weeks the course of the experiment. For instance, the concen-

tration of VP-immunoreactive material (optical density)The immunocytochemical study showed that the number

decreased gradually over 3 weeks of osmotic stimulation,of VP-immunoreactive cell bodies did not change (Fig. 4)

while the relative content of VP-immunoreactive materialwhile their area increased significantly in the third week of

was maintained at the control level when taking intosalt loading (Fig. 5). Although the optical density of these

account the progressive hypertrophy of VPergic neurons incells decreased slightly, the relative amount of VP-im-

this experiment (see Results). The dynamics of the VPmunoreactive material per neuron exceeded that in rats

content per neuron was compared with that of the totalsalt-loaded for 2 weeks (Fig. 7).

amount of VP mRNA in all neurons in the SON perIn contrast to VP-immunoreactive neurons, the number

section. This approach appeared to be suitable, as the greatof TH-immunoreactive neurons increased significantly

majority of VPergic neurons were recently demonstrated toduring the third week of salt loading (Fig. 4). Neverthe-

contain both VP mRNA and VP peptide — more than 90%less, they were less numerous than VP-immunoreactive

in intact rats and about 80% in rats salt-loaded for 1 weekneurons in the same experimental group. The area of

[1].TH-immunoreactive neurons increased to the same extent

The concentration (optical density) of VP-immuno-as that of VP-immunoreactive neurons (Fig. 3). The optical

reactive material in the posterior lobe is proportional to thedensity and the relative amount of TH-immunoreactive

difference between the rate of VP transfer from the cellmaterial per neuron increased during the third week of salt

bodies to the axon terminals and the rate of VP releaseloading (Figs. 6 and 7).

from the axons to the general circulation.The optical density of the VP-immunoreactive material

in the posterior lobe did not change during the third week4.2. The dynamics of TH and VP expression in

of salt loading (Fig. 8).magnocellular neurons during long-term osmotic

From the in situ hybridization study with both VP andstimulation

TH radioactive probes, it follows that there was no changein the optical density of the SON in comparison with rats

This study first demonstrated a correlation between thesalt-loaded for 2 weeks (Figs. 3 and 9).

relative contents of VP mRNA and TH mRNA in mag-nocellular neurons in the course of their long-term osmoticstimulation. The increase of the contents of TH and VP4. DiscussionmRNAs during the first week of osmotic stimulation (in

4.1. Technical remarks agreement with earlier studies [15,33]) was followed by adecrease in the second week, and this level was maintained

This is the first study attempting to compare the until the end of the experiment. The synchronization of VP

74 M. Abramova et al. / Brain Research 925 (2002) 67 –75

and TH mRNA syntheses suggests similar mechanisms for Minor modifications (steady state) of TH and VPexpression and VP depletion were observed during the lastthe regulation of VP and TH gene expression (transcriptionstage of VP neuron reaction to osmotic stimulation,factors, etc.).corresponding to the third week of salt loading of rats.The dynamics of the TH mRNA content observed in thisTH-immunoreactive neurons increased in number by thestudy are similar to those described earlier under long-termend of the experiment and were as numerous as VP-immobilization stress [23]. Moreover, other stress stimuliimmunoreactive neurons in intact and 1-week salt-loaded(hypoxia, cold), or the administration of drugs modulatinganimals. Nevertheless, these neurons remained less numer-the noradrenergic afferent (reserpine, cocaine, nicotine,ous than VP-immunoreactive neurons in rats salt-loaded forforskoline), which influence the VPergic neurons, could2 and 3 weeks. These data suggest that, by the end of theinduce TH mRNA production. This might be a result of thethird week of osmotic stimulation, all VPergic neuronsincrease of either the transcription rate or of the stabilityoriginally detected in the SON of intact rats co-express(half-life) of mRNA. For example, hypoxia provokes anTH. In this context, VP-immunoreactive neurons that didincrease of both the rate of TH transcription and thenot synthesize TH during the third week of osmoticstability of TH mRNA [9,24]. Second messengers such as

21 stimulation were most probably represented by oxytociner-Ca and cAMP are assumed to be involved in thisgic neurons, which begin to produce VP under stimulationregulation [12,24]. On the other hand, osmotic stimulation

21 [6,21], but fail to co-express TH [19,20].results in elevated levels of intracellular cAMP, Ca , andAlthough previous attempts failed to elucidate thegenes of c-fos and the jun families in the SON of rats

functional significance of TH in VPergic neurons, they[7,17,31], which, in turn, are capable of modifying THprovide important information about the peculiarities ofmRNA transcription and stability levels [24,25].TH in these neurons. TH failed to convert L-tyrosine toIn this study, three subsequent stages were distinguishedL-DOPA in VPergic neurons, because of the lack of thein the reaction of VPergic neurons to osmotic stimulation.specific co-factor, BH [19]. In light of the present data,4The initial stage, corresponding to the first week of saltthe hypothesis concerning the similar regulation of TH and

loading, is characterized by an activation of VP neurons.VP transcription and translation appears to be promising

This was manifested in: (a) an increased number ofand should be tested in a future study.

VPergic neurons expressing TH; (b) activated synthesis of Thus, our study demonstrates a correlation between THVP mRNA and TH mRNA; (c) increased synthesis of TH and VP expression during long-term salt loading, sug-and VP; (d) intense depletion of VP from the axons in the gesting similar mechanisms for their regulation. Moreover,posterior lobe; and (e) functional hypertrophy of VPergic a possible functional interaction between both molecules,neurons. The VP content of cell bodies did not change for example the involvement of TH in VP release, cannotduring the first week of osmotic stimulation in spite of the be excluded.dramatic depletion of VP from the axons in the posteriorlobe. These data are in agreement with the radioimmunoas-say observation of a two-fold increased VP level in the Acknowledgementsplasma of rats salt-loaded for 1 week [14]. VP release fromosmotically activated VPergic neurons was also demon- The present study was supported by the followingstrated by the increased number of neurons containing VP grants: PICS (98-04-22018), NATO (OUTR CRGmRNA, but lacking VP peptide, in 10% of rats after salt 970131), INTAS-RFBR (95-IN-RU-1246), RFBR (96-04-loading for 1 week [1]. 49441; 00-15-97840), the French Ministry of Education

The next stage, the second week of salt loading, is and Science, the French Ministry of Foreign Affairs.considered to be a period of adaptation of the VPergicneurons to chronic osmotic stimulation. It is characterizedby a dramatic increase in the number of VP-immuno- Referencesreactive neurons and VPergic neurons probably co-ex-pressing TH, as well as by an increase of the VP level in [1] F. Amaya, M. Tanaka, Y. Tamada, Y. Tanaka, G. Nilaver, Y. Ibata,plasma [29]. The influence of salt loading on vasopressin gene expression in

The most convincing evidence of neuronal adaptation to magno- and parvocellular hypothalamic neurons: an immuno-cytochemical and in situ hybridization analysis, Neuroscience 89osmotic stimulation during the second stage appears to be(1999) 515–523.the synchronous decrease of the content of both TH

[2] M. Arluison, M. Dietl, J. Thibault, Ultrastructural morphology ofmRNA and VP mRNA in cell bodies. The TH content of dopaminergic nerve terminals and synapses in the striatum of the ratcell bodies increased progressively during the second week using tyrosine hydroxylase immunocytochemistry: a topographicalof osmotic stimulation, while the VP content remained at study, Brain Res. Bull. 13 (1984) 169–185.

[3] M.J. Brownstein, J.T. Russel, H. Gainer, Synthesis, transport, andthe same level as in intact animals. This means that, evenrelease of posterior pituitary hormones, Science 207 (1980) 373–at a relatively low level of TH mRNA, the rate of TH378.

protein greatly prevailed over that of enzymatic degra- [4] M.J. Brownstein, E. Mezey, Multiple chemical messengers indation of cytosolic TH, whereas the rate of VP remained hypothalamic magnocellular neurons, Prog. Brain Res. 68 (1986)unchanged. 161–168.

M. Abramova et al. / Brain Research 925 (2002) 67 –75 75

´ ¨[5] A. Calas, La versatilite neuronale, La Vie des Sciences Comptes [20] B. Meister, R. Cortes, M.J. Villar, M. Schaulling, T. Hokfelt,Rendus 11 (1994) 271–285. Peptides and transmitter enzymes in hypothalamic magnocellular

`[6] A. Calas, M. Landry, D. Roche, A. Trembleau, Un modele de neurons after administration of hyperosmotic stimuli: comparison´ ´plasticite phenotypique: les neurones hypothalamo-post- between messenger RNA and peptide-protein levels, Cell Tissue

hypophysaires, C.R. Soc. Biol. 188 (1994) 187–206. Res. 260 (1990) 279–298.[7] D.A. Carter, D. Murphy, Cyclic nucleotide dynamics in the rat [21] E. Mezey, J.Z. Kiss, Coexpression of vasopressin and oxytocin in

hypothalamus during osmotic stimulation: in vivo and in vitro hypothalamic supraoptic neurons of lactating rats, Endocrinologystudies, Brain Res. 487 (1989) 350–356. 129 (1991) 1814–1820.

[8] M. Castel, H. Gainer, H.D. Dellmann, Neuronal secretory systems, [22] I. Nagatsu, M. Sakai, T. Takeuchi, R. Arai, N. Karasawa, K.Int. Rev. Cytol. 88 (1984) 303–458. Yamada, T. Nagatsu, Tyrosine hydroxylase (TH)-only-immuno-

[9] M.F. Czyzyk-Krzeska, W.R. Paulding, J.E. Beresh, S.L. Kroll, Post- reactive non-catecholaminergic neurons in the brain of wild mice ortranscriptional regulation of tyrosine hydroxylase gene expression the human TH transgenic mice do not contain GTP cyclohydrolase I,by oxygen in PC12 cells, Kidney Int. 51 (1997) 585–590. Neurosci. Lett. 228 (1997) 55–57.

[10] E. Fernandez, G.L. Craviso, Protein synthesis blockade differentially [23] M. Rusnak, S. Zorad, P. Buckendahl, E.L. Sabban, R. Kvetnansky,affects the degradation of constitutive and nicotinic receptor-induced Tyrosine hydroxylase mRNA levels in locus coeruleus of rats duringtyrosine hydroxylase protein level in isolated bovine chromaffin adaptation to long-term immobilization stress exposure, Mol. Chem.cells, J. Neurochem. 73 (1999) 169–178. Neuropathol. 33 (1998) 249–258.

¨[11] S. Fetissov, F. Marsais, S. Nicolaıdis, A. Calas, Expression of [24] R. Raymond, D. Millhorn, Regulation of tyrosine hydroxylase gene21tyrosine hydroxylase in magnocellular hypothalamic neurons of expression during hypoxia: role of Ca and PKC, Kidney Int. 51

obese (fa / fa) and lean heterozygous (Fa / fa) Zucker rats, Mol. Brain (1997) 536–541.Res. 50 (1997) 314–318. [25] E.L. Sabban, Control of tyrosine hydroxylase gene expression in

[12] L.H. Fossom, C.R. Sterling, A.W. Tank, Regulation of tyrosine chromaffin and PC12 cells, Cell Dev. Biol. 8 (1997) 101–111.hydroxylase gene transcription rate and tyrosine hydroxylase mRNA [26] F.J. Seil, M.L. Johnson, R. Nishi, G. Nilaver, Tyrosine hydroxylasestability by cyclic AMP and glucocorticoid, Mol. Pharmacol. 42 expression in non-catecholaminergic cells in cerebellar cultures,(1992) 898–908. Brain Res. 569 (1992) 164–168.

¨[13] T. Hokfelt, B. Meister, T. Melander, B. Everitt, Coexistence of [27] A.J. Smolen, Image analytic techniques for quantification of im-classical transmitters and peptides with special reference to the munohistochemical staining in the nervous system, in: P.M. Connarcuate nucleus–median eminence complex, in: D. Nerozzi (Ed.), (Ed.), Quantitative and Qualitative Microscopy, Methods in Neuro-Hypothalamic Dysfunction in Neuropsychiatric Disorders, Raven sciences, Vol. 3, Academic Press, San Diego, 1990, pp. 208–229.Press, New York, 1987, pp. 21–34. `[28] A. Trembleau, A. Calas, M. Fevre-Montange, Combination of

[14] S.C. Hooi, G.S. Richardson, J.K. McDonald, J.M. Allen, J.B. immunocytochemistry and in situ hybridization with syntheticMartin, J.I. Koenig, Neuropeptide Y (NPY) and vasopressin (AVP) oligonucleotide probes to localize simultaneously vasopressin, oxy-in the hypothalamo–neurohypophysial axis of salt-loaded or Brattle- tocin and their mRNAs in hypothalamic magnocellular neurons,boro rats, Brain Res. 486 (1989) 214–220. Bull. Assoc. Anat. 72 (1988) 101–106.

[15] S. Hyodo, M. Fujiwara, M. Sato, A. Urano, Molecular- and [29] H.H.M. Van Tol, D.Th.A.M. Voorhuis, J.P.H. Burbach, Oxytocinimmuno-histochemical study on expressions of vasopressin and gene expression in discrete hypothalamic magnocellular cell groupsoxytocin genes following salt loading, Zool. Sci. 5 (1988) 1033– is stimulated by prolonged salt loading, Endocrinology 120 (1987)1042. 71–76.

[16] J.Z. Kiss, E. Mezey, Tyrosine hydroxylase in magnocellular neuro- [30] C. Verney, P. Gaspar, A. Febvret, B. Berger, Transient tyrosinesecretory neurons: response to physiological manipulation, Neuroen- hydroxylase-like immunoreactive neurons contain somatostatin anddocrinology 43 (1986) 519–525. substance P in the developing amygdala and bed nucleus of the stria

[17] M. Lafarga, M.T. Berciano, F.J. Martinez-Guijarro, M.A. Andres, B. terminalis of the rat, Brain Res. 470 (1988) 45–58.Mellstrom, C. Lopez-Garcia, J.R. Naranjo, Fos-like expression and [31] K. Wang, S.E.F. Guldenaar, J.T. McCabe, Fos and Jun expression innuclear size in osmotically stimulated supraoptic nucleus neurons, rat supraoptic nucleus neurons after acute vs. repeated osmoticNeuroscience 50 (1992) 867–875. stimulation, Brain Res. 746 (1997) 117–125.

` ´[18] P. Laniece, H. Le Hir, S. Bodeau-Pean, Y. Charon, L. Valentin, C. [32] K. Yagita, H. Okamura, Y. Ibata, Rehydration process from salt-Thermes, J. Mallet, S. Dumas, A novel rat tyrosine hydroxylase loading: recovery of vasopressin and its coexisting galanin, dynor-mRNA species generated by alternative splicing, J. Neurochem. 66 phin and tyrosine hydroxylase immunoreactivities in the supraoptic(1996) 1819–1825. and paraventricular nuclei, Brain Res. 667 (1994) 13–23.

[19] F. Marsais, A. Calas, Ectopic expression of non-catecholaminergic [33] W.S. Young, M. Warden, E. Mezey, Tyrosine hydroxylase mRNA istyrosine hydroxylase in rat hypothalamic magnocellular neurons, increased by hyperosmotic stimuli in the paraventricular and supra-Neuroscience 94 (1999) 151–161. optic nuclei, Neuroendocrinology 46 (1987) 439–444.