help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology, doi:10.1210/en.2008-0387
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Alonso, G.
Right arrow Articles by Guillon, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Alonso, G.
Right arrow Articles by Guillon, G.
Endocrinology Vol. 149, No. 9 4279-4288
Copyright © 2008 by The Endocrine Society

Intrahypothalamic Angiogenesis Induced by Osmotic Stimuli Correlates with Local Hypoxia: A Potential Role of Confined Vasoconstriction Induced by Dendritic Secretion of Vasopressin

Gérard Alonso, Evelyne Gallibert, Chrystel Lafont and Gilles Guillon

Institut de Génomique Fonctionnelle, Département d’Endocrinologie; Centre National de la Recherche Scientifique Unité Mixte de Recherche 5203; Institut National de la Santé et de la Recherche Médicale Unité 661; and Université Montpellier I, Université Montpellier II, F-34094 Montpellier, France

Address all correspondence and requests for reprints to: Gilles Guillon, Institut de Génomique Fonctionnelle, 141 rue de la Cardonille, F-34094 Montpellier Cedex 05, France. E-mail: gilles.guillon{at}igf.cnrs.fr.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
We have previously shown that hyperosmotic stimulation of adult Wistar rats induces local angiogenesis within hypothalamic magnocellular nuclei, in relation to the secretion of vascular endothelial growth factor (VEGF) by the magnocellular neurons. The present study aimed at understanding how osmotic stimulus relates to increased VEGF secretion. We first demonstrate a correlation between increased VEGF secretion and local hypoxia. Osmotic stimulation is known to stimulate the metabolic activity of hypothalamic magnocellular neurons producing arginine vasopressin (AVP) and to increase the secretion of AVP, both by axon terminals into the circulation and by dendrites into the extracellular space. In AVP-deficient Brattleboro rats, the dramatic activation of magnocellular hypothalamic neurons failed to induce hypoxia, VEGF expression, or angiogenesis, suggesting a major role of hypothalamic AVP. A possible involvement of dendritic AVP release is supported by the findings that 1) hypoxia and angiogenesis were not observed in non osmotically stimulated Wistar rats in which circulating AVP was increased by the prolonged infusion of exogenous AVP, 2) contractile arterioles afferent to the magnocellular nuclei were strongly constricted by the perivascular application of AVP via V1a receptors (V1a-R) stimulation, and 3) after the intracerebral or ip administrations of selective V1a-R antagonists to osmotically stimulated rats, hypothalamic hypoxia and angiogenesis were or were not inhibited, respectively. Together, these data strongly suggest that the angiogenesis induced by osmotic stimulation relates to tissue hypoxia resulting from the constriction of local arterioles, via the stimulation of perivascular V1a-R by AVP locally released from dendrites.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
IN MAMMALS, THE central nervous system (CNS) vasculature is formed by a complex network of blood vessels and capillaries that supply oxygen and various nutriments to neurons. Under basal physiological conditions, the adult CNS vasculature is essentially quiescent. As for many other cell types, neural cells are able to detect variations in oxygen levels and promote adaptive cellular responses that lead to angiogenesis. For instance, various pathological conditions that induce local tissue hypoxia related to blood circulation impairment, such as stroke, head trauma, and neurodegenerative disease can provoke local angiogenesis. This has mainly been found to result from the activation of hypoxia-inducible genes, including the potent angiogenic factor vascular endothelial growth factor (VEGF). Results from a number of studies also suggest that angiogenesis can be locally induced within regions of the adult CNS by the chronic stimulation of specific neuronal systems. For instance, rearing rats in a complex environment was found to increase the capillary density within the visual cortex (1), whereas prolonged motor activity was reported to induce angiogenesis within the cerebellar cortex (2) and primary motor cortex (3). Within the hypothalamus, the supraoptic and paraventricular nuclei (SON and PVN) contain a dense population of magnocellular neurons that synthesize arginine vasopressin (AVP) or oxytocin (OT), two peptidergic hormones playing major roles in, respectively, water balance and vasoconstriction or reproduction. Recently, we have shown that local angiogenesis could be induced within these hypothalamic nuclei by a prolonged osmotic stimulus in association with an increased secretion of VEGF by the magnocellular neurons localized to these nuclei (4).

The present study aims at understanding how osmotic stimulus relates to increased VEGF secretion by the hypothalamic magnocellular neurons. A main effect of osmotic stimulus is the activation of magnocellular hypothalamic neurons secreting AVP, thus leading to an increased secretion of this potent vasoconstrictor. Because hypoxia is a potent stimulus for VEGF secretion, we examined the hypothesis that increased AVP secretion may stimulate the angiogenic process via local vasoconstriction-induced hypoxia.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Materials
AVP and desmopressin (dDAVP) were from Bachem (Bubendorf, Switzerland). The selective V1a receptor (V1a-R) antagonists SR49059 and d(CH2)5TyrMe2AVP (Manning compound) were kindly provided by Sanofi-Aventis (Toulouse, France) and by Dr. M. Manning (Toledo, OH), respectively. Alzet osmotic minipumps were from Charles River Laboratories (L’arbresle, France). Bromodeoxyuridine (BrdU), dimethylsulfoxide, and Cremophor were from Sigma (Lisle d’Abeau Chesne, France). Rhodamine-conjugated wheat germ agglutinin (R-WGA) was from Vector Laboratories (Peterborough, UK). The hypoxyprobe-1 kit was from Chemicon-Millipore (Saint Quentin en Yvelines, France).

Animals
Male Wistar and Long Evans rats were provided by CERJ (Le Genest, France) and male homozygous Brattleboro rats by the laboratory of A. Burlet (Nancy, France). The rats were bred in the local animal facility and maintained on a 12-h dark, 12-h light cycle (0700–1900 h) with food and water ad libitum. Animals were divided into three groups including 1) control rats with free access to standard dry food and tap water, 2) hyperosmotically stimulated rats given 2% NaCl solution as drinking fluid for 1–6 d, and 3) rehydrated rats that were hyperosmotically stimulated for 6 d and then given tap water as drinking fluid for 1–5 d. Animal manipulations were performed according to the recommendations of the French Ethical committee (Agreement 75-178, 05/16/2000) and under the supervision of authorized investigators. All protocols performed have been approved by the local institution’s ethics committee.

Administration of exogenous AVP
Rats were sc implanted with an Alzet 7-d osmotic pump (model 1007D, delivery rate 0.5 µl/h for 7 d) filled with 100 µl saline either alone or containing 91 µg AVP (infusing 13 µg peptide per day). Animals were then euthanized at 3 or 6 d after implantation.

Administration of V1a-R antagonists
For peripheral administration, we used SR49059, a selective nonpeptidic V1a-R antagonist that does not cross the blood-brain barrier (5). Due to its low solubility in water, it was dissolved in saline containing 5% dimethylsulfoxide plus 7.5% Cremophor (vol/vol) and administered via two daily ip injections (1 mg/kg) for 3 d in osmotically stimulated rats drinking a 2% saline solution. For central administration, we used d(CH2)5TyrMe2AVP (Manning compound), a selective peptidic V1a-R antagonist that is water soluble (and thus usable within the Alzet osmotic minipumps. After deep anesthesia with equithesin (3 ml/kg), animals were placed in a David Kopf stereotaxic device and a stainless-steel cannula (26 gauge, 10 mm long) placed into the lateral ventricle according to the stereotaxic atlas of Paxinos and Watson (24): 1.3 mm posterior, 1.6 mm lateral, and 4 mm dorsoventral to bregma (to prevent tissue lesions in the vicinity of the SON, the cannula was implanted into the lateral ventricle). The cannula was connected via polyethylene tubing to an Alzet 7-d osmotic pump filled with either 100 µl saline alone or containing 100 µg of the Manning antagonist. The cannula was cemented in place adjacent to an anchor screw inserted in the skull, and the Alzet pump was inserted into a sc pocket made with sterile forceps over the rat neck and shoulder blades. Twenty-four hours after the surgery, animals were given 2% saline as drinking fluid for 5 d. Animals were fixed by intracardiac perfusion after pimonidazole hydrochloride and BrdU administration as described below.

Detection of tissue hypoxia
To detect tissue hypoxia, we used the immunocytochemical detection of pimonidazole adducts 1 h after an ip injection of pimonidazole hydrochloride (60 mg/kg; hypoxyprobe-1 kit; Chemicon-Millipore).

BrdU administration
BrdU (Sigma) was administered ip (150 mg/kg in 0.5 ml 0.01 N hydroxychloride solution) 10 and 5 h before the fixation of the animals.

Immunocytochemistry
After deep anesthesia with equithesin, animals were perfused through the ascending aorta with PBS (pH 7.4) followed by 500 ml fixative composed of 4% paraformaldehyde (confocal microscopy) or 4% paraformaldehyde plus 0.5% glutaraldehyde (electron microscopy). The forebrain was dissected and fixed by immersion overnight in the same fixative. It was then cut frontally with a vibratome into 50-µm-thick sections that were carefully rinsed in PBS and subsequently treated for single or multiple immunocytochemical labeling.

Confocal microscopy.
The vibratome sections were incubated for 48 h at 4 C with one or two primary antibodies including polyclonal rabbit IgG, polyclonal goat IgG, and monoclonal mouse IgG antibodies. The different markers, suppliers, and concentrations used in this study are listed in Table 1Go. After rinsing in PBS, sections were incubated for 4 h at 4 C with corresponding secondary antibodies conjugated with either Alexa-488 (Molecular Probes, Eugene, OR) or Cy3 (Jackson Laboratories, Suffolk, UK). The primary and secondary antibodies were diluted in PBS containing 1% normal goat or donkey serum and 0.1% Triton X-100. Sections treated for the immunodetection of BrdU were incubated in 2 N HCl for 30 min at room temperature, carefully rinsed in PBS, and then incubated with the anti-BrdU antibody. Sections treated for the immunodetection of hypoxia-inducible factor (HIF)-1{alpha} were heated for 20 min in citrate buffer (pH 6), rinsed in cold PBS, and then incubated with the HIF-1{alpha} antibody.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Antibodies used for immunochemistry

 
After rinsing in PBS, labeled sections were mounted in Mowiol and observed under a Bio-Rad MRC 1024 confocal laser scanning microscope equipped with a krypton/argon mixed-gas laser. Two laser lines emitting at 488 and 568 nm were used for exciting the Alexa-488- or Cy3-conjugated secondary markers. The background noise of each confocal image was reduced by averaging four image inputs. The organization of the immunostained structures was studied on bidimensional reconstructed images obtained by collecting 10 consecutive confocal images 1 µm apart through the whole vibratome section thickness and then projecting on the same plane. The specificity of the AVP, OT, and commercial antibodies was assessed by absorption tests. Additional controls consisted of 1) omitting the primary antibodies and applying the secondary antibodies alone, 2) applying each primary antibody sequentially and then reacting these with an inappropriate secondary antibody, and 3) exciting each fluorochrome by an inappropriate illumination. This allowed us to confirm that the two secondary antibodies used in double-immunostaining experiments induced no artifactual fluorescent labeling and that there was no overlap of the emission spectra of the two fluorochromes.

Quantitative analysis was performed on series of sections passing through the middle portions of the SON (i.e. the largest SON areas). Cell proliferation and tissue hypoxia were quantified by counting the BrdU-labeled nuclei and the hypoxyprobe-1-labeled neurons detected on adjacent sections in four SON areas per rat, with at least three rats per experiment. Data were statistically compared using the nonparametric test of Mann and Whitney.

Electron microscopy.
After rinsing in PBS, sections were incubated with the primary antibody for 48 h at 4 C, rinsed in PBS, and incubated for 12 h at 4 C with a peroxidase-labeled Fab fragment of goat antirabbit IgG. The primary and secondary antibodies were diluted in PBS containing 2% normal goat and 0.1% saponin. After rinsing in Tris buffer (pH 7.4), they were then incubated with 0.1% 3,3'-diaminobenzidine in Tris buffer, rinsed in 0.1 M cacodylate buffer (pH 7.3), and postfixed in 1% osmium tetroxide in the same buffer. After dehydration in graded concentrations of ethanol, they were then embedded in Spur (Sigma). Punches of 1.5 mm in diameter were cut through the arcuate nucleus and the median eminence and mounted on araldite blocks. Ultrathin sections were cut, moderately counterstained with uranyl acetate, and examined with a Hitachi H 7100 electron microscope (Hitachi, Krefeld, Germany).

Visualization of arteriolar vasoconstriction
Rats were anesthetized as described above and cerebral vessels labeled by the intraaortic injection of 10 ml PBS followed by 2 ml PBS containing 0.2 mg R-WGA. The brain was then rapidly dissected, glued with its ventral surface up in a petri dish filled with an oxygenated KREBS buffer, and viewed under a fluorescence stereo microscope equipped with a CCD camera. To evaluate the contractile effects of AVP, the original incubating medium was replaced by buffer solutions containing AVP (at 10–9 to 10–7 M), the Manning V1a-R antagonist alone (10–7 M) or both the Manning antagonist (10–7 M) and AVP (10–9 M).

AVP RIA
Plasma AVP was determined by collecting the trunk blood after decapitation of conscious animals. To prevent manipulation-induced stress, animals were manipulated two times per day for the 3 d before decapitation. Blood was centrifuged in prechilled tubes containing EDTA for 15 min at 3000 x g and 4 C. The plasma concentration of AVP was determined by the direct Bühlman RIA kit (Bühlman Laboratories AG, Basel, Switzerland).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
The observations of the present study were restricted to brain sections passing through the hypothalamus that contained the different magnocellular nuclei, including the SON, PVN, and accessory nucleus (AN). As compared with the other hypothalamic magnocellular nuclei, the SON has a clearly defined anatomical localization (along the lateral border of the optic chiasm) and is homogeneous, essentially containing AVP and OT neurons. For these reasons, the data reported here will mostly concern the SON, although the present findings apply to all the hypothalamic magnocellular nuclei. In light of our previous observations that the cell proliferation induced by osmotic stimuli within hypothalamic magnocellular nuclei essentially concerns endothelial cells (4), we used the detection of proliferating (BrdU-labeled) cells as an index of angiogenesis within these nuclei.

VEGF expression induced by hyperosmotic stimulation correlates with tissue hypoxia
Within adult tissue, tissue hypoxia is a major stimulus for angiogenesis. We therefore investigated the possible occurrence of tissue hypoxia within the hypothalamus of osmotically stimulated rats. Tissue hypoxia was detected using the hypoxyprobe-1 kit (Chemicon-Millipore) based on the immunocytochemical detection of pimonidazole adducts formed in vivo after the administration of pimonidazole hydrochloride. In a preliminary control study, transient cerebral ischemia was induced in adult rats by clamping one of the carotid arteries for 20 min. After 2 h, the animals were injected with pimonidazole hydrochloride and fixed 1 h later. In these animals, strong immunostaining for pimonidazole adducts (hypoxyprobe-1 immunostaining) was detected within a large number of cells dispersed throughout the hypothalamic and other brain regions ipsilateral to the clamp (Fig. 1AGo). In contrast, a faint if any immunostaining was detected in the brain of control, normally hydrated animals (Fig. 1BGo). We then injected pimonidazole hydrochloride into rats osmotically stimulated for 3 or 6 d. Whatever the osmotic stimulus duration, we detected an intense hypoxyprobe-1 immunostaining within the hypothalamic magnocellular nuclei (Fig. 1Go, C and D), compared with very faint if any immunostaining in the surrounding hypothalamic and extrahypothalamic regions. Rats that were normally hydrated for 2 d after 6 d of hyperosmotic stimulation showed a very faint hypoxyprobe-1 immunostaining within the hypothalamic magnocellular nuclei (data not shown).


Figure 1
View larger version (193K):
[in this window]
[in a new window]

  		FIG. 1. Hyperosmotic stimulation induces tissue hypoxia within hypothalamic magnocellular nuclei. Pimonidazole hydrochloride was injected into rats before fixation, and sections passing through the SON were immunostained for the pimonidazole hydrochloride adducts (hypoxyprobe-1 immunostaining). A and B, Intense hypoxyprobe-1 immunostaining is associated with a large number of cells dispersed through the optic chiasm (oc) and the surrounding hypothalamic region of a rat submitted to cerebral ischemia (A), whereas very faint if any immunostaining is detected in the SON region of a control, normally hydrated rat (B); C and D, in a 6-d osmotically stimulated rat, strong hypoxyprobe-1 immunostaining associates with a large number of neurons localized to the SON (C) or the PVN (D). The hatched line in B indicates the anatomical limits of the SON. V, Third ventricle. Scale bar, 100 µm (A–D).

 
We then examined whether such a local tissue hypoxia induced by a prolonged osmotic stimulation correlated with an increase in VEGF expression. In osmotically stimulated rats, every examined brain section double immunostained for hypoxyprobe-1 and VEGF revealed an intense neuronal VEGF immunostaining restricted to those hypothalamic magnocellular nuclei also exhibiting intense hypoxyprobe-1 immunostaining (Fig. 2Go, A–D). Lastly, we examined whether osmotic stimulation is associated with the activation of the transcription factor HIF-1{alpha}, the most recognized hypoxia-sensitive regulator of VEGF gene expression. Sections double immunostained for HIF-1{alpha} and VEGF indicated that, as compared with normally hydrated rats (Fig. 2Go, E and F), both immunostainings were highly increased in the hypothalamic magnocellular nuclei of osmotically stimulated rats, the HIF-1{alpha}-immunostaining being essentially localized to neuronal profiles exhibiting intense VEGF immunostaining (Fig. 2Go, G and H). Together, these data indicate that both angiogenesis and VEGF secretion, induced by a prolonged osmotic stimulus, correlate with tissue hypoxia.


Figure 2
View larger version (44K):
[in this window]
[in a new window]

  		FIG. 2. Correlation between tissue hypoxia and VEGF expression. Pimonidazole hydrochloride was injected into the rats before fixation. A–D, Sections double immunostained for hypoxyprobe-1 and VEGF showing that, as compared with control (A and B), both immunostainings are highly increased within SON neurons of osmotically stimulated rats (C and D); E–H, sections double immunostained for HIF-1{alpha} and VEGF showing that, as compared with control normally hydrated rats (E and F), both immunostainings are highly increased in the SON of a 6-d osmotically stimulated rats (G and H); inset in H, color image of superimposed double immunostaining showing that HIF-1{alpha} immunostaining essentially associates with the nucleus of intensely VEGF-immunostained neurons. oc, Optic chiasm. Scale bar, 100 µm (in D for A–D; in H for E–H).

 
Hypothalamic AVP plays a major role in the hypoxia and angiogenesis processes induced by osmotic stimuli
We next tried to identify the link between hyperosmotic stimulus and local tissue hypoxia and angiogenesis. One major consequence of hyperosmotic stimulation is the functional activation of hypothalamic AVP neurons and the subsequent secretion of large amounts of AVP by these neurons. Tissue hypoxia could be associated with both events because 1) the functional activation of magnocellular neurons induces a local increase of oxygen metabolism and 2) secreted AVP is a potent vasoconstrictor that may locally alter blood and oxygen supply. The Brattleboro rat is a mutant form of the Long Evans rat strain that suffers from a genetic inability to produce functional AVP (6). The AVP deficiency in homozygous Brattleboro rats leads to continuous osmotic stimulation of the magnocellular hypothalamic neurons (7) with a total absence of AVP secretion. We first verified that, as in Wistar rats, hyperosmotic stimuli produced hypoxia and angiogenesis within the hypothalamic magnocellular nuclei of Long Evans rats. For this, pimonidazole hydrochloride and BrdU were injected as described above into Long Evans rats either normally hydrated or drinking a 2% NaCl solution for 3 or 6 d. As was observed in Wistar rats, hyperosmotic stimuli induced local hypoxia within the hypothalamic magnocellular nuclei of Long Evans rats, correlated with increased cell proliferation and neuronal expression of VEGF (Fig. 3Go, A–H). In Brattleboro homozygous rats, the conditions of osmotic stimulation were attested by plasma osmolality levels (380 ± 7 mosmol; n = 4) comparable to that detected in 6-d osmotically stimulated Long Evans rats (360 ± 5 mosmol; n = 3), and higher than that detected in normally hydrated Long Evans rats (340 ± 6 mosmol; n = 3). Contrasting with nonmutant osmotically stimulated rats however, the hypothalamic magnocellular nuclei of Brattleboro rats were always devoid of hypoxia or cell proliferation markers, and their neurons exhibited very low VEGF expression (Fig. 3Go, E–H). These data indicate a strong association between the hypoxia and angiogenesis processes induced by hyperosmotic stimulation and the secretion of hypothalamic AVP, with no relation to the functional activation of magnocellular neurons.


Figure 3
View larger version (101K):
[in this window]
[in a new window]

  		FIG. 3. Comparison of the effects of prolonged osmotic stimulation in the SON of Long Evans and homozygous Brattleboro rats. Pimonidazole hydrochloride and BrdU were injected into the rats before fixation, and adjacent sections passing through the SON were double immunostained for AVP and VEGF (A, B, E, F, I, and J) or single immunostained for hypoxyprobe-1 and BrdU (C, D, G, H, K, and L). A–H, In the SON of Long Evans rats containing numerous AVP-immunostained neurons (A and E), prolonged osmotic stimulation induces a marked increase of 1) neuronal immunostainings for VEGF (F vs. B) and hypoxyprobe-1 (G vs. C) and 2) the number of proliferative (BrdU-immunostained) cells (H vs. D); I–L, in the SON of a homozygous Brattleboro rat, the absence of AVP-immunostained structures (I) associates with faint immunostainings for VEGF (J), hypoxyprobe-1 (K), and BrdU (L). oc, Optic chiasm. Scale bar, 100 µm (A–L).

 
Tissue hypoxia and angiogenesis are not induced by increased circulating AVP
We then tested whether an increased concentration of circulating AVP may affect intrahypothalamic oxygenation and angiogenesis, independently of the functional activation of magnocellular neurons. For this, an osmotic minipump containing AVP (diffusing 13 µg AVP per day) was implanted into normally hydrated Wistar rats (AVP-infused rats) and the effects on hypothalamic hypoxia and angiogenesis compared with that observed in normally hydrated or osmotically stimulated Wistar rats implanted with an osmotic minipump containing PBS. Control, osmotically stimulated, and AVP-infused rats were fixed 6 d after the minipump implantation, after the administration of both pimonidazole hydrochloride and BrdU (as described in Materials and Methods). As compared with osmotically stimulated rats, no effect on tissue hypoxia or cell proliferation was detected throughout the hypothalamus of PBS- or AVP-infused normally hydrated rats (Fig. 4Go, A–F). Measurements of plasma AVP concentrations indicated that AVP-infused rats exhibited very high AVP circulating levels (74 pg/ml ± 7; n = 4) compared with osmotically stimulated (15 pg/ml ± 5; n = 3) and control (5 pg/ml ± 1; n = 3) rats (Fig. 4Go). These data indicate that increased levels of circulating AVP are not responsible for the angiogenic processes detected in the hypothalamus after osmotic stimuli.


Figure 4
View larger version (59K):
[in this window]
[in a new window]

  		FIG. 4. Correlation between circulating AVP levels and SON hypoxia and cell proliferation. Pimonidazole hydrochloride and BrdU were injected into the rats before fixation, and adjacent sections passing through the SON were immunostained for hypoxyprobe-1 or BrdU. Intense tissue hypoxia and cell proliferation are observed in the SON of the osmotically stimulated rat (C and D) but are very low in the control (A and B) and the AVP-infused (E and F) rat. The lower panel shows plasma AVP concentrations measured in the control, osmotically stimulated, and AVP-infused rats. oc, Optic chiasm. Scale bar, 100 µm (A–F).

 
Contractile arterioles afferent to the SON are closely surrounded by AVP-containing dendrites
An osmotic stimulus increases the release of endogenous AVP at two levels of the hypothalamic neurons: 1) the axon terminals located in the neurohypophysis, from where AVP enters the general circulation, and 2) the dendrites and cell bodies located in the hypothalamic magnocellular nuclei, from where AVP enters the intercellular space within the nuclei (8, 9). AVP is generally considered a potent vasoconstrictor that acts on the V1a-R expressed by perivascular smooth muscle cells (10). We therefore wondered whether dendritic release of AVP could influence the oxygen supply to hypothalamic magnocellular nuclei, via vasoconstrictive effects on the contractile arterioles afferent to these nuclei. To investigate this theory, we first examined whether AVP-containing dendrites present within the hypothalamic magnocellular nuclei exhibited intimate anatomical relationships with the contractile arterioles running through these nuclei. Brain sections were first immunostained for either rat endothelial cell antigen 1 (RECA-1), a general marker of rat CNS vessels, or {alpha}-actin, a marker of perivascular smooth muscle cells. The {alpha}-actin immunostaining within the SON was found associated with a minority of the blood vessels projecting through the nucleus, including one or two large pial arteries running along the ventral surface and one to three connected large vessels crossing ventrodorsally through the nucleus (Fig. 5Go, A and B). Sections double immunostained for AVP and {alpha}-actin indicated that the contractile arterioles crossing through the SON are surrounded by numerous AVP-containing processes (Fig. 5CGo). Electron microscopy further revealed that throughout the SON, dendritic processes containing AVP-immunostained secretory granules frequently formed direct contacts with smooth muscle-like cells surrounding the endothelial cells of large arteriole-like blood vessels (Fig. 5DGo). These data strongly suggest that AVP locally released by SON dendrites have a privileged access to the perivascular smooth muscle cells known to express V1a-R.


Figure 5
View larger version (81K):
[in this window]
[in a new window]

  		FIG. 5. Anatomical relationships between contractile arterioles afferent to the SON and AVP containing dendrites. A, Section immunostained for the endothelial cell marker RECA-1 showing that, in addition to numerous interconnected capillaries, the SON contains large vessels connected with pial arteries located on the ventral brain surface (arrow); B, section immunostained for {alpha}-actin, a marker of perivascular smooth muscle cells, showing that the contractile cells correspond to large vessels crossing through the SON (arrows) or located on its ventral surface (arrowhead); C, color image of superimposed double immunostaining for {alpha}-actin and AVP showing that contractile vessels crossing through the SON are closely surrounded by numerous AVP-containing structures; D, electron microscope image of an AVP-immunostained section showing close apposition of a dendrite containing numerous AVP-immunostained secretory granules (sg) and a perivascular smooth muscle cell (smc). BV, Blood vessel; oc, optic chiasm. Scale bars, 100 µm (A–C) and 1 µm (D).

 
Arterioles afferent to the SON are constricted by perivascular AVP application
We next decided to test the vasoconstrictor effects of perivascular application of AVP on arterioles afferent to the SON. For this, cerebral vessels were labeled by the intraaortic injection of R-WGA. The brain was then rapidly dissected and the pial arterioles branching toward the SON positioned under a fluorescence stereo microscope equipped with a CCD camera. We found that the addition of AVP (10–9 M) but not OT to the incubation medium induced a rapid (within 1 min) and long-lasting (more than 15 min) constriction of the arterioles adjacent to the SON (Fig. 6Go, A–C). This vasoconstriction was reversed within 5 min after rinsing with incubation medium containing no AVP (not shown). Such a constriction was not observed when the brain was preincubated for 15 min before AVP addition with d(CH2)5TyrMe2AVP, a selective, water-soluble V1a-R antagonist (Fig. 6Go, D–E). These data indicate that local perivascular application of AVP induces a very strong, V1a-R-dependent, constriction of the arterioles afferent to the hypothalamic magnocellular nuclei, strongly suggesting that the diameter of these arterioles can be influenced by the AVP released from adjacent AVP-containing dendrites.


Figure 6
View larger version (110K):
[in this window]
[in a new window]

  		FIG. 6. Effects of perivascular application of AVP on the constriction of SON afferent arterioles. After deep anesthesia of the rats, cerebral vessels were labeled by intraaortic injection of PBS followed by R-WGA. The brain was dissected and glued in a petri dish filled with an oxygenated buffer milieu under a fluorescence stereo microscope equipped with an EM-CCD camera. A, Low-magnification image showing the organization of WGA-labeled vessels at the ventral surface of the hypothalamus, with arrows pointing to the arteriolar segments afferent to the SON; B and C, high-magnification images of arteriolar segments afferent to the SON incubated in KREBS buffer alone and 5 min after the addition of AVP (10–9 M); D and E, high-magnification images of the arteriolar segments afferent to the SON incubated 15 min in KREBS buffer containing a V1a-R antagonist (Manning antagonist, 10–7 M), and 5 min after the addition of AVP (10–9 M) to the same medium. Note that the strong constriction of the arteriolar segments induced by addition of AVP to the incubation medium (C vs. B) is completely inhibited by prior incubation of the preparation in the presence of the V1a-R antagonist (E vs. D). oc, Optic chiasm; on, optic nerve. Scale bar, 100 µm.

 
Intracerebral administration of a specific antagonist to V1a-R inhibits hypoxia and angiogenesis induced by osmotic stimuli
Because the vasoconstrictor effects of AVP involve its action on V1a-R expressed by perivascular smooth muscle cells, we tested whether inhibiting these receptors in vivo modifies the earlier observed angiogenic response to hyperosmotic stimulation. For this, specific antagonists to V1a-R were administrated to osmotically stimulated rats either ip (to inhibit the peripheral receptors) or intracerebrally (to inhibit the central receptors) (see Materials and Methods). After 3 d of peripheral or central administration, rats received injections of pimonidazole hydrochloride and BrdU and were then fixed as described above. In ip injected rats, we detected a strong labeling for hypoxia and cell proliferation markers within the magnocellular hypothalamic nuclei of rats injected with either the vehicle or the V1a-R antagonist (Fig. 7Go, A and B). We observed similar hypoxyprobe-1 and BrdU labeling to that found in osmotically stimulated rats in intracerebrally infused rats implanted with PBS-infusing pumps (not shown). In contrast to these rats, we observed markedly decreased labeling within the hypothalamic magnocellular nuclei of rats implanted with a V1a-R antagonist intracerebral infusing pump (Fig. 7Go, C and D). Although detectable on both hypothalamic sides, such modifications always appeared more important in the hypothalamic nuclei ipsilateral to the cannula implantation. We also determined whether the effects of the intracerebral infusion of V1a-R antagonist evidenced here could be related to an inhibition of the osmotically induced AVP release. Our observations indicated that the response of AVP neurons to prolonged osmotic stimulus was not affected by the intracerebral infusion of V1a-R antagonist; as compared with normally hydrated rats, AVP immunostaining of cell bodies and axon terminals was markedly decreased in osmotically stimulated rats receiving intracerebroventricular infusions of PBS either alone or containing the V1a-R antagonist (data not shown). Together, these data indicate that V1a-R located within the brain, but not the periphery, are involved in the hypoxia and angiogenesis processes induced by osmotic stimulation.


Figure 7
View larger version (72K):
[in this window]
[in a new window]

  		FIG. 7. Effects of ip and intracerebral administrations of V1a-R antagonists on hypoxia and angiogenesis. Pimonidazole hydrochloride and BrdU were injected before fixation in 3-d osmotically stimulated rats receiving or not a V1a-R antagonist. Adjacent sections passing through the SON were then immunostained for either hypoxyprobe-1 or BrdU. SON hypoxia and cell proliferation induced by osmotic stimulation are unaffected by the ip administration of the V1a-R antagonist SR49059 (200 µg/d for 3 d) (A–D) but markedly inhibited after intracerebral administration of the Manning V1a-R antagonist (14 µg/d for 3 d) (E–H). The schematic drawing illustrates the mode of intracerebral administration, with a cannula connected to an Alzet osmotic pump positioned into the lateral ventricle. Lower panel, Quantitative evaluation of the effects of ip or intracerebral administrations of V1a-R antagonists. The number of cells labeled for hypoxyprobe-1 or BrdU were counted within the SON and expressed as means ± SEM for the number of animals (n). **, P < 0.01. lv, Lateral ventricle; oc, optic chiasm; OS 3d, 3-d osmotic stimulation; veh, vehicle. Scale bar, 100 µm (A–H).

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
We have previously shown in the adult rat that prolonged osmotic stimuli induce a local and reversible angiogenesis within hypothalamic magnocellular nuclei, in association with an increased expression of VEGF by the magnocellular neurons (4). Yet, the mechanisms linking the osmotic stimulus to VEGF synthesis remained obscure. Because hypoxia is considered a main regulator of VEGF expression (11), we decided to test whether tissue hypoxia was involved in the angiogenic processes induced by osmotic stimulation.

Under physiological or pathological conditions, cells are able to sense a decrease in oxygen tension by activating HIF-1{alpha}. During normoxia, HIF-1{alpha} protein is continuously expressed and rapidly degraded, whereas during hypoxia, suppression of this degradation leads to an accumulation of the protein in the nucleus where it promotes the expression of several hypoxia-inducible genes including VEGF (12). Here we have demonstrated that, after prolonged osmotic stimulation, the increased expression of VEGF by neurons of the hypothalamic magnocellular nuclei is closely related to the occurrence of both local tissue hypoxia and intranuclear accumulation within these neurons of the transcription factor HIF-1{alpha}. Together, these data strongly support the idea that local hypoxia is the main stimulus for the increased expression of VEGF observed within the hypothalamic magnocellular neurons.

The question then was what cellular mechanisms could link an osmotic stimulus to local tissue hypoxia. Because osmotic stimulation is known to lead to strong functional activation of AVP magnocellular neurons (13), a first possibility was that tissue hypoxia within the hypothalamic magnocellular nuclei results from the increase in oxygen metabolism within these nuclei. Three main arguments, however, contradict this hypothesis: 1) increases in neuronal activity induce only very transient local hypoxia, and the increase in oxygen metabolism is rapidly compensated by increased local blood flow (14, 15); 2) hypoxia and angiogenesis were not observed in the hypothalamic magnocellular nuclei of lactating female, rats, despite the chronic activation of the OT-secreting neurons (G. Alonso, unpublished data); and 3) within the hypothalamic magnocellular nuclei of Brattleboro AVP-deficient rats, neither hypoxia nor angiogenesis occur, although the magnocellular hypothalamic neurons remain continuously stimulated. Interestingly, this latter finding clearly demonstrates the major role played by hypothalamic AVP in the induction of hypoxia and angiogenesis by osmotic stimulation. Because AVP is a potent constrictor of arterioles (16, 17, 18), we examined the hypothesis that the tissue hypoxia occurring after osmotic stimuli may result from vasoconstriction of the arterioles afferent to the hypothalamic magnocellular nuclei. After osmotic stimulation, hypothalamic AVP is secreted at the level of neurohypophyseal axon terminals from where it enters the general circulation (19), but also at the level of neuronal cell bodies and dendrites, from where it diffuses into the extracellular space (20, 21). Our data clearly indicate that increased levels of circulating AVP cannot account for the intrahypothalamic hypoxia and angiogenesis induced by osmotic stimulation. Indeed, neither hypoxia nor angiogenesis could be detected within the hypothalamic nuclei with increased levels of circulating AVP independent of activated AVP neurons (i.e. by the infusion of large amounts of exogenous peptide to normally hydrated rats). On the other hand, several arguments support the hypothesis that locally released AVP by dendrites and cell bodies may induce constriction of the arterioles afferent to the hypothalamic nuclei: 1) the concentration of AVP in the extracellular fluid of the SON has been reported to be 100- to 1000-fold higher and its half-life longer (20 vs. 2 min) than that in the blood (20); 2) AVP-containing dendrites establish preferential anatomical relationships with the contractile arteries penetrating the SON; 3) perivascular application of AVP induced strong, V1a-R-dependent constriction of arterioles afferent to the SON; and 4) hypoxia and angiogenesis induced by osmotic stimulation were strongly inhibited by intracerebral but not ip administration of specific antagonists to V1a-R.

Because it is likely that the local angiogenesis induced by osmotic stimulation surpasses the simple adjustment of local blood flow, one may ask whether this process responds to specific physiological needs. A surprising feature of hypothalamic magnocellular neurons is that their cell size is reversibly modified in response to modifications of the body fluid osmolality (22, 23). Within the SON for instance, the cell size of both AVP and OT neurons is increased by approximately 2-fold in 7-d osmotically stimulated rats (23). It can thus reasonably be assumed that the induction of angiogenesis in these nuclei serves to provide vascularization of the expanded volume occupied by these neurons.

In conclusion, our data strongly support the idea that the angiogenesis induced by osmotic stimulation within the hypothalamic magnocellular nuclei of adult rats results from local tissue hypoxia induced by the constriction of afferent arterioles via the secretion of dendritic AVP and the activation of perivascular V1a-R. This is the first illustration of a new, surprising role of AVP within the adult CNS.


   Acknowledgments
 
We thank Dr. C. Serradeil-LeGal (Sanofi-Aventis, Toulouse, France) and Dr. M. Manning (Toledo, OH) for their generous gifts of selective AVP receptor antagonists. Confocal and electron microscopy were achieved using the facilities of CRIC (Montpellier, France).


   Footnotes
 
Disclosure Statement: The authors have nothing to disclose.

First Published Online May 15, 2008

Abbreviations: AN, Accessory nucleus; AVP, arginine vasopressin; BrdU, bromodeoxyuridine; CNS, central nervous system; HIF, hypoxia-inducible factor; OT, oxytocin; PVN, paraventricular nucleus; RECA-1, rat endothelial cell antigen 1; R-WGA: rhodamine-conjugated wheat germ agglutinin; SON, supraoptic nucleus; V1a-R, V1a receptor; VEGF, vascular endothelial growth factor.

Received March 25, 2008.

Accepted for publication May 6, 2008.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Black JE, Sirevaag AM, Greenough WT 1987 Complex experience promotes capillary formation in young rat visual cortex. Neurosci Lett 83:351–355[CrossRef][Medline]
  2. Black JE, Isaacs KR, Anderson BJ, Alcantara AA, Greenough WT 1990 Learning causes synaptogenesis, whereas motor activity causes angiogenesis, in cerebellar cortex of adult rats. Proc Natl Acad Sci USA 87:5568–5572[Abstract/Free Full Text]
  3. Swain RA, Harris AB, Wiener EC, Dutka MV, Morris HD, Theien BE, Konda S, Engberg K, Lauterbur PC, Greenough WT 2003 Prolonged exercise induces angiogenesis and increases cerebral blood volume in primary motor cortex of the rat. Neuroscience 117:1037–1046[CrossRef][Medline]
  4. Alonso G, Galibert E, Duvoid-Guillou A, Vincent A 2005 Hyperosmotic stimulus induces reversible angiogenesis within the hypothalamic magnocellular nuclei of the adult rat: a potential role for neuronal vascular endothelial growth factor. BMC Neurosci 6:20[CrossRef][Medline]
  5. Tribollet E, Raufaste D, Maffrand J, Serradeil-Le Gal C 1999 Binding of the non-peptide vasopressin V1a receptor antagonist SR-49059 in the rat brain: an in vitro and in vivo autoradiographic study. Neuroendocrinology 69:113–120[CrossRef][Medline]
  6. Schmale H, Ivell R, Breindl M, Darmer D, Richter D 1984 The mutant vasopressin gene from diabetes insipidus (Brattleboro) rats is transcribed but the message is not efficiently translated. EMBO J 3:3289–3293[Medline]
  7. Morris JF 1982 The Brattleboro magnocellular neurosecretory system: a model for the study of peptidergic neurons. Ann NY Acad Sci 394:54–71[CrossRef][Medline]
  8. Ludwig M, Horn T, Callahan MF, Grosche A, Morris M, Landgraf R 1994 Osmotic stimulation of the supraoptic nucleus: central and peripheral vasopressin release and blood pressure. Am J Physiol 266:E351–E356
  9. Ludwig M, Callahan MF, Neumann I, Landgraf R, Morris M 1994 Systemic osmotic stimulation increases vasopressin and oxytocin release within the supraoptic nucleus. J Neuroendocrinol 6:369–373[CrossRef][Medline]
  10. Serradeil-Le Gal C, Herbert JM, Delisee C, Schaeffer P, Raufaste D, Garcia C, Dol F, Marty E, Maffrand JP, Le Fur G 1995 Effect of SR-49059, a vasopressin V1a antagonist, on human vascular smooth muscle cells. Am J Physiol 268:H404–H410
  11. Kuo NT, Benhayon D, Przybylski RJ, Martin RJ, LaManna JC 1999 Prolonged hypoxia increases vascular endothelial growth factor mRNA and protein in adult mouse brain. J Appl Physiol 86:260–264[Abstract/Free Full Text]
  12. Jewell UR, Kvietikova I, Scheid A, Bauer C, Wenger RH, Gassmann M 2001 Induction of HIF-1{alpha} in response to hypoxia is instantaneous. FASEB J 15:1312–1314[Free Full Text]
  13. Bisset GW, Chowdrey HS 1988 Control of release of vasopressin by neuroendocrine reflexes. Q J Exp Physiol 73:811–872[Free Full Text]
  14. Lindauer U, Vogt J, Schuh-Hofer S, Dreier JP, Dirnagl U 2003 Cerebrovascular vasodilation to extraluminal acidosis occurs via combined activation of ATP-sensitive and Ca2+-activated potassium channels. J Cereb Blood Flow Metab 23:1227–1238[CrossRef][Medline]
  15. Lindauer U, Royl G, Leithner C, Kuhl M, Gold L, Gethmann J, Kohl-Bareis M, Villringer A, Dirnagl U 2001 No evidence for early decrease in blood oxygenation in rat whisker cortex in response to functional activation. Neuroimage 13:988–1001[Medline]
  16. Fernandez N, Martinez MA, Garcia-Villalon AL, Monge L, Dieguez G 2001 Cerebral vasoconstriction produced by vasopressin in conscious goats: role of vasopressin V1 and V2 receptors and nitric oxide. Br J Pharmacol 132:1837–1844[CrossRef][Medline]
  17. Bein B, Cavus E, Dorges V, Stadlbauer KH, Tonner PH, Steinfath M, Scholz J 2005 Arginine vasopressin reduces cerebral oxygenation and cerebral blood volume during intact circulation in swine: a near infrared spectroscopy study. Eur J Anaesthesiol 22:62–66[CrossRef][Medline]
  18. Cavus E, Dorges V, Wagner-Berger H, Stadlbauer KH, Steinfath M, Wenzel V, Bein B, Scholz J 2005 Changes of local brain tissue oxygen pressure after vasopressin during spontaneous circulation. Acta Neurochir (Wien) 147:283–290; discussion 290[CrossRef][Medline]
  19. Andersson B, Leksell LG, Rundgren M 1982 Regulation of water intake. Annu Rev Nutr 2:73–89[CrossRef][Medline]
  20. Ludwig M, Leng G 2006 Dendritic peptide release and peptide-dependent behaviours. Nat Rev Neurosci 7:126–136[CrossRef][Medline]
  21. Ludwig M, Bull PM, Tobin VA, Sabatier N, Landgraf R, Dayanithi G, Leng G 2005 Regulation of activity-dependent dendritic vasopressin release from rat supraoptic neurones. J Physiol 564:515–522[Abstract/Free Full Text]
  22. Hatton GI, Walters JK 1973 Induced multiple nucleoli, nucleolar margination, and cell size changes in supraoptic neurons during dehydration and rehydration in the rat. Brain Res 59:137–154[CrossRef][Medline]
  23. Zhang B, Glasgow E, Murase T, Verbalis JG, Gainer H 2001 Chronic hypoosmolality induces a selective decrease in magnocellular neurone soma and nuclear size in the rat hypothalamic supraoptic nucleus. J Neuroendocrinol 13:29–36[CrossRef][Medline]
  24. Paxinos G, Watson C1986 The rat brain in stereotaxic coordinates. 2nd ed. London: Academic Press Ltd.



This article has been cited by other articles:


Home page
 Physiol. Rev.Home page
C. Clapp, S. Thebault, M. C. Jeziorski, and G. Martinez De La Escalera
Peptide Hormone Regulation of Angiogenesis
Physiol Rev, October 1, 2009; 89(4): 1177 - 1215.
[Abstract] [Full Text] [PDF]

Home page
 EndocrinologyHome page
Q. J. Pittman and S. J. Mulligan
Dendritic Vasopressin Release: Reducing the Flow Makes Blood Vessels Grow
Endocrinology, September 1, 2008; 149(9): 4276 - 4278.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Alonso, G.
Right arrow Articles by Guillon, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Alonso, G.
Right arrow Articles by Guillon, G.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals