Fluorescent Vasotocin Conjugate for Identification of the Target Cells

Publication Date (Web): June 19, 2004. Copyright © 2004 American Chemical .... R MIDDLETON , B KELLAM. Current Opinion in Chemical Biology 2005 9 (5)...
0 downloads 0 Views 164KB Size
Download PDF
Bioconjugate Chem. 2004, 15, 909−914

909

Fluorescent Vasotocin Conjugate for Identification of the Target Cells for Brain Actions of Vasotocin Christine M. Lewis,† E. Kurt Dolence,†,§ Zhaojie Zhang,‡ and James D. Rose†,‡,* Neuroscience Program, School of Pharmacy, and Department of Zoology and Physiology, University of Wyoming, Laramie, Wyoming 82071-3166. Received March 19, 2004; Revised Manuscript Received May 5, 2004

The effects of neuropeptides on the brain are a major focus of neuroendocrine research, and little progress has been made in the identification of the target neurons for many neuropeptides. Arginine8vasotocin (AVT) is a neurohypophyseal peptide present in nonmammalian vertebrates that controls many neural and behavioral functions. Here we describe synthesis and functional characterization of an AVT-Oregon green conjugate 1 (AVT-OG 1) that can be used in vivo to identify AVT target neurons. Application of AVT-OG 1 to the brainstem of an amphibian produces rapid, endosome-like internalization together with typical AVT-like neurophysiological effects. Thus, preparation of AVTOG 1, which preserves the peptide’s neurophysiological effects, is useful as a fluorescent marker for AVT target neurons. Consequently, AVT-OG 1 conjugate will have considerable utility for analyzing the neural actions of AVT in the intact brain.

INTRODUCTION

Vasopressin-like and oxytocin-like peptides are members of a family of highly related neurohypophyseal nonapeptides that are present in all animals. Arginine8vasotocin (AVT) is the specific peptide that serves antidiuretic and pressor functions in all vertebrate taxa except mammals, most of which produce arginine8vasopressin (AVP) for these functions (1). These two neuropeptides differ only in that AVP has phenylalanine rather than isoleucine in position 3 (2). In addition to the antidiuretic and pressor effects of these neuropeptides, AVT and AVP appear to act in the vertebrate brain and spinal cord to mediate a diverse array of complex functions, particularly regulation of reproductive behaviors (3, 4). In addition, neural development and control of cellular responses to neuronal injury may be influenced by vasopressin-like neuropeptides (5, 6). Numerous studies of mammalian and nonmammalian vertebrate species suggest that the neurobehavioral effects of AVT and AVP are mediated by V1-like receptors in the central nervous system (3, 7-10). Agonist activation of G-protein-coupled receptors, such as the receptors mediating AVT and AVP behavioral effects, triggers a cascade of events, including internalization of the ligandreceptor complex (11, 12). This internalization may modulate cell responsiveness by causing the loss of receptors from the cell surface, making them inaccessible to continued ligand challenge (13, 14). Current understanding of the cellular actions of vasopressin-like neuropeptides is very uneven, ranging from a detailed account of the target cells and mechanisms of action for antidiuretic functions involving the kidney, to much less complete mechanistic understanding of neural * To whom correspondence should be addressed: Department of Zoology and Physiology, Box 3166, University of Wyoming, Laramie, WY 82071-3166. E-mail: [email protected], Phone: 307-766-6719 Fax: 307-766-2921. † Neuroscience Program. § School of Pharmacy. ‡ Department of Zoology and Physiology.

and behavioral actions. A limiting factor in elucidation of cellular mechanisms underlying neural-behavioral effects of vasopressin-like neuropeptides has been the absence of a means to identify the specific target neurons for AVT or AVP. Although receptors for these neuropeptides have been localized in certain neurons, the demonstration of receptors, per se, does not reveal the functional effects mediated by such receptors or even guarantee that histologically demonstrated receptors are functional. In addition, the diversity of structural and functional phenotypes of neurons in the brain, unlike structurally simpler peripheral target tissues such as the kidney, further complicates the matter of relating neuropeptide target cells to a role in neural function. The purpose of the present paper is to document the preparation of a fluorescent Oregon Green 488 AVT conjugate 1 (AVT-OG 1), validate its functional effects on neurons, and demonstrate its utility as a means of identifying the neuronal target cells for AVT. These steps in validating the utility of the fluorescent AVT conjugate will permit the use of this conjugate in the elucidation of this neuropeptide’s mechanism of neurobehavioral action. Research in this laboratory uses a comparatively simple model for analysis of the neurobehavioral actions of AVT. Roughskin newts (Taricha granulosa) exhibit a highly stereotyped pattern of mating behavior in which the male clasps the female with all four legs from a dorsal approach for a period of hours as an obligatory step in the courtship process leading to fertilization (15). It has been established that AVT is one of the hormones normally facilitating this behavior and the lower brainstem is the key brain region for AVT action in facilitating clasping (4). In addition, we have shown that the claspfacilitating neural action of AVT appears to entail an enhancement of the responsiveness of medullary neurons to sensory stimuli applied to various locations on the body surface (16). To visualize peptide internalization in endosome-like vesicles, previous studies of peripheral tissues have utilized fluorescent AVT (17), fluorescent AVP (12), and

10.1021/bc049928x CCC: $27.50 © 2004 American Chemical Society Published on Web 06/19/2004

910 Bioconjugate Chem., Vol. 15, No. 4, 2004

Lewis et al.

Scheme 1. Synthesis of Oregon Green 488-[Arg8]-Vasotocin Conjugate AVT-OG (1)a

a (a) 5′-Isomer Oregon Green 488, DMF, NOHSu, EDC‚HCl, 21 h, 22 °C; (b) [Arg8]-vasotocin, 0.1 M aq NaHCO , 24 h, 22 °C; 3 (c) lyophilization; (d) C18 RP-TLC followed by preparative C18 RP-HPLC.

fluorescent receptor agonists or antagonists (18). However, the use of such conjugates has been confined to in vitro studies of nonneural cells (11, 12, 14, 17-20). The most significant aspect of our use of this fluorescent vasopressin-like neuropeptide AVT-OG 1 is in vivo application to the brain in such a way that specific neuronal target cells can be visualized while the concurrent neuronal effects of AVT are neurophysiologically monitored. In addition, the fluorescent AVT-OG 1 that we are presenting appears to be comparable in neuropotency to unlabeled AVT and retains excellent fluorescence. Using this approach, we aim, for the first time, to connect the functional effects of AVT with the neuropeptide’s specific target neurons and their structural characteristics. The analytic power of this in vivo use of the fluorescent neuropeptide AVT-OG 1, particularly in the roughskin newt model, is considerable, in that it can also be used in conjunction with real-time techniques such as optical imaging of signaling events. EXPERIMENTAL PROCEDURES

Synthesis of Oregon Green 488 Conjugated Arginine8-Vasotocin 1 (AVT-OG 1) (Scheme 1). Reagents. Unless otherwise stated, reagents and solvents were obtained from commercial sources without further purification. Distilled deionized water was obtained from a Millipore NanoPure system. C18 reverse phase TLC was conducted using Merck RP-18 F254 glass-backed TLC plates. Preparative C18 RP-HPLC was conducted using a Phenomenex Juptiter 10 µm 300A 250 × 21.2 mm column while monitoring at both 254 and 494 nm with a flow rate of 1.0 mL/min. ESI mass spectrometry was obtained using a mixture of 50:50 CH3CN:H2O containing 0.1% TFA. Oregon Green 488-[Arg8]-Vasotocin Conjugate (1) (AVT-OG 1). The 5′-isomer of Oregon Green 488 (0.5 mg, 1.1 µM, Molecular Probes) was dissolved in DMF (56 µL) in a foil-covered vial. To this solution was added 22 µL (2.2 µmol) of a stock solution of N-hydroxysuccinimide (0.1 µM in DMF) and 22 µL (2.2 µmol) of a stock solution of 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride (0.1 µM in DMF). This mixture was stirred at 22 °C for 21 h. This active ester solution was

transferred to the manufacturer’s vial of [Arg8]-vasotocin acetate salt (1.0 mg, 1.0 µM Sigma, free base MW 1050.2) dissolved in 100 µL of 0.1 M aq NaHCO3 with the aid of 100 µL of DMF and 100 µL of 0.1 M aq NaHCO3. The vial was covered with foil, and this mixture was stirred vigorously for 24 h at 22 °C. At this time, the mixture was transferred to a 25 mL pear shaped round-bottom flask with the aid of water, frozen, and lyophilized to afford a yellow film. This film was applied using 50:50 CH3CN:H2O containing 0.1% TFA to one 20 × 20 cm RP18 F254 TLC plate and eluted with 40:60 CH3CN:H2O containing 0.1% TFA. The major yellow band (Rf 0.31) was removed by scraping and eluted with 50:50 CH3CN: H2O containing 0.1% TFA, and the solvents were removed by evaporation in vacuo. This material was further purified by gradient preparative RP-HPLC (5% CH3CN: H2O containing 0.1% TFA (5 min) followed by a gradient over 40 min to a composition of 80:20 CH3CN:H2O containing 0.1% TFA) (tR ) 26.3 min). This afforded 0.6 mg (41% yield) of 1 as a water soluble yellow film. ESI mass spectrometry calcd for C64H75N15O18S2F2 (M + H)+ m/z 1444.5 found: C64H75N15O18S2F2 (M + H)+ m/z 1444.3. Neurophysiological Validation of the Physiological Activity of AVT-OG 1. Adult male roughskin newts (Taricha granulosa) were held in community tanks supplied with continuously flowing, aerated well water. The animals were fed chopped beef heart and maintained on a natural photoperiod. The University of Wyoming Animal Care and Use Committee approved all procedures involving animals. Procedures used for recording and analysis of neuronal activity were similar to those previously reported (16) and will be only briefly described here. Newts were anesthetized by immersion in 0.1% MS-222 (tricaine methanesulfonate). The rostral medulla was surgically exposed, and the head was stabilized with a support system attached to the skull. After surgical preparation was complete, the newt was removed from the anesthetic and partially immersed in fresh, aerated well water with the skin covered by wet tissue paper to facilitate transcutaneous respiration. For electrophysiological recording, newts were immobilized by administration of 2% solution of gallamine triethiodide in Ringer’s solution (0.2 mL, ip).

Brain Target Cells for Vasotocin

Recording began after sufficient time elapsed for the newt to recover from the anesthetic (approximately 2-4 h). Single neuron activity was recorded with 50 µm Diamel-insulated stainless steel microelectrodes (impedances of 100-500 kΩ) placed on the surface of the rostromedial medulla in the lower brainstem. At most recording sites, spikes were recorded simultaneously from two to five neurons. Two AVT-sensitive parameters of neuronal function were evaluated: spontaneous firing rates and responses to skin stimulation. Sensory responsiveness was tested by application of light pressure (4 s) to locations on the face, individual feet, and trunk with a cone-shaped probe mounted on a piano wire stylus to ensure constant force of stimuli. The stylus was fixed to a pressure transducer to record stimulus onset, duration, and offset. Stimuli were administered at 3 min intervals throughout each experiment. Prior to administration of a peptide or vehicle, four trials of stimulation were done to establish a baseline neuronal response. All newts received hormone or vehicle administration in 4 µL of physiological saline administered directly to the surface of the medulla via microsyringe. Neuronal activity, pressure transducer recordings, and a voice narrative were recorded on videotape in pulse code-modulated format by means of a Neurodata Neurocorder for later analysis. Firing by individual neurons was identified with Spike II software (Cambridge Electronic Designs). The criterion required for identification of neuronal sensory responses and the assessment of hormone effects on neuronal sensory responses and activity levels were similar to those described previously (16). Changes in activity were assessed by comparison of prehormone and posthormone response magnitudes. In order for a neuron’s spontaneous firing to be considered changed by AVT-OG 1, or AVT, it’s second-bysecond firing rate after the neuropeptide application had to be above or below each of the four prepeptide samples of firing by at least 30%. Likewise for a neuron’s response to a sensory stimulus to be considered increased, its second-by-second firing rate during a sensory stimulus after the neuropeptide application had to be above or below the four prepeptide records of firing during the same stimulus by at least 30%. In addition to evaluating the neurophysiological effectiveness of AVT-OG 1, we examined three aspects of the conjugate’s internalization by neurons: time course, minimally effective dose, and specificity. To this end, 4 µL of physiological saline containing one of three doses of AVT-OG 1 (10 ng, 40 ng, or 80 ng), AVT (80 ng), or vehicle alone were administered to the surface of the medulla and rinsed at the end of 30 min (n ) 4 newts for each group). To assess time course of internalization, two additional groups of newts (n ) 4) received 80 ng of AVT-OG 1 for 5, 30, or 60 min prior to rinsing. Small numbers of additional newts received 80 ng of AVT-OG 1 for various exposure times between 5 and 60 min (e.g. 15, 45 min) to provide a more complete qualitative picture of the time course of internalization. These dosages of the peptide and application volumes of vehicle were based on previous neurophysiological and behavioral experiments showing that a 4 µL volume of solution containing AVT in these relative concentrations and applied directly to the surface of the medulla produced reliable neurophysiological effects (16). To determine the specificity of neuronal internalization of AVT-OG 1 the medulla was pretreated with 4 µL of physiological saline containing either unlabeled AVT or Manning compound (21, 22), a V1-receptor antagonist. For each treatment condition (n ) 4), 1.76 mg of Manning compound (d(CH2)5-

Bioconjugate Chem., Vol. 15, No. 4, 2004 911

[Tyr(Me)2]AVP), or physiological saline (VEH) was administered followed 30 min later by 80 ng of AVT-OG 1 in 4 µL of physiological saline. Histological Localization of Fluorescent AVTOG 1 Internalization by Neurons. After completion of each experimental treatment, the surface of the medulla was rinsed twice with ice cold PBS (1-4 °C), the newt was immediately decapitated, and the head was rinsed for 30 s twice in ice-cold PBS (1-4 °C) before being placed in ice-cold 30% sucrose-formalin (1-4 °C) for a minimum of 24 h. Brains were removed, embedded in Tissue Tek OCT compound, and transversely sectioned using a cryostat. Sections cut through the medulla were selected at 100 µm intervals, beginning just caudal to the cerebellum, thaw-mounted on microscope slides, and dried for a minimum of 8 h at room temperature. Neuronal labeling by AVT-OG 1 was examined and digital images collected with a Leica TCS SP2 confocal laser-scanning microscope in the University of Wyoming Microscopy Core as well as with a conventional epifluoresence microscope (Olympus BX40, FITC filter set, and Optronics digital video camera system). Neuronal AVTOG 1 internalization was determined by counting the number of nucleated neurons with 10 or more fluorescent endosomes in sections through the rostral medulla. This value is consistent with other studies in which endosomes have been counted for quantification of target cells (13, 17-20). An endosome was defined as a spot of fluorescence greater than 0.1 µm but less than 0.5 µm in diameter (12). In each transverse section, neurons were counted from a zone extending from the lateral border of the medulla to the midline. This encompassed an area 1800 µm in width and 600 µm in depth in each section. Neurons were counted from sections every 100 µm in a zone extending from 0 µm to 2000 µm caudal to the posterior extent of the cerebellum. Numbers of labeled neurons were compared across treatment conditions with analysis of variance followed by Bonferroni critical value procedure. RESULTS

Neuronal Internalization of AVT-OG 1. Application of AVT-OG 1 to the surface of the medulla produced a time- and dose-dependent internalization of fluorescence (Figure 1) that exhibited properties consistent with previously reported (12, 14) demonstrations of neuropeptide internalization. When viewed by confocal microscopy, the intracellular labeling consisted of endosome-like foci of fluorescence approximately 0.2 µm in diameter (Figure 1b). The nuclei of labeled neurons were always devoid of labeling (Figures 1a and 1c). The endosome-like uptake of fluorescence was evident in numerous medullary neurons of all sizes, particularly those in the medial reticular nuclei. The labeled neurons tended to be in clusters that were separated by unlabeled ones (Figures 1c and 1e). Fluorescently labeled cells were present at the lowest AVT-OG 1 dose examined (10 ng); however, these cells were present primarily at the most dorsal aspects of the tissue section examined. At both 40 ng and 80 ng doses, the number of labeled cells increased, but consistent labeling throughout the depth of the medulla appeared only after administration of 80 ng of AVT-OG 1 (Figure 2). Exposure of medullary neurons to AVT-OG 1 for durations of 5, 30, or 60 min revealed intracellular fluorescent labeling, but the pattern of labeling differed significantly across the three time intervals (p < 0.01) (Figure 3). At 5 min, labeled cells were present, but the mean number of cells for the group

912 Bioconjugate Chem., Vol. 15, No. 4, 2004

Lewis et al.

Figure 1. Confocal and epifluorescent images of rostromedial medullary neurons showing AVT-OG 1 internalization. (A and C) Single confocal optical scans (2 µm) illustrating the distribution of labeling within neurons. The cytoplasm of the cell body and a proximal dendrite (d) were intensely labeled but not the nucleus (n) (scale bar ) 10 µm). (B) Confocal image of internalized AVTOG 1 localized in puncta similar in size and appearance to endosomes (scale bar ) 5 µm). (D and E) Epifluorescence images illustrating peripheral distribution of puncta in a neuronal cell body (D) at 15 min after AVT-OG 1 administration and redistribution of puncta throughout the cytoplasm of two neurons (E; small arrows) and a dendrite (d) at 30 min after AVT-OG 1 administration (scale bar ) 10 µm).

Figure 2. Mean number of labeled neurons in each group of newts as a function of dose after AVT-OG 1 administration (10 ng, 40 ng, or 80 ng) to the medulla. Neurons with AVT-OG 1 internalization were present at the lowest dose (10 ng) and the number of labeled cells increased substantially at the highest dose examined (80 ng).

Figure 3. Mean number of labeled neurons in each group of newts as a function of time after AVT-OG 1 administration. AVT-OG 1 was applied to the medulla and rinsed after 5, 30, or 60 min. Neurons with AVT-OG 1 internalization were present at 5 min and the number of labeled cells peaked at 30 min.

of newts was significantly less (mean ) 34.75) than at 30 min of exposure (mean ) 85.75; p < 0.01) when the number of fluorescently labeled neurons was highest. Labeling was typically clustered at the periphery of the neuronal cell bodies at 15 min (Figure 1d) but at 30 min was distributed throughout the cytoplasm of each cell as well as proximal portions of primary dendrites (Figure 1e). Labeling of fine diameter processes thin enough to be axons was not evident. The number of fluorescently labeled cells peaked at 30 min and was significantly lower by 60 min of exposure (mean ) 15.25). AVT-OG 1 Internalization Is Specific and Blocked by AVT Receptor Ligands. Fluorescent labeling of medullary neurons appeared to be specific to AVT receptors in that it differed significantly across pretreatment competition conditions (Figure 4). Labeling was greatly diminished by pretreatment with either unlabeled AVT (p < 0.01) or Manning compound (p < 0.01), a specific

V1-receptor antagonist. Vehicle pretreatment had no effect on AVT-OG 1 internalization and Oregon Green, by itself, did not label cells. AVT-OG 1 Internalization Is Associated with AVT-like Neurophysiological Effects. Application of AVT-OG 1 to the medulla produced changes in neuronal activity and sensory responsiveness (Figure 5) of the same types and magnitudes as previously seen with unlabeled AVT. For the AVT-OG 1 dose used in the present study (80 ng), most recorded neurons showed a pronounced increase in spontaneous firing within 3 min. Most of these neurons showed complete reversal of spontaneous firing to pretreatment values within 3254 min of hormone administration. The majority (60.5%) of these neurons showed increased magnitudes of response to somatic sensory stimuli applied to various body locations after hormone administration. These changes in spontaneous activity and sensory responsiveness were

Brain Target Cells for Vasotocin

Figure 4. Competition of AVT-OG 1 neuronal labeling due to pretreatment with AVT or Manning compound. Unlabeled AVT or the specific V1 receptor antagonist, Manning compound, administered immediately prior to AVT-OG 1, significantly decreased the mean number of labeled cells, whereas vehicle pretreatment had no effect on AVT-OG 1 internalization. Significant differences were found in comparisons between A vs E, B vs C, and B vs D. Oregon Green 488 alone (E) did not produce any labeling.

Bioconjugate Chem., Vol. 15, No. 4, 2004 913

AVT. The potent enhancement of neuronal spontaneous activity and responsiveness to somatic sensory stimuli were essentially identical to results previously obtained (16) with a comparable dose of unlabeled AVT. The cytological pattern of AVT-OG 1 uptake is consistent with an endosome-like pattern of internalization, like that previously described for toad bladder (18, 19) with an AVT agonist conjugate and for vasopressin agonist conjugates (11, 12, 14) in nonneuronal cells. The time course of AVT-OG 1 uptake and dispersal in endosome-like foci of fluorescence began in the perimeter of the cell body and progressed through the extranuclear cytoplasm of the neuronal somata and proximal dendrites. This pattern is also consistent with endosome formation. Our results are compatible with a V1 receptor-based process of AVT-OG 1 internalization. In the intact brain, competition by pretreatment with unlabeled AVT or a specific V1 receptor antagonist blocked AVT-OG 1 uptake, thus verifying the specificity of action of the conjugate. In addition, the fact that Manning compound pretreatment produced very pronounced blockade of AVT-OG 1 uptake shows that the neurons in the newt medulla may have predominantly V1 receptors. This result validates the utility of this conjugate 1 for investigations of neural effects of AVT due to the fact that V1like receptors mediate such neural effects (4, 7). In summary, this is the first use of a fluorescent AVT conjugate to investigate neuroendocrine function. In addition, we have verified that the AVT-OG conjugate 1 retains its receptor specificity and physiological effects. The availability of this fluorescent AVT makes new and analytically powerful investigations of neural and other cellular actions of AVT more feasible. ACKNOWLEDGMENT

Figure 5. Both AVT and AVT-OG 1 caused significant increases in neuronal sensory responses rapidly after hormone administration. Sensory responsiveness increased in 22 of 34 neurons after administration of AVT and 23 of 38 neurons increased in sensory responsiveness after AVT-OG 1 administration. Changes of this magnitude were not seen after vehicle administration (n ) 22).

not observed after vehicle application to the medulla (Figure 5). Collectively, these neurophysiological effects of AVT-OG 1 are essentially the same as those we have observed previously with unlabeled AVT, were similar to changes seen after administration of unconjugated AVT (Figure 5), and showed that the conjugate was functionally effective with a time course paralleling the process of internalization. DISCUSSION

The fluorescent conjugate AVT-OG 1 described here provides a significant new tool for analysis of this neuropeptide’s functions in cellular systems. One of the most active areas of investigation concerns AVT effects on brain functions, particularly those controlling behavior. Our results show that the AVT-OG conjugate 1 effectively labels target neurons for the neuropeptide and concurrently initiates the usual physiological actions of

This research was made possible by NIH-NCRR grant P20 RRO15553 to J.D.R. and the University of Wyoming’s Center of Biomedical Research Excellence in Cellular Signaling and NIH grant GM062139-01 to E.K.D. This research was also supported by the University of Wyoming COBRE Microscopy Core and the Department of Chemistry’s ESI-MS facilities. Supporting Information Available: ESI-MS and reverse phase HPLC for compound 1. This material is available free of charge via the Internet at http:// pubs.acs.org/BC. LITERATURE CITED (1) Norris, D. O. (1997) Vertebrate Endocrinology, pp 157-240, Academic Press, San Diego. (2) Moore, F. L. (1992) Evolutionary precedents for behavioral actions of oxytocin and vasopressin. Ann. N. Y. Acad. Sci. 652, 156-165. (3) Goodson, J. L., and Bass, A. H. (2001) Social behavior functions and related anatomical characteristics of vasotocin/ vasopressin systems in vertebrates. Brain Res. Rev. 35, 246265. (4) Rose, J. D., and Moore, F. L. (2002) Behavioral Neuroendocrinology of vasotocin and vasopressin and the sensorimotor processing hypothesis. Front. Neuroendocrinol. 23, 317-341. (5) Tribollet, E., Arsenijevic, Y., Marguerat, A., Barberis, C., and Dreifuss, J. J. (1994) Axotomy induces the expression of vasopressin receptors in cranial and spinal motor nuclei in the adult rat. Proc. Natl. Acad. Sci. U.S.A. 91, 9636-9640.

914 Bioconjugate Chem., Vol. 15, No. 4, 2004 (6) Oz, M., Kolaj, M., and Renaud, L. P. (2001) Electrophysiological evidence for vasopressin V(1) receptors on neonatal motoneurons, premotor and other ventral horn neurons. J. Neurophys. 86, 1202-1210. (7) Albers, H. E., Pollock, J., Simmons, W. H., and Ferris, C. F. (1986) A V1-like receptor mediates vasopressin-induced flank marking behavior in hamster hypothalamus. J. Neurosci. 6, 2085-2089. (8) Iwata, T., Toyoda, F., Yamamoto, K., and Kikuyama, S. (2000) Hormonal control of urodele reproductive behavior. Comp. Biochem. Physiol. B 126, 221-229. (9) Insel, T. R., Winslow, J. T., Wang, Z., and Young, L. J. (1998) Oxytocin, vasopressin, and the neuroendocrine basis of pair bond formation. Adv. Exp. Med. Biol. 449, 215-224. (10) Miller, F. L. and Moore, L. J. (1983) Arginine vasotocin induces sexual behavior of newts by acting on cells in the brain. Peptides 4, 97-102. (11) Briner, V. A., Williams, B., Tsai, P., and Schrier, R. W. (1992) Demonstration of processing and recycling of biologically active V1 vasopressin receptors in vascular smooth muscle. Proc. Natl. Acad. Sci. U.S.A. 89, 2854-2858. (12) Lutz, W., Salisbury, J. L., and Kumar, R. (1991) Vasopressin receptor-mediated endocytosis: current view. Am. J. Physiol. 261, F1-F13. (13) Tsao, P., Cao, T., and von Zastrow, M. (2001) Role of endocytosis in mediating downregulation of G-protein-coupled receptors. Trends. Pharmacol. Sci. 22, 91-96. (14) Innamorati, G., Le Gouill, C., Balamotis, M., and Birnbaumer, M. (2001) The long and the short cycle: Alternative

Lewis et al. intracellular routes for trafficking of G-protein-coupled receptors. J. Biol. Chem. 276, 13096-13103. (15) Propper, C. R. (1991) Courtship in the roush-skinned newt Taricha granulosa. Animal Behav. 41, 547-554. (16) Rose, J. D., Kinnaird, J. R. and Moore, F. L. (1995) Neurophysiological effects of vasotocin and corticosterone on medullary neurons: Implications for hormonal control of amphibian courtship behavior. Neuroendocrinology 62, 406417. (17) Eggena, P., and Buku, A. (1989) Synthesis and characterization of a long-acting fluorescent analog of vasotocin. Biol. Cell 66, 1-6. (18) Eggena, P., Lu, M., and Buku, A. (1990) Internalization of fluorescent vasotocin-receptor agonist and antagonist in the toad bladder. Am. J. Physiol. 259, C462- C470. (19) Eggena, P., Ma, C.-L., Lu, M., and Buku, A. (1990) A fluorescent analogue of hydrin 1: a new probe for vasotocin receptors. Am. J. Physiol. 259, E524-E528. (20) Buku, A., Masur, S., and Eggena, P. (1989) Synthesis and characterization of fluorescein- and rhodamine-labeled probes for vasotocin receptors. Am. J. Physiol. 257, E804-E808. (21) Sawyer, W. H., and Manning, M. (1984) The development of vasopressin antagonists. Fed. Proc. 43, 87-90. (22) Tripp, S. K., and Moore, F. L. (1988) Autoradiographic characterization of binding sites labeled with vasopressin in the brain of a urodele amphibian. Neuroendocrinology 48, 87-92.

BC049928X