Synthesis and biological activities of a fluorescent photoaffinity analog of vasopressin

Visualizing microtubule-dependent vasopressin type 2 receptor trafficking using a new high-affinity fluorescent vasopressin ligand

The vasopressin receptor type 2 (V2R) is the major target of vasopressin (VP) in renal epithelial cells. Although it is known that VP induces V2R internalization, accumulation in the perinuclear area, and degradation, the V2R intracellular trafficking pathways remain elusive. We visualized this process by developing a new fluorescent VP analog tagged by tetramethylrhodamine (TMR)-[Lys-(PEG)(2)-Suc-TMR(8)]VP or (VP(TMR)). This ligand is fully functional as revealed by its high binding affinity toward V2R [(K(d)) =157 ± 52 nM] and ability to increase intracellular cAMP 32-fold. VP(TMR) induced V2R internalization in LLC-PK1 cells expressing either a FLAG-tagged receptor (FLAG-V2R) or V2R C-terminally tagged with green fluorescent protein (GFP) (V2R-GFP). After internalization, VP(TMR) and V2R-GFP colocalized in the perinuclear area, suggesting that the hormone and receptor traffic along the same pathway. VP(TMR) and V2R colocalized initially with the early endosome markers EEA1 and Rab5, and later with the recycling and late endosome markers Rab11 and Rab25. Epifluorescence microscopy of LLC-PK1 cells expressing GFP-tagged microtubules (MT) showed that VP(TMR)-containing vesicles travel along the MT network, and even remain attached to MT during the metaphase and anaphase of mitosis. Colchicine, a MT-depolymerizing agent, abolished perinuclear accumulation of VP(TMR), and Western blot analysis showed that VP-induced V2R-GFP degradation is markedly retarded, but not abolished, by colchicine (10 μM). We conclude that the new VP(TMR) ligand is suitable for dissecting V2R and VP internalization and trafficking in cells, and that V2R trafficking and down-regulation is an MT-dependent mechanism.

Fluorescent agonists and antagonists for vasopressin/oxytocin G protein-coupled receptors: usefulness in ligand screening assays and receptor studies

Different series of fluorescent agonists and antagonists have been developed and characterized for arginine-vasopressin and oxytocin G protein-coupled receptors. Both cyclic and linear peptide analogs of the neurohypophysial hormones are useful tools for investigating receptor localization and trafficking, analysing receptor structural organization, and developing new receptor-selective high-throughput ligand screening assays.

An investigation of position 3 in arginine vasopressin with aliphatic, aromatic, conformationally-restricted, polar and charged amino acids

We report the solid-phase synthesis and some pharmacological properties of 23 new analogs of arginine vasopressin (AVP) which have the Phe3 residue replaced by a broad variety of amino acids. Peptides 1-9 have at position 3: (1) the mixed aromatic/aliphatic amino acid thienylalanine (Thi) and the aliphatic amino acids; (2) cyclohexylalanine (Cha); (3) norleucine (Nle); (4) Leu; (5) norvaline (Nva); (6) Val; (7) alpha-aminobutyric acid (Abu); (8) Ala; (9) Gly. Peptides 10-23 have at position 3: the aromatic amino acids, (10) homophenylalanine (Hphe): (11) Tyr; (12) Trp; (13) 2-naphthylalanine (2-Nal); the conformationally-restricted amino acids (14) Pro; (15) 2-aminotetraline-2-carboxylic acid (Atc); the polar amino acids (16) Ser; (17) Thr; (18) Gln; and the charged amino acids (19) Asp; (20) Glu; (21) Arg; (22) Lys; (23) Orn. All 23 new peptides were evaluated for agonistic and, where appropriate, antagonistic activities in in vivo antidiuretic (V2-receptor) and vasopressor (V1a-receptor) assays and in in vitro (no Mg2+) oxytocic assays. The corresponding potencies (units/mg) in these assays for AVP are: 323+/-16; 369+/-6 and 13.9+/-0.5. Peptides 1-9 exhibit the following potencies (units/mg) in these three assays: (1) 379+/-14; 360+/-9; 36.2+/-1.9; (2) 294+/-21: 73.4+/-2.7; 0.33+/-0.02; (3) 249+/-28; 84.6+/-4.3; 4.72+/-0.16; (4) 229+19; 21.4+/-0.6; 2.1+/-0.2; (5) 134+/-5; 31.2+/-0.9; 28.4+/-0.2; (6) 114+/-9; 45.3+2.3; 11.3+/-1.6; (7) 86.7+/-2.5; 4.29+/-0.13; 0.45+/-0.03; (8) 15.5+/-1.5; 0.16+/-0.01; approximately 0.02: (9) 3.76+/-0.03; < 0.02; in vitro oxytocic agonism was not detected. These data show that the aliphatic amino acids Cha, Nle, Leu, Nva and Val are well-tolerated at position 3 in AVP with retention of surprisingly high levels of antidiuretic activity. Peptides 2-9 exhibit significant gains in both antidiuretic/vasopressor (A/P) and antidiuretic/oxytocic (A/O) selectivities relative to AVP. [Thi3]AVP appears to be a more potent antidiuretic and oxytocic agonist than AVP and is equipotent with AVP as a vasopressor agonist. The antidiuretic potencies of peptides 10-23 exhibit drastic losses relative to AVP. They range from a low of 0.018+/-0.001 units/mg for the Lys3 analog (peptide 22) to a high of 24.6+/-4.6 units,mg for the Hphe3 analog (peptide 10). Their vasopressor potencies are also drastically reduced. These range from a low of < 0.002 units/mg for peptide 22 to a high of 8.99+0.44 units/mg for the Atc3 analog (peptide 15). Peptides 10-23 exhibit negligible or undetectable in vitro oxytocic agonism. The findings on peptides 10-23 show that position 3 in AVP is highly intolerant of changes with aromatic, conformationally-restricted, polar and charged amino acids. Furthermore, these findings are in striking contrast to our recent discovery that position 3 in the potent V2/V1a/OT antagonist d(CH2)5D-Tyr(Et)2VAVP tolerates a broad latitude of structural change at position 3 with many of the same amino acids, to give excellent retention of antagonistic potencies. The data on peptides 1-4 offer promising clues to the design of more potent and selective AVP V2 agonists.

Position three in vasopressin antagonist tolerates conformationally restricted and aromatic amino acid substitutions: a striking contrast with vasopressin agonists

We report the solid-phase synthesis and some pharmacological properties of 12 position three modified analogues (peptides 1-12) of the potent non-selective antagonist of the antidiuretic (V2-receptor), vasopressor (V1a-receptor) responses to arginine vasopressin (AVP) and of the uterine contracting (OT-receptor) responses to oxytocin (OT), [1(-beta mercapto-beta,beta-pentamethylenepropionic acid)-2-O-ethyl-D-tyrosine 4-valine] arginine vasopressin [d(CH2)5D-Tyr(Et)2VAVP] (A) and two analogues of (B) (peptides 13,14), the 1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid3 (Tic3) analogue of (A). Peptides 1-12 have the following substituents at position three in (A): (1) Pro; (2) Oic; (3) Atc; (4) D-Atc; (6) D-Phe; (7) Ile; (8) Leu; (9) Tyr; (10) Trp; (11) Hphe; (12) [HO]Tic; Peptide (13) is the Tyr-NH2(9) analogue of (B): Peptide (14) is the D-Cys(6) analogue of (B). All 14 new peptides were evaluated for agonistic and antagonistic activities in in vivo V2 and V1a assays and in vitro (no Mg2+)n oxytocic assays. With the exception of the D-Phe3 peptide (No. 6), which exhibits very weak V2 agonism (approximately 0.0017 U/mg), none of the remaining 13 peptides exhibit any agonistic activities in these assays. In striking contrast to their deleterious effects on agonistic activities in AVP, the Pro3, Oic3, Tyr3 and Hphe3 substitutions in (A) are very well tolerated, leading to excellent retention of V2, V1a and OT antagonistic potencies. All are more potent as V2 antagonists than the Ile3 and Leu3 analogues of (A). The Tyr-NH2(9) and D-Cys(6) substitutions in (B) are also well tolerated. The anti-V2 pA2 values of peptides 1-5 and 7-14 are as follows (1) 7.77 +/- 0.03; (2) 7.41 +/- 0.05; (3) 6.86 +/- 0.02; (4) 5.66 +/- 0.09; (5) approximately 5.2; (7) 7.25 +/- 0.08; (8) 6.82 +/- 0.06; (9) 7.58 +/- 0.05; (10) 7.61 +/- 0.08; (11) 7.59 +/- 0.07; (12) 7.20 +/- 0.05; (13) 7.57 +/- 0.1; (14) 7.52 +/- 0.06. All analogues antagonize the vasopressor responses to AVP, with anti-V1a pA2 values ranging from 5.62 to 7.64, and the in vitro responses to OT, with anti-OT pA2 values ranging from 5.79 to 7.94. With an anti-V2 potency of 7.77 +/- 0.03, the Pro3 analogue of (A) is surprisingly equipotent with (A), (anti-V2 pA2 = 7.81 +/- 0.07). These findings clearly indicate that position three in AVP V2/V1a antagonists, in contrast to position three in AVP agonists, is much more amenable to structural modification than had heretofore been anticipated. Furthermore, the surprising retention of V2 antagonism exhibited by the Pro3, Oic3, Tyr3, Trp3 and Hphe3 analogues of (A), together with the excellent retention of V2 antagonism by the Tyr-NH2(9) and D-Cys6 analogues of (B) are promising new leads to the design of potent and possibly orally active V2 antagonists for use as pharmacological tools and/or as radioiodinatable ligands and for development as potential therapeutic agents for the treatment of the hyponatremia caused by the syndrome of the inappropriate secretion of the antidiuretic hormone (SIADH).

Generation of anti-idiotypic monoclonal antibodies recognizing vasopressin receptors in cultured cells and kidney sections

To produce anti-idiotypic antibodies against receptors for the neurohypophyseal hormone vasopressin, an anti-vasopressin monoclonal antibody with a ligand specificity similar to that of vasopressin receptors was employed for immunization. Three anti-idiotypic monoclonal antibodies were obtained which induced, like vasopressin, plasminogen activator production in the renal epithelial cell line LLC-PK1 (expressing V2-receptors). Induction of plasminogen activator synthesis by the anti-idiotypic antibodies could be inhibited by coincubation with a vasopressin antagonist. In a fashion similar to that of vasopressin itself, the anti-idiotypic antibodies induced receptor down-regulation. The anti-idiotypic antibodies were employed to visualize vasopressin receptors on LLC-PK1 and A7r5 (V1-receptor-expressing) smooth muscle cells by immunofluorescence. Antibody-mediated fluorescence was not observed in receptor-deficient mutant cell lines or vasopressin-receptor-down-regulated cells. Furthermore, these antibodies were used for immunohistochemical localization of vasopressin receptors in rat and bovine kidney preparations. In accordance with earlier physiological and biochemical observations, vasopressin receptors were detected predominantly in collecting ducts in cortex and medulla. On the cellular level, a differential staining pattern was observed.

Vasopressin V2-receptor mobile fraction and ligand-dependent adenylate cyclase activity are directly correlated in LLC-PK1 renal epithelial cells

The role of hormone receptor lateral mobility in signal transduction was studied using a cellular system in which the receptor mobile fraction could be reversibly modulated to largely varying extents. The G-protein-coupled vasopressin V2-type receptor was labeled in LLC-PK1 renal epithelial cells using a fluorescent analogue of vasopressin, and receptor lateral mobility measured using fluorescence microphotolysis (fluorescence photobleaching recovery). The receptor mobile fraction (f) was approximately 0.9 at 37 degrees C and less than 0.1 at 10 degrees C, in accordance with previous studies. When cells were incubated for 1 h at 4 degrees C without hormone, and then warmed up to 37 degrees C and labeled with the vasopressin analogue, f increased from approximately 0.4 to 0.8 over approximately 1 h. The apparent lateral diffusion coefficient was not markedly affected by temperature pretreatment. Studies with radiolabeled vasopressin indicated that temperature pretreatment influenced neither receptor number nor binding/internalization kinetics. F-actin staining revealed that temperature change resulted in reversible changes of cytoskeletal structure. The maximal rate of in vivo cAMP production at 37 degrees C in response to vasopressin, but not to forskolin (receptor-independent agonist), was also markedly influenced by preincubation of cells at 4 degrees C, thus paralleling the effects of temperature preincubation on f. A linear correlation between f and maximal cAMP production was observed, suggesting that the receptor mobile fraction is a key parameter in hormone signal transduction in vivo. We conclude that mobile receptors are required to activate G-proteins, and discuss the implications of this for signal transduction mechanisms.

Internalization of fluorescent vasotocin-receptor agonist and antagonist in the toad bladder

This study compares hydrosmotic action, receptor binding, and fluorescent uptake of an agonist, d9phe(flu)AVT, and an antagonist, d4lys(flu)AVT, in the toad bladder. D9phe(flu)AVT increased osmotic water flow across the bladder with a 50% effective dose of 2 nM, whereas d4lys(flu)AVT inhibited water flow with a 50% effective dose of 0.1 microM. D9phe(flu)aVT displaced 10 nM [3H]arginine vasopressin (AVP) from plasma membranes by 50% (IC50) with 10 nM, whereas d4lys(flu)AVT had an IC50 of 3 microM. The fluorescent agonist induced a persistent increase in membrane permeability to water after removal from the serosal bathing solution. This residual response was diminished by preincubation with an agonist (AVP), but not with an antagonist [d4lys(N3)AVT]. The agonist, d9phe(flu)AVT, was internalized into toad bladder epithelial cells, as seen by epifluorescence microscopy, and this uptake was blocked by d4lys(N3)AVT. The antagonist, by contrast, was not internalized but remained at the cell surface. After stimulation with forskolin, however, the fluorescent antagonist was also internalized. These experiments suggest that agonists, but not antagonists, form functional complexes with receptors that, via formation of cAMP, trigger not only an increase in membrane permeability to water but also facilitate the clearance of hormone from the cell surface by endocytic uptake.

Synthesis and biological activity of a biotinylated vasopressin analog

To study vasopressin receptor-mediated endocytosis using electronmicroscopy methods and to develop avidin affinity columns for receptor purification, we synthesized and tested the biological properties of a biotinylated vasopressin (VP) analog [1-(2-mercapto) propionic acid] 8-[lysine-N6-biotin] VP (B-MLVP). B-MLVP was prepared by coupling biotin to the epsilon amine of the lysine residue in [1-(2-mercapto) propionic acid] 8-(lysine) VP (MLVP). The structure of HPLC purified B-MLVP was confirmed by fast atom bombardment mass spectrometry. B-MLVP effectively competed for arginine vasopressin (AVP) binding sites in canine renal plasma membranes on the surface of LLC-PK1 kidney cells. Dissociation constants of 15 nM and 202 nM were calculated from the results of competition binding assays conducted with membranes and cells, respectively. B-MLVP stimulated adenylate cyclase activity and elevated cellular 3',5',cyclic-AMP (cAMP) content in a manner similar to AVP, indicating it is an agonist of VP action in renal tissue. These observations indicate that B-MLVP is an agonist of VP action and may be used to study renal VP receptors by employing avidin coupled to various reporter groups.

Synthesis and characterization of fluorescein- and rhodamine-labeled probes for vasotocin receptors

Fluorescent analogues of vasotocin, [1-(beta-mercaptopropionic acid), 9-(p-aminofluoresceinylphenylalanine)]arginine vasotocin [[MPA1, (p-NH2flu)Phe9]AVT] and [1-(beta'-mercaptopropionic acid), 9-(p-amino rhodaminylphenylalanine)]AVT [[MPA1, (p-NH2rhod)-Phe9]AVT], were synthesized by the solid-phase method. These compounds yielded half-maximal hydrosmotic responses (half-maximal values) in the toad urinary bladder at 2 X 10(-9) M. Epifluorescence microscopy showed punctate basal localization of analogues on the majority of bladder epithelial cells within 20 min. The cellular localization was prevented by vasotocin. These fluorescent analogues may serve as useful probes for vasotocin receptors in toad bladder and in tissues from other species that use vasotocin as their antidiuretic-pressor hormone.

Binding and internalization of a fluorescent vasopressin analogue by collecting duct cells

The kinetics of binding and internalization of a fluorescent vasopressin analogue [1-desamino-8-rhodamine lysine vasopressin (rhoda LVP)] by principal cells within the microperfused rabbit cortical collecting tubule were quantitatively assessed with computer-assisted video microscopy. At 25 degrees C, binding of rhoda LVP exhibited saturation kinetics with half-maximal binding at 2 nM and maximal binding at concentrations greater than 5 nM. Rhoda LVP binding could be prevented by the simultaneous addition of a 10-fold higher concentration of arginine vasopressin (AVP) or the V2-receptor agonist, 1-desamino-8-D-arginine vasopressin (desmopressin). No obvious internalization or rhoda LVP was detected at 25 degrees C, i.e., the rhoda LVP fluorescence remained localized to the basal pole of each principal cell for at least 100 min after rhoda LVP addition and could be largely reversed by the subsequent addition of AVP. Conversely, warming the cells to 38 degrees C after binding was initiated resulted in a rapid (less than 30 min) migration of the fluorescence into the cell interior and a loss of AVP-displaceable binding from the cell surface. These results document the utility of this noninvasive optical strategy for quantitatively monitoring hormone binding to intact collecting tubule cells and demonstrate that rhoda LVP binds reversibly and with high affinity to V2 receptors on principal cells in the collecting tubule. The internalization (and presumed inactivation) of the hormone-receptor complex at 38 degrees C may contribute to the desensitization of collecting tubule cells to vasopressin at physiological temperature.

Fluorescent, photoaffinity, and biotinyl analogs of oxytocin

This study reports the solid phase synthesis and biological activities of two oxytocin analogs, [1-desamino, 4-lysine,7-(L-3,4,-dehydroproline)]oxytocin and [1-desamino, 4-threonine,7-(L-3,4-dehydroproline),8-lysine]oxytocin, and several fluorescent, photoaffinity, or biotinylated derivatives of these analogs and of oxytocin. The activities (in IU/mg) of the lysine-containing parent compounds, respectively, were as follows: uterus (without Mg++) 4.8 and 54; uterus (with Mg++) 19 and 440; milk ejection 65 and 414. The above analogs were coupled through the chemically reactive epsilon-amino group of lysine in position 4 or 8 or, in the case of oxytocin, through the N-terminal amino group of fluoresceine, photoaffinity, or biotinyl ligands. Fluoresceine coupled in position 1 of oxytocin gave an analog of low to moderate uterine (3.8 without Mg+ and 1.9 with Mg++) and milk ejection (7.9) activities. Analogs with biotin or fluoresceine coupled to lysine in position 4 had moderate uterine (11 and 23 without Mg++; 38 and 11 with Mg++) and milk ejection (33 and 13) activities. Analogs with fluoresceine, photoaffinity, or biotinyl labels coupled to lysine in position 8 retained good uterine (106, 62, and 147 without Mg++; 79, 78, and 509 with Mg++) and milk ejection (101, 181, and 247) activities and represent potentially useful experimental tools for studying hormone-receptor interactions and for receptor localization and isolation.

Synthesis and biological activities of a photoaffinity probe for vasotocin and oxytocin receptors

The present study describes the synthesis and biological activities of the photoreactive vasotocin analog 1-deamino[8-lysine(N epsilon-4-azidobenzoyl)] vasotocin ([Mpa1, Lys(N epsilon-4-azidobenzoyl)8]vasotocin). The analog was obtained by introducing the photoreactive aryl azido group at the epsilon-amino group of Lys8 in [Mpa1, Lys8]-vasotocin, which was synthesized by the solid phase method. In the isolated toad urinary bladder the photoaffinity analog of vasotocin retained hydroosmotic activity in the absence of u.v.-light. After irradiation the osmotic water flow across the bladder wall increased. Moreover, the water permeability remained high during repeated periods of washout, suggesting that the analog formed covalent complexes with vasotocin receptors in the toad bladder. In the rat uterotonic assay the photoreactive vasotocin analog was without photoactivation a mild agonist. These studies suggest that the photoaffinity analog of vasotocin might be useful for the isolation of vasotocin receptors in low vertebrates and oxytocin receptors in mammals.

Cell specificity of vasopressin binding in renal collecting duct: computer-enhanced imaging of a fluorescent hormone analog

A noninvasive microscopic method was used to assess the cell specificity of vasopressin binding within the heterogeneous collecting duct. The binding of a fluorescent vasopressin analog (1-desamino-8-rhodamine-L-lysine vasopressin) to cells of the microperfused rabbit cortical collecting tubule was visualized and quantitated with image-intensified video microscopy and digital image processing. Binding to the basolateral membranes of a subpopulation of cells could be detected within 1-2 min of addition of the fluorescent analog (10 nM) to the peritubular bath. Binding could be prevented or reversed by the addition of a 10-fold excess of the native hormone, which indicates that the fluorescent analog binds specifically to vasopressin receptors. The time course of binding paralleled and slightly preceded hyperpolarization of the lumen-negative transepithelial voltage, an electrical response that is also elicited by the native hormone. Double-label experiments in which the intercalated cell population was stained with fluorescein-labeled peanut lectin revealed that binding of the vasopressin analog was localized to the remaining cell type, the principal cell. Our results support the following conclusions. First, the principal cell constitutes the primary target cell for vasopressin in the rabbit cortical collecting tubule, although the intercalated cell may possess a limited number of receptors at a density below the detection limit of this optical approach. Second, computer-enhanced video microscopy is a powerful, noninvasive method for assessing the kinetics and spatial pattern of hormone binding.

Covalent labeling of hydrosmotic toad bladder receptors with an antagonist of vasotocin

A photoreactive analogue of vasotocin, [1-desamino,4-lysine(azidobenzoyl),8-arginine]vasotocin (4-N3-AVT), has been examined in the isolated toad urinary bladder for biological activity and binding to hormonal receptors. Although 4-N3-AVT induced only a small increase in bladder permeability to water, it behaved as a potent inhibitor of hydrosmotic action of [8-arginine]vasotocin (AVT) and [8-arginine]vasopressin (AVP). The inhibitory action of 4-N3-AVT was readily reversed on removal of the analogue from the serosal bathing solution. On the other hand, when bladders were exposed to 4-N3-AVT in the presence of long wavelength UV light (365 nm), the inhibition by 4-N3-AVT was not reversed on washout of the analogue. The dose of vasopressin required for a half-maximal response (ED50 value) was increased from 5 X 10(-9) to 1.3 X 10(-7) M in bladders photolabeled with 4-N3-AVT and the maximal response capacity of the tissue (intrinsic activity) was reduced to 79% of nonphotolabeled controls. A crude membrane preparation derived from bladders photolabeled with 4-N3-AVT contained 72 fmol of specific binding sites for tritium-labeled vasopressin per milligram protein, whereas nonphotolabeled controls had 136 fmol of specific binding sites per milligram protein. These observations suggest that 4-N3-AVT forms a covalent bond with hydrosmotic receptors in the presence of UV light. This is the first antagonistic photoaffinity analogue observed in the toad bladder and it may serve as a useful tool for analyzing the cellular mechanism of action of antidiuretic hormone.

Vasopressin binding sites in toad bladder: studies with the photoaffinity analogue [Phe(p-N3)3]AVP

Toad bladders were exposed to [Phe(p-N3)3]AVP (N3-AVP), an analogue of vasopressin with a photoreactive p-azido group in position three, in the presence and absence of ultraviolet (UV) light. Bladders exposed to the analogue in the presence of UV light showed an increase in membrane permeability to water, which persisted in spite of repeated and prolonged washout of analogue. In contrast, the hydroosmotic response induced by the analogue in the absence of UV light was readily reversed on washout. Aliquots of a broken epithelial cell preparation, derived from bladders that had been exposed to the analogue in the presence of UV light, bound less tritium-labeled vasopressin ([3H]AVP) than control aliquots that had been exposed to the analogue in the absence of UV irradiation or irradiated in the absence of the analogue. Membrane preparations that had not been photolabeled had specific binding sites for [3H]AVP in excess of 1,800 fmol/mg protein without evidence of saturation at a [3H]AVP concentration of 250 nM. Conversely, photolabeled membranes were saturated at a [3H]AVP concentration of 100 nM. The present studies demonstrate that a high proportion of [3H]AVP binding sites can be covalently labeled with N3-AVP and that at least some of these N3-AVP-bound sites are functional in triggering an increase in membrane permeability to water.