Comparison of the pharmacological profiles of arginine vasopressin and oxytocin analogs at marmoset, macaque, and human vasopressin 1a receptor

Abstract

Arginine vasopressin (AVP) and oxytocin (OT) are nonapeptides that bind to G-protein coupled receptors and influence social behaviors. Consensus mammalian AVP and OT (Leu⁸-OT) sequences are highly conserved. In marmosets, an amino acid change in the 8th position of the peptide (Pro⁸-OT) exhibits unique structural and functional properties. There is ~85% structural homology between the OT receptor (OTR) and vasopressin 1a receptor (V1aR) resulting in significant cross-reactivity between the ligands and receptors. Chinese hamster ovary (CHO) cells expressing marmoset (mV1aR), macaque (qV1aR), or human vasopressin receptor 1a (hV1aR) were used to assess AVP, Leu⁸-OT and Pro⁸-OT pharmacological profiles. To assess activation of Gq, functional assays were performed using Fluo-3 to measure ligand-induced Ca²⁺ mobilization. In all three V1aR-expressing cell lines, AVP was more potent than the OT ligands. To assess ligand-induced hyperpolarization, FLIPR Membrane Potential (FMP) assays were performed. In all three V1aR lines, AVP was more potent than the OT analogs. The distinctive U-shaped concentration-response curve displayed by AVP may reflect enhanced desensitization of the mV1aR and hV1aR, which is not observed with qV1aR. Evaluation of Ca²⁺-activated potassium (K⁺) channels using the inhibitors apamin, paxilline, and TRAM-34 demonstrated that both intermediate and large conductance Ca²⁺-activated K⁺ channels contributed to membrane hyperpolarization, with different pharmacological profiles identified for distinct ligand-receptor combinations. Taken together, these data suggest differences in ligand-receptor signaling that may underlie differences in social behavior. Integrative studies of behavior, genetics and ligand-receptor interaction will help elucidate the connection between receptor pharmacology and social behaviors.

Graphical Abstract

1. INTRODUCTION

AVP and OT are peptide ligands synthesized in the hypothalamus and secreted into the circulation affecting a host of peripheral physiological functions, including renal water absorption, vasoconstriction, parturition, and lactation. In the central nervous system, these neuropeptides are expressed in the ‘Social Brain Network’ and contribute to a variety of social behaviors [1-4] with AVP and OT generally exhibiting opposing behavioral roles and physiological functions [5-8]. These central AVP and OT projections are regulated by genetic, epigenetic, physiological, social, and environmental factors that affect mating preferences, parental care, decision making, social recognition, learning, and memory, and perturbations are associated with social behavioral deficits and psychopathologies [5, 9].

The OT-AVP neuropeptide family is highly evolutionarily conserved, with members or homologs identified in invertebrates and vertebrate taxa. AVP is completely conserved in mammals, and differs from the consensus mammalian OT sequence at only amino acids 3 and 8 [10-12]. In New World Monkeys (NWMs), six novel OT-analogs have been identified [11-15]. In mammals, the OT-AVP receptor family is comprised of three distinct vasopressin receptors (V1aR, V1bR, V2R) and one oxytocin receptor (OTR) [16]. In the vertebrate brain, V1aR is the most abundant and widely expressed vasopressin receptor [17] and differences in V1aR receptor genetic variability and expression patterns are associated with various social phenotypes [12, 18-21]. Across the phylogenetic tree, coevolution is observed for OT and AVP ligands and their receptors [11, 12, 22]. With ~85% structural homology between OTR and V1aR, significant cross-reactivity is observed between OT-AVP family ligands and receptors [12, 23-26]. AVP binds with higher affinity than oxytocin analogs at marmoset (mV1aR), macaque (qV1aR) and human (hV1aRs) vasopressin 1a receptors. Interestingly, OT analogs display 10-fold less binding affinity at the mV1aR than hV1aR and qV1aR, which may provide important taxon-specific insights into ligand-receptor signaling [27]. Due to high conservation in both ligands and receptors, the OT-AVP family is an excellent model system for studying the effects of ligand and receptor structure on G-protein coupled receptor (GPCR)-mediated signaling.

GPCRs bind to heterotrimeric G-proteins that are classified into four subfamilies: G_s (stimulation of adenylate cyclase), G_i/o (inhibition of adenylate cyclase and/or opening of potassium currents), G_q (activation of phospholipase C; calcium mobilization) and G₁₂ (activatior of Rho GTPases) [28]. As a GPCR, V1aR couples to G protein(s), activating signaling cascades that can result in Ca²⁺ mobilization and alteration of voltage and ligand-gated ion channels [6, 29-31]. In rat motorneurons, AVP can affect neuronal excitability by exerting a dual effect that suppresses K⁺ currents and enhances cation conductance [32]. Ligand binding can have agonistic or antagonistic effects on downstream signaling. At the hV1aR, OT analogs partially antagonize AVP binding, and are partial agonists at inducing Ca²⁺ influx [27].

In this report, we assess differences in GPCR signaling due to natural genetic variation in V1aRs associated with phenotypic consequences: marmoset (NWM, socially monogamous), macaque (OWM, nonmanogamous), and human (apes, socially monogamous) [11, 12, 21, 27]. We stably transfected CHO cells expressing mV1aR, qV1aR, and hV1aR and characterized AVP- and OT-analog G-protein signaling pathways by monitoring Ca²⁺ mobilization and membrane hyperpolarization.

2. MATERIALS AND METHODS

2.1. Chinese hamster ovary (CHO) cell cultures.

Transcriptome profiling confirmed that neuronal cell lines ND7/23, F-11 and SH-SY5Y express OT receptor [33]. Likewise, RT-PCR showed neuroblastoma cell lines SK-N-SH, SH-SY5Y, IMR-32 and astrocytoma cell line MOG-G-UVW all express OT receptor [34]. We excluded cell lines that express OT receptors from our studies inasmuch as endogenous expression would contribute to observed ligand-induced signaling. Therefore, we selected CHO cells as a common transfection background since they do not express OT receptors [35]. Wild-type Chinese hamster ovarian-K1 (CHO-K1) cells were purchased from ATCC (CCL-61) and cultured in Ham’s F12 (Hyclone SH30026.01), 10% fetal bovine serum (FBS) (Atlanta Biologicals S11550), 1.5% HEPES 1M Solution (Hyclone SH30231.01), 1% Penicillin-Streptomycin (10,000 U/mL; Life Technologies 15140-163). CHO-K1 cells stably transfected with mV1aR, qV1aR and hV1aR as described [27, 36] and were received from Dr. Myron Toews lab (University of Nebraska Medical Center; UNMC). mV1aR, qV1aR and hV1aR expressing cells were cultured in Ham’s F12 (Hyclone SH30026.01), 10% FBS (Atlanta Biologicals S11550), 1.5% HEPES 1M Solution (Hyclone SH30231.01), 1% Penicillin-Streptomycin (10,000 U/mL; Life Technologies 15140-163) and 400 mg/mL G418 (RPI Corp. G64000-5.0). Cells were cultured at 37°C in 5% CO₂ and 90% humidity.

2.2. Drugs.

AVP, Leu⁸-OT, and Pro⁸-OT (Anaspec 58863), 1,2-Bis(2-aminophenoxy)ethane-N,N, N9,N9-tetraacetic acid tetrakis(acetoxymethyl ester) (BAPTA-AM) (A1076; Sigma-Aldrich), NS-1619 (Sigma-Aldrich N170), Paxilline (Sigma-Aldrich P2928), SKA-31 (Sigma-Aldrich S5573), thapsigargin (Sigma-Aldrich T9033), and TRAM-34 (Sigma-Aldrich T6700) were reconstituted in DMSO. Pertussis toxin (Sigma-Aldrich P7208) was reconstituted in ultrapure water with 5 mg/mL bovine serum albumin (Fisher Scientific BP1600-100). Apamin (Sigma-Aldrich A1289) was reconstituted in 0.05 M acetic acid.

2.3. Ca²⁺ Mobilization Assay.

The effects of AVP and OT analog addition on Ca²⁺ mobilization were examined using Fluo3-AM fluorescence (Molecular Probes F1241) monitored with a FLIPR2 plate reader (Molecular Devices, Sunnyvale, CA). FLIPR operates by illuminating the bottom of a 96-well microplate with an air-cooled laser and measuring the fluorescence emissions from cell-permeant dyes in all 96 wells simultaneously using a cooled CCD camera. This instrument is equipped with an automated 96-well pipettor, which can be programmed to deliver precise quantities of solutions simultaneously to all 96 culture wells from two separate 96-well source plates.

Cells were plated at 0.3 million cells/mL in 96-well plates (MidSci P9803) and cultured overnight in culture media at 37°C in 5% CO₂ and 95% humidity. On the day of assay, growth medium was aspirated and replaced with 100 μl dye-loading medium per well containing 4 μM Fluo-3 AM and 0.04% pluronic acid (Molecular Probes P3000MP) in Locke's buffer (8.6 mM HEPES, 5.6 mM KCl, 154 mM NaCl, 5.6 mM glucose, 1.0 mM MgCl₂, 2.3 mM CaCl₂, 0.5 mM probenecid; pH 7.4). Cells were incubated for 1 h at 37°C in 5% CO2 and 95% humidity and then washed four times in 180μl fresh Locke's buffer using an automated microplate washer (Bio-Tek Instruments Inc., VT). Baseline fluorescence was recorded for 60 s, prior to a 20 μl addition of various concentrations of Leu⁸-OT and Pro⁸-OT. Cells were excited at 488 nm and Ca²⁺-bound Fluo-3 emission was recorded at 538 nm at 2 s intervals for an additional 200 s.

To assess the role of [Ca²⁺]_i in OT-mediated mobilization of Ca²⁺, the sarco/endoplasmic reticulum calcium ATPase (SERCA) inhibitor thapsigargin was used to deplete [Ca²⁺]_i stores. Cells were incubated in 100 μl dye-loading medium per well containing 4μM Fluo-3 AM and 0.04% pluronic acid in Locke's buffer (8.6 mM HEPES, 5.6 mM KCl, 154 mM NaCl, 5.6 mM glucose, 1.0 mM MgCl₂, 2.3 mM CaCl₂, 0.5 mM probenecid; pH 7.4). Cells were incubated at 37°C in a 5% CO2 and 95% humidity for 60 min prior to washing four times in 180 μl Locke’s buffer and 10 μl addition of thapsigargin (1 μM final concentration) and incubated for an additional five minutes. [Ca²⁺]_i mobilization assays were performed as described above.

2.4. Membrane Potential Assay.

To assess changes in membrane potential the FLIPR Membrane Potential Assay (FMP blue; Molecular Probes F1241) was used. Cells were plated at 0.3 million cells/ml in 96-well plates (MidSci P9803) and cultured overnight in culture media at 37°C in 5% CO₂ and 95% humidity. Growth medium was removed and replaced with 190 μl per well of FMP Blue in Locke's buffer (8.6 mM HEPES, 5.6 mM KCl, 154 mM NaCl, 5.6 mM glucose, 1.0 mM MgCl₂, 2.3 mM CaCl₂ pH 7.4). Cells were incubated at 37°C in 5% CO₂ and 95% humidity for 45 min. Baseline fluorescence was recorded for 60 s, prior to a 10 μl addition of log concentrations of AVP, Leu⁸-OT and Pro⁸-OT. Cells were excited at 530 nm and emission was recorded at 565 nm at 2 s intervals for an additional 200 s.

To ensure the veracity of comparisons of EC₅₀ and maximum response achievable (E_MAX) values of AVP and OT variants, all compounds were evaluated in parallel on the same 96-well plate, with the same split of cells and with identical reagent solutions. This experimental design was used for all AVP and OT peptide comparisons throughout this study, using mV1aR, qV1aR, and hV1aR-expressing cells. Inasmuch as all assays were performed in the same CHO cell line, we can exclude differences in cellular context as a source of observed differences in peptide potency or efficacy.

To assess the role of G_i/o in the OT ligand-induced membrane hyperpolarization, cells were incubated overnight with pertussis toxin (PTX) to inactivate G_i/o [37]. Cells were plated at 125,000 cells/mL in 96-well plates. PTX (150 ng/ml) was added 24 hours after plating and incubated for an additional 24 hours. Membrane potential assay was performed as described above.

BAPTA-AM is an calcium chelator [38]. If changes in calcium are responsible for activation of the Ca²⁺-activated potassium channels, the response should be BAPTA-AM sensitive. Cells were incubated at 37°C in 5% CO2 and 95% humidity for 35 minutes prior to a 10 ml addition of BAPTA-AM. Cells were incubated for an additional 10 minutes after drug addition.

To assess the role of intracellular Ca²⁺ in AVP and OT ligand-induced changes in membrane potential, thapsigargin was used. Cells were incubated with FMP in Locke’s buffer at 37°C in a 5% CO2 and 95% humidity for 40 min prior to a 10 μl addition of thapsigargin. Cells were incubated for an additional 5 min after drug addition.

To assess potential OT ligand-induced membrane hyperpolarization through Ca²⁺-activated K⁺ channels, we tested three inhibitors targeting Ca²⁺-activated K⁺ channel subtypes. G_q-mediated activation of protein kinase-C (PKC) causes an increase in cytosolic Ca²⁺ [28] with attendant activation of Ca²⁺ activated K⁺ channels. Ca²⁺-activated K⁺ channels are separated into three subtypes of small (SK_Ca), intermediate (IK_Ca), and large conductance (BK_Ca) channels [39]. Apamin is a selective inhibitor of SK_Ca channels [40, 41]. TRAM-34 is a selective inhibitor of the IK channel, K_Ca3.1 [42, 43]. Paxilline is a selective inhibitor of the BK_Ca channel [44]. Cells were incubated with FMP in Locke’s buffer at 37°C in a 5% CO2 and 95% humidity for 35 min prior to a 10 μl addition of apamin, paxilline, TRAM-34, or paxilline + TRAM-34. Cells were incubated for an additional 10 min after drug addition. Membrane potential assays were performed as described above.

SKA-31 is an activator of IK_Ca channel K_Ca3.1 [45, 46]. If changes in [Ca²⁺]_i are responsible for the activation of K_Ca3.1, the response should be SKA-31 sensitive. Cells were incubated at 37°C in a 5% CO2 and 95% humidity for 35 min prior to a 10 μl addition of TRAM-34. Cells were incubated for an additional 10 min after TRAM-34 addition. Membrane potential assays were performed as described above, with the exception of stimulation with SKA-31 rather than OT analogs.

NS-1619 is a BK_Ca channel activator [47, 48]. If changes in [Ca²⁺]_i are responsible for activation of the BK_Ca, the response should be NS-1619 sensitive. Cells were incubated at 37°C in a 5% CO2 and 95% humidity for 35 min prior to a 10 μl addition of paxilline. Cells were incubated for an additional 10 min after paxilline addition. Membrane potential assays were performed as described above, with the exception of stimulation with NS-1619 rather than OT analogs.

2.5. Data Analysis.

All concentration-response data were analyzed and graphs generated using GraphPad Prism (San Diego, CA, U.S.A.) software. EC₅₀ and E_MAX values for AVP- and OT peptide-stimulated increases in Fluo-3 fluorescence or IC₅₀ and E_MAX values decreases in FMP Blue fluorescence were determined by nonlinear least-squares fitting of a logistic equation to the peptide concentration versus fluorescence area under the curve data. The 95% confidence intervals for all EC₅₀/IC₅₀ and E_MAX were used to assess differences in potency/efficacy. R² was used to assess goodness of fit. A one-way ANOVA was performed with Sidek’s multiple comparisons to determine statistical significance and the adjusted p-values reported.

3. RESULTS

3.1. Comparison of marmoset and macaque V1aR amino acid sequence to human V1aR

In contrast to the natural variation in OT analogs in NWM’s, the AVP amino acid sequence is completely conserved in NWMs, OWMs, and humans. However, variation in V1aRs in NWMs is implicated in social monogamy, and may also contribute to other social phenotypes [11, 12, 21]. Marmoset, macaque and human V1aR amino acid sequences were accessed from the National Center for Biotechnology Information (NCBI). Alignment using the NCBI basic local alignment search tool (BLAST) [49] indicates that human and macaque (Macaca mulatta) V1aR are 97% conserved; and human and marmoset (Callithrix jacchus) are 92% conserved. Physiochemical changes to marmoset and macaque V1aR sequences were classified as radical when the amino acid substitution differed by size, polarity and/or charge, and conservative if the substitution did not differ by these properties (Figure 1; Supplementary Table 1). Radical amino acid substitutions were primarily observed in the mV1aR N-terminus and extracellular regions that may play a role in ligand binding. In intracellular regions and the C-terminus radical substitutions may affect G protein coupling and downstream signaling. Radical changes were more numerous for mV1aR than for qV1aR. Amino acids 3, 245, 264, 319 and 369 in mV1aR and qV1aR both differed from hV1aR. In this manuscript, we assess how this natural variation V1aRs impacts AVP and OT-induced receptor activation and cellular signaling.

Figure 1.

Comparison of marmoset and macaque V1aR amino acid sequence to human V1aR. Identification of amino acid substitutions in marmoset (red), macaque (blue), or both (red and blue striped) V1aRs relative to human V1aR. Numbers represent the location of the amino acid substitution. Radical physiochemical substitutions that differ in size, polarity and/or charge are indicated by diamonds and amino acid substitutions that do not differ by these properties are conservative changes indicated by circles (Supplementary Table 1).

3.2. AVP and OT analogs induce G_q-mediated intracellular Ca²⁺ Mobilization

G_q activates the phospholipase Cβ (PLC) and inositol 3-phosphate (IP₃; PLC-IP₃) signaling pathway [50] resulting in calcium mobilization. To compare AVP and OT analog activation of V1aR-coupled G_q, we performed functional assays using the Ca²⁺ indicator dye Fluo3-AM. At the mV1aR and qV1aR, AVP was significantly more potent, with a subpicomolar EC₅₀, whereas Pro⁸- and Leu⁸-OT displayed EC_50s in the nanomolar range (Table 1; Figure 2A-B; Supplementary Figure 1A-F) There was no significant difference in efficacy between AVP and OT analogs at these receptors (E_MAX; Supplementary Figure 1A-F). At the hV1aR, AVP was 60-175X more potent than Pro⁸- and Leu⁸-OT and significantly more efficacious (E_MAX) (Table 1; Figure 2C, Supplementary Figure 1G-I). AVP, Leu⁸-OT, and Pro⁸-OT did not mobilize Ca²_i in non-transfected CHO-K1 cells [35](Pierce et al. 2020). These data demonstrate that the dose-dependent response observed in V1aR-expressing lines is attributable to V1aR expression.

Table 1.

Potency of AVP, Leu⁸-OT and Pro⁸-OT at inducing calcium mobilization in mV1aR, qV1aR and hV1aR-expressing CHO cells. Sigmoidal curves (Figure 2). Raw data (Supplementary Figure 1).

		Ligand
Line	Parameter	AVP	Leu8-OT	Pro8-OT
mV1aR	EC50	0.77 pM	7.23 nM	3.39 nM
	95% CI	0.25 to 2.25 pM	2.81 to 18.58 nM	0.81 to 14.18 nM
	R2	0.81	0.88	0.78
qV1aR	EC50	0.34 pM	54.98 nM	19.40 nM
	95% CI	0.11 to 1.02 pM	25.77 to 117.30 nM	10.80 to 34.87 nM
	R2	0.82	0.91	0.95
hV1aR	EC50	1.45 nM	258.4 nM	92.78 nM
	95% CI	0.28 to 7.42 nM	7.81 to 8556.00 nM	13.56 to 635.00 nM
	R2	0.68	0.55	0.62

Figure 2.

AVP, Leu8-OT and Pro8-OT induced calcium mobilization in mV1aR, qV1aR and hV1aR expressing CHO cells. AVP, Leu⁸-OT, and Pro⁸-OT concentration-response relationships (A) in mV1aR cells. AVP, Leu⁸-OT, and Pro⁸-OT concentration-response relationships (B) in qV1aR cells. AVP, Leu⁸-OT, and Pro⁸-OT concentration-response relationships (C) in hV1aR cells. N=3 experiments (3-4 replicates per dose per experiment).

Inhibition of the sarco/endoplasmic reticulum calcium ATPase (SERCA) pump results in depletion of intracellular Ca²⁺ stores, because SERCA is responsible for maintaining the Ca²⁺ gradient between the cytosol and endoplasmic reticulum [51, 52]. To confirm the role of intracellular Ca²⁺ stores in AVP- and OT-analog mediated Ca²⁺ influx, cells were pretreated with thapsigargin, a potent SERCA inhibitor. In control assays AVP and OT analogs produced concentration dependent increases in intracellular Ca²⁺ in all three V1aR cell lines, however pretreatment with thapsigargin blocked AVP and OT-analog induced calcium mobilization Ca²⁺ (Figure 3; Supplementary Figure 2), and thus demonstrated AVP and OT-mediated increases in cytosolic Ca²⁺ levels are dependent on intracellular Ca²⁺ stores.

Figure 3.

Effects of pretreatment with thapsigargin (Tg) on AVP, Leu⁸-OT and Pro⁸-OT induced changes in intracellular Ca²⁺ mobilization in mV1aR-, qV1aR-, and hV1aR-expressing CHO cells. In mV1aR-expressing cells, control and Tg-pretreated concentration response relationships with AVP (A), Leu⁸-OT (B), and Pro⁸-OT (C). In qV1aR-expressing cells, control and Tg-pretreated concentration response relationships with AVP (D), Leu⁸-OT (E), and Pro⁸-OT (F). In hV1aR-expressing cells, control and Tg-pretreated concentration response relationships with AVP (G), Leu⁸-OT (H), and Pro⁸-OT (I). N=3 experiments (five replicates per dose per experiment).

3.3. AVP and OT Analog-Induced Changes in Membrane Potential are Dependent on G_q Mediated Ca²⁺ Moblization

Functional assays using FMP Blue, a membrane potential-sensitive dye, were used to assess AVP and OT-induced membrane hyperpolarization [53, 54]. In all three V1aR expressing cell lines, AVP and both OT-analogs produced concentration-dependent decreases in fluorescence consistent with a hyperpolarizing response (Table 2; Figure 4; Supplementary Figure 3). In mV1aR and hV1aR expressing cells, AVP-induced membrane hyperpolarization displayed a distinctive U-shaped concentration response curve, which was not observed with qV1aR (Figure 4) or with OT-analogs at mV1aR-, qV1aR- and hV1aR-expressing cells. In mV1aR, qV1aR and hV1aR-expressing cell lines, there was no difference in potency between Leu⁸- and Pro⁸-OT-induced hyperpolarizing responses (Table 2; Figure 4; Supplementary Table 3). At the hV1aR, no significant difference in efficacy is observed between the three ligands (Supplementary Figure 3), suggesting the OT analogs function as full agonists in this signaling pathway. The absence of AVP or OT-analog induced hyperpolarizing response in untransfected CHO-K1 cells demonstrated the requirement for V1aR transfection in the observed hyperpolarizing responses [35](Pierce et al. 2020).

Table 2.

Potency of AVP, Leu⁸-OT and Pro⁸-OT at inducing membrane hyperpolarization in mV1aR, qV1aR and hV1aR CHO cells. Sigmoidal curves (Figure 4). Raw data (Supplementary Figure 4).

		Ligand
Line	Parameter	AVP	Leu8-OT	Pro8-OT
mV1aR	IC50_1	274.17 pM	61.88 pM	68.26 pM
	95% CI		10.01 to 382.20 pM	14.62 to 318.7 pM
	IC50_2	2.51 nM
	R2	0.64	0.65	0.73
qV1aR	IC50	0.56 pM	125.00 pM	203.00 pM
	95% CI	0.17 to 1.86 pM	50.30 to 310.60 pM	62.56 to 658.70 pM
	R2	0.62	0.84	0.78
hV1aR	IC50_1	0.70 nM	22.70 nM	12.26 nM
	95% CI		1.98 to 260.70 nM	1.57 to 95.89 nM
	IC50_2	1.70 nM
	R2	0.57	0.55	0.62

Figure 4.

AVP, Leu8-OT and Pro8-OT induced membrane hyperpolarization in mV1aR, qV1aR and hV1aR expressing CHO cells. AVP, Leu⁸-OT, and Pro⁸-OT concentration-response relationships in mV1aR cells(A), in qV1aR cells (B), and in hV1aR cells (C). N=3 experiments (3-4 replicates per dose per experiment).

To assess if G_i/o contributes to the AVP and OT-analog induced hyperpolarizing response, cells were pretreated with PTX, which catalyzes the ADP-ribosylation of the Gα_i/o subunit, preventing this subunit from interacting with a GPCR [55]. Pretreatment with PTX did not inhibit AVP or OT-analog induced hyperpolarizing responses in any of the V1aR-expressing cell lines (Figure 5; Supplementary Figure 4; Supplementary Table 2), suggesting G_i/o does not contribute to the hyperpolarizing response. As a positive control, we used a human kappa-opioid-receptor expressing CHO cell line, which couples to G_i, and demonstrated that its agonist DynorphinA 1-13-NH₂ produced a robust hyperpolarizing response that was abolished in cells pretreated with PTX (data not shown) [35].

Figure 5.

Lack of effect for pretreatment with PTX on AVP, Leu⁸-OT and Pro⁸-OT induced changes in membrane hyperpolarization in mV1aR-, qV1aR-, and hV1aR-expressing CHO cells. In mV1aR-expressing cells, control and PTX-pretreated concentration response relationships with AVP (A), Leu⁸-OT (B), and Pro⁸-OT (C). In qV1aR-expressing cells, control and PTX-pretreated concentration response relationships with AVP (D), Leu⁸-OT, (E), and Pro⁸-OT (F). In hV1aR-expressing cells, control and PTX-pretreated concentration response relationships with AVP (G), Leu⁸-OT (H), and Pro⁸-OT (I). N=3 experiments (five replicates per dose per experiment).

To directly assess the role of calcium in AVP and OT-mediated membrane hyperpolarization, cells were pretreated with the calcium chelator BAPTA-AM. In all three V1aR-expressing cell lines, BAPTA-AM exposure blocked AVP and OT-analog induced hyperpolarization. Interestingly, in BAPTA-AM treated V1aR-expressing cells, an AVP and OT-analog induced depolarization was observed (Supplementary Figure 5), indicating a possible dual modulation of ion channels [32]. To confirm the role of intracellular calcium stores in OT-mediated changes in membrane potential, we pretreated cells with the SERCA inhibitor thapsigargin. As expected, pretreatment with thapsigargin eliminated hyperpolarization produced by AVP and OT-analogs in V1aR-expressing cells. Notably, in thapsigargin pretreated V1aR-expressing cells, AVP and OT-analogs again produced a depolarization response (Supplementary Figure 6).

To explore the role of Ca²⁺-activated K⁺ channels in AVP and OT-analog induced membrane hyperpolarization, which involves activating the G_q/phosphoinositide-phospholipase C pathway, we used a pharmacological approach with inhibitors that discriminate between subtypes of Ca²⁺-activated K⁺ channels; small conductance (SK_Ca), intermediate conductance (IK_Ca), and large conductance (BK_Ca) channels [39]. Cells were pretreated with the SK_Ca-selective blocker apamin [41] to assess the role of SK_Ca channels in AVP and OT-analog induced membrane hyperpolarization. In mV1aR and qV1aR-expressing cells, apamin did not substantially affect (≤10%) AVP or OT-analog induced membrane hyperpolarization (Figure 6A-F; Supplementary Table 3; Supplementary Figures 7A-C and 8A-C). In hV1aR-expressing cells, apamin did not markedly affect AVP or Pro⁸-OT-induced hyperpolarizing response, however, did significantly reduce Leu⁸-OT-induced membrane hyperpolarization by ~50% (Figure 6G-I, Supplementary Table 3; Supplementary Figure 9A-C). V1aR cells were treated with the IK_Ca channel blocker TRAM-34, which specifically blocks K_Ca3.1 [42]. In all three V1aR-expressing cell lines, TRAM-34 produced the most effective inhibition of AVP and OT-analog induced membrane hyperpolarization (Figure 6; Supplementary Table 3; Supplementary Figures 7G-I, 8G-I, 9G-I). To confirm the contribution of K_Ca3.1 in AVP and OT-analog induced membrane hyperpolarization, we used K_Ca3.1 activator SKA-31. In all three V1aR-expressing cell lines, TRAM-34 inhibited SKA-31-induced membrane hyperpolarization in a dose-dependent manner (Supplementary Figure 10; Supplementary Table 4), supporting a role for K_Ca3.1 in the AVP and OT-induced membrane hyperpolarization in V1aR-expressing cells. V1aR-expressing cells were pretreated with the BK_Ca-blocker paxilline to assess the role of BK_Ca channels in AVP and OT-induced hyperpolarizing responses. In mV1aR and hV1aR-expressing cells, pretreatment with paxilline had little to no effect on AVP or OT-analog induced hyperpolarizing responses (Figure 6A-C, G-I; Supplementary Table 3; Supplementary Figures 7D-F, 9D-F), whereas in qV1aR-expressing cells, paxilline modestly inhibited AVP and OT-analog induced hyperpolarizing responses by ~15-30% (Figure 8D-F, Supplementary Table 3, Supplementary Figure 6D-F). To confirm the contribution of BK_Ca channels in AVP and OT-analog induced hyperpolarizing response, we used the BK_Ca activator NS-1619. In all three V1aR-expressing cell lines, paxilline inhibited NS-1619-induced membrane hyperpolarization in a dose-dependent manner (Supplementary Figure 11; Supplementary Table 4), supporting a role for the regulation of membrane potential by BK_Ca channels in qV1aR-expressing cells. To determine the combined contribution of BK_Ca and K_Ca3.1 channels in AVP and OT-induced changes in membrane potential, cells were pretreated with both paxilline and TRAM-34. In mV1aR and hV1aR expressing cells, the combined pretreatment of paxilline and TRAM-34 robustly inhibited AVP and OT-analog induced hyperpolarizing responses, which was primarily attributable to TRAM-34. However, in qV1aR-expressing cells, the combined pretreatment of paxilline and TRAM-34 demonstrated an additive effect of these inhibitors (Figure 6; Supplementary Table 3; Supplementary Figures 7J-L, 8J-L, 9J-L), suggesting that BK_Ca and K_Ca3.1 both contribute to the AVP and OT-analog hyperpolarizing response in qV1aR-expressing cells.

Figure 6.

Effects of pretreatment with Ca²⁺-activated K⁺ channel inhibitors on AVP and/or Leu⁸-OT or Pro⁸-OT induced changes in membrane potential in mV1aR-, qV1aR-, and hV1aR-expressing CHO cells. Apamin (asparagus), paxilline (light blue), TRAM-34 (maroon), and paxilline + TRAM-34 (lavender) inhibitor fluorescence was normalized to AVP (green; A, D, G), Leu⁸-OT (blue; B, E, H), or Pro⁸-OT-induced (red; C, F, I) membrane hyperpolarization in mV1aR- (A-C), qV1aR- (D-F), and hV1aR expressing cells. Area under the curve (negative peaks only) were assessed and a one-way Anova was performed with Sidek’s multiple comparisons to determine statistical significance. N=3 experiments for each inhibitor (10 replicates per dose per experiment). Adjusted P-values (Supplementary Table 2).

4. DISCUSSION

In this study, we compared the pharmacological profiles of AVP, Leu⁸-OT, and Pro⁸-OT at the marmoset, macaque and human V1aRs. Previous studies have demonstrated modest differences in primate OT ligand and receptor binding and signaling profiles [35, 36], as well as differences in AVP and OT binding properties and Ca²⁺ mobilization at these V1aRs [27]. The natural variation in V1aR across NWM, OWM, and humans provides a unique opportunity to assess signaling cascades in relation to genetic variation in receptors. Comparison of mV1aR, qV1aR and hV1aR amino acid sequences demonstrated that variation was most commonly observed in the N-terminus, which may affect ligand binding, and the C-terminus, which has implications in G-protein coupling. By comparing these receptors expressed in a common cell line (CHO cells), we exclude differences in cellular context for variations in cellular signaling. Effects on intracellular signaling cascades can result in changes in cellular function, as well as physiological and behavioral effects at the organismal level [56, 57].

Our results confirm that all three ligands activated G_q signaling in concentration-dependent increases in intracellular Ca²⁺ through all three V1aRs. Consistent with binding data, AVP was much more potent at mobilizing intracellular Ca²⁺ than either Leu⁸-OT and Pro⁸-OT at all three V1aRs. Notably, in hV1aR expressing cells, AVP was also significantly more efficacious than either OT analog [27], suggesting that receptor variation in hV1aR may contribute to partial agonism by OT ligands. In hV1aR, amino acids at positions 3, 245, 264, 319 and 369 differ from both mV1aR and qV1aR, any or all of which may contribute to this effect. Notably, differences in Leu⁸-OT and Pro⁸-OT ligand structure did not show appreciable differences in signaling profiles at each primate V1aR.

In all three V1aR expressing cell lines, AVP, Leu⁸-OT, and Pro⁸-OT induced a dose-dependent hyperpolarizing response. AVP was much more potent at inducing a hyperpolarizing response in all three cell lines than either Leu⁸-OT and Pro⁸-OT. Although efficacy of both OT analogs trended lower at the human V1aR, these differences did not reach statistical significance as previously observed with the Fluo3-AM calcium mobilization assays. Notably, AVP and OT-analog hyperpolarizing response in qV1aR cells did not return to baseline as quickly as mV1aR and hV1aR. Additionally, AVP-induced membrane hyperpolarization at mV1aR and hV1aR demonstrated a u-shaped dose response curve that wasn’t observed in qV1aR-expressing cells, which may be reflective of enhanced desensitization with higher doses at mV1aR and hV1aR. AVP and OT-analog induced hyperpolarizing responses were insensitive to pretreatment with the G_i/o inhibitor PTX in all three V1aR expressing cell lines, suggesting G_i/o coupling does not contribute to these responses. Pretreatment with BAPTA-AM blocked AVP and OT-analog induced hyperpolarizing responses, demonstrating the effect is dependent on Ca²⁺ mobilization. Moreover, pretreatment with thapsigargin blocked AVP and OT-analog induced hyperpolarizing responses in all three V1aR expressing cell lines, establishing that the response is dependent on intracellular Ca²⁺ mobilization.

The Gq/PI-PLC pathway is PTX insensitive and can contribute to hyperpolarizing responses through Ca²⁺ activated K⁺ channels, so we assessed this pathway by using blockers specific for SK_Ca, BK_Ca(K_Ca1.1), and IK_Ca (K_Ca3.1) channels. Apamin selectively blocks SK_Ca channels, and pretreatment with this inhibitor resulted in minimal inhibition (≤ 10%) of AVP and OT analogs in all three lines with the exception of Leu8-OT in hV1aR-expressing cells. These data demonstrate that SK_Ca channels provide minimal contribution to AVP and OT-analog induced membrane hyperpolarization. Paxilline selectively blocks BK_Ca channels, and pretreatment with this inhibitor demonstrated a modest reduction in q-V1aR-expressing cells with AVP and OT-analog induced membrane hyperpolarization, suggesting that BK_Ca channels contribute to this response. Paxilline also inhibited the hyperpolarizing response to BK_Ca channel opener NS-1619, further supporting a role for BK_Ca channels in the hyperpolarizing response. TRAM-34 selectively blocks the IK_Ca channel K_Ca3.1, and pretreatment with TRAM-34 demonstrated a modest inhibition of AVP and OT-analog induced hyperpolarizing responses in qV1aR-expressing cells, whereas it robustly inhibited these responses in mV1aR and hV1aR-expressing cells. TRAM-34 also inhibited the hyperpolarizing response to K_Ca3.1 channel activator SKA-31, supporting a role for K_Ca3.1 channel contribution to the hyperpolarizing response in all three V1aR expressing cell lines. In qV1aR-expressing cells pretreatment with paxilline + TRAM-34 demonstrated an additive effect, suggesting that BK_Ca and K_Ca3.1 channels contribute to the hyperpolarizing response. Notably, the qV1aR sequence is more similar to the hV1aR than the mV1aR, so a study to assess qV1aR-specific variation at amino acid P17L in the N-terminus or I414L in the C-terminus affected AVP and OT-analog induced signaling responses would be interesting to parse out whether one or both amino acid substitution contributes to these signaling differences.

We recognize the importance of cellular context in GPCR signaling and understand that signaling profiles in the heterologous expression system (CHO cells) that we selected likely differs from those of intact neurons. However, primary neuronal cultures expressing the V1a receptors from the species we are evaluating are not currently available. Established human neuroblastoma and glioma cell lines reportedly express the endogenous OT receptor [34], which would cross react with AVP and OT ligands. Thus, we believe that these initial in vitro experiments to define the signaling profiles of AVP and OT variants are a fundamental first step to understanding ligand-receptor function.

Identification of the effects of natural variation on V1aR-mediated cellular signaling is crucial to translating pharmacological profiles at the cellular level to sociobehavioral processes at the organismal level. Species-specific differences in selectivity and affinity, as well as in vitro and in vivo differences complicate drug development [26]. The findings OT analogs have partial antagonist binding properties and partial agonist signaling properties at the hV1aR [27] has important implications for analyzing behavioral outcomes and for therapeutic development. Characterization of additional AVP and OT agonists may provide insight into functionally selective elements of the V1aR and how that may contribute to species-specific responses in AVP and OT-mediated cellular signaling. Together, these factors likely contribute to a disparate outcomes between animal studies and human clinical trials. Although there have been substantial increases in peptide and non-peptide agonists and antagonists for OT-AVP family as research tools [8, 26, 58-62], therapeutic development has been slow [26].

The OT-AVP neuropeptide family mediates social behavior and physiological processes, with perturbations associated with social behavioral deficits [5]. One major challenge is connecting ligand-receptor activation and cellular signaling to physiological and behavioral changes at the organismal level [57]. The present results show that AVP displays functionally distinct responses from OT analogs at the primate V1aRs examined. Together, these data provide insights into species-specific differences in V1aR sequence, resulting in differences in AVP, Leu⁸-OT and Pro⁸-OT pharmacological profiles, which likely contribute to differences in social behavior. Integration these pharmacological profiles, as well as genetic, physiological, and behavioral data is essential for advancing AVP and OT mediated therapeutic development.

ABBREVIATIONS

AVP

arginine vasopressin

BAPTA-AM

1,2-Bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis (acetoxymethyl ester)

BSA

bovine serum albumin

Ca²⁺

calcium

CHO

Chinese hamster ovary

CI

confidence interval

CNS

central nervous system

EC₅₀

the half-maximal response

E_MAX

is the maximum response achievable

FMP

Flipr membrane potential

GPCR

G protein-coupled receptor

hV1aR

human vasopressin receptor 1a

IC₅₀

is the half-inhibitory response

K⁺

potassium

Leu⁸-OT

consensus mammalian oxytocin sequence

mV1aR

marmoset vasopressin receptor 1a

NWM

New World Monkeys

OT

oxytocin

OTR

oxytocin receptor

OWM

Old World Monkeys

Pro⁸-OT

oxytocin sequence with proline in 8^th position

PKC

protein kinase C

PTX

pertussis toxin

qV1aR

macaque vasopressin receptor 1a

V1aR

vasopressin receptor 1a