U50,488H was obtained from the NIDA Drug Supply Program (Research Triangle Institute, NC). Filter papers (grade no. 1) were purchased from Advantec MFS (Dublin, CA). A corticosterone ELISA kit (Cat# KGE009) was purchased from R&D Systems (Minneapolis, MN).

Adult male and female wildtype C57BL/6N mice and homozygous K4A mice on a C57BL/6N background (25–40 g, 5–8 months old) were used. Wildtype mice were purchased from Jackson Laboratory (Bar Harbor, ME). The K4A knockin line was customed generated by homologous recombination (Cyagen, Santa Clara, CA), bred in-house, and genotyped as we described previously [15]. K4A mice were cryopreserved at Jackson Laboratory [#415239, Oprk1tm2.2(K4A)Llch]. Mice were housed under standard conditions (22 °C, 50%–60% humidity, 12:12 h light/dark cycle) with food and water ad libitum.

Adult male CD-1 mice purchased from Jackson Laboratory were used at about 12 weeks old for comparison of effects of U50,488H and nalfurafine on serum corticosterone levels.

All the procedures were approved by the Temple University Institutional Animal Care and Use Committee (ACUP 4956).

Mice were habituated to handling and injection once a day for 3 days. Mice then received U50,488H (5 mg/kg, s.c.), nalfurafine (20 μg/kg, s.c.) or saline (both 10 μL/g body weight). One hour later, 50–100 μL blood was collected from the submandibular vein according to NIH guidelines for survival blood collection. ¹ One week later, mice received the alternate treatment in a counterbalanced design. Serum corticosterone was measured using a commercial ELISA kit following published procedures [20, 21]. Serum samples were processed according to the manufacturer’s instructions, and absorbance was measured at 450 nm with background subtraction at 550 nm. Concentrations were calculated from standard curves using Prism 10 (GraphPad Software).

Following ≥1 h habituation, mice received U50,488H (5 mg/kg, s.c.) or saline and were returned to home cages for 10 min. Mice were then placed on pre-weighed filter papers in plexiglass cylinders for 1 h. Urine output was determined by filter paper weight gain after removal of feces and normalized to body weight. Treatments were separated by a 72-h washout period.

Data are presented as mean ± S.E.M. and were analyzed using one-way or two-way ANOVA.

Two-way ANOVA revealed a significant main effect of treatment, but no effect of genotype or treatment × genotype interaction. Thus, U50,488H (5 mg/kg, s.c.) significantly increased corticosterone levels in male and female wildtype and K4A mice, with no genotype differences (Figures 1A,B).

Basal corticosterone levels following saline treatment were compared between sexes (Figure 1C). Two-way ANOVA showed a significant main effect of sex, but no effect of genotype or sex × genotype interaction. Therefore, female mice exhibited higher corticosterone levels than males. Analysis of U50,488H-induced net increases also revealed no significant effects of genotype or sex.

U50,488H-induced net increases in serum were analyzed (Figure 1D). Two-way ANOVA results showed no significant main effects of sex or genotype, indicating that there are no significant differences between males and females or between wildtype and K4A mice.

We compared effects of nalfurafine and U50,488H on serum corticosterone levels. We found that nalfurafine (20 μg/kg, s.c.) and U50,488H (5 mg/kg, s.c.) increased corticosterone to similar extents (Supplementary Figure S1). Nalfurafine was administered at 20 μg/kg (s.c.) as at this dose it produced maximal analgesic and anti-scratch effects (ED₅₀ values of 5.8 mg and 8.0 mg/kg, respectively) [16].

Two-way ANOVA revealed a significant main effect of treatment, but no effect of genotype or treatment × genotype interaction. Accordingly, U50,488H (5 mg/kg, s.c.) significantly increased urine output in both sexes regardless of genotype (Figures 2A,B).

Net increases in urine output were calculated by subtracting the mean saline value from each U50,488H-treated data point. Two-way ANOVA showed a significant main effect of sex, but no effect of genotype or sex × genotype interaction (Figure 2C). Thus, U50,488H increased urine output to a greater extent in males than females, independent of genotype.

We found that nalfurafine (20 μg/kg, s.c.) elevated serum corticosterone to similar extents as U50,488H (5 mg/kg, s.c.). We showed previously that treatment of mice with U50,488H (5 mg/kg, s.c.), but not with nalfurafine (20 μg/kg, s.c.), promotes KOR phosphorylation in the mouse brain and KOR internalization in the mouse ventral tegmental area [22], suggesting that nalfurafine does not activate the associated β-arrestins-mediated signaling. As both compounds produced similar elevation in serum corticosterone, KOR phosphorylation and internalization, proxies of β-arrestin-mediated signaling, are unlikely to be required. This is consistent with the observation that U50,488H increased serum corticosterone similarly in wildtype and K4A mice. It is noteworthy that in vitro studies on functional selectivity of nalfurafine have yielded different results. It was demonstrated to be G protein-biased [23–25], un-biased [16] and β-arrestin-biased [26], likely due to differences in assay conditions.

Our findings that KOR phosphorylation is not required for U50,488H-promoted increases in urine outputs is consistent with those of Inan, Lee [18] who showed that in the dose ranges producing effective anti-scratch effects U50,488H (1 mg, 3 mg, 10 mg/kg) and nalfurafine (5 μg, 10 μg, 20 μg/kg) increased urine output dose-dependently in rats, with similar maximal responses. As mentioned above, we have shown that in vivo U50,488H (5 mg/kg), but not nalfurafine (20 μg/kg), promotes KOR phosphorylation in the mouse brain and KOR internalization in the mouse ventral tegmental area [22]. Since both compounds produced effective diuresis, KOR phosphorylation and internalization, proxies of β-arrestin-mediated signaling, are unlikely to be involved.

In contrast to our findings, triazole 1.1 and triazole 187, demonstrated to be G protein-biased in vitro, produced diuresis with lower maximal effects than U50,488H [27], suggesting possible β-arrestin involvement. Differences in experimental approaches may account for this discrepancy.

As β-arrestins-mediated signaling is not involved, one may infer that G protein-mediated signaling should be required for U50,488H-induced diuresis and elevated serum corticosterone. However, testing the involvement of G proteins in vivo using pertussis toxin (PTX) or small molecule inhibitors presents serious methodological issues. Systemic PTX administration causes severe toxicity in animals, including vascular permeability changes [28], disruption of blood-brain barrier [29] and intense systemic endoplasmic reticulum stress [30]. These effects would confound experimental results. Most of small-molecule drugs, including G protein inhibitors, do not readily cross the blood-brain barrier (BBB) [31]. Treatment of mice with PTX or small molecule inhibitors via intracerebroventricular injections or microinjection into specific brain sites to bypass the BBB introduces invasive surgical manipulations that interfere with behavioral and endocrine baselines.

If KOR phosphorylation is involved in receptor desensitization, in theory, K4A mice should show higher responses to U50,488H compared with wildtype mice. However, that is not the case in U50,488H-induced increases in urine output and serum corticosterone. This is reminiscent of our previous finding that K4A mutations did not affect acute U50,488H-induced anti-pruritic and hypo-locomotor effects [15]. It may be because the acute exposure to U50,488H (5 mg/kg) is not sufficient to cause KOR desensitization. Rather, higher doses or repeated exposures may be required. We previously found that 5x exposures to 80 mg/kg, but not 5 mg/kg, U50,488H caused tolerance to the anti-scratch effect of U50,488H [15].

Previous studies have demonstrated that KOR-mediated analgesic and anti-pruritic effects are mediated by G protein signaling [for reviews, see [13, 32, 33]], based on findings from mutant mouse lines (GRK3^−/−, β-arrestin2−/−, p38α MAPK^−/−, K4A), distinct KOR agonists, and G protein-biased agonists. Our data indicate that KOR-induced increases in serum corticosterone and urine output are not mediated by β-arrestins-mediated pathway, and likely involved G proteins. In contrast, mechanisms underlying KOR-mediated side effects, especially CPA, remain unresolved. Studies using genetic and pharmacological approaches indicate distinct mechanisms for CPA, hypolocomotion, and motor incoordination. CPA involves GRK-, p38α mitogen-activated protein kinase (MAPK), mammalian target of rapamycin (mTOR), cannabinoid receptor 1 (CB1), and protein kinase C (PKC) activation, but not β-arrestin2. KOR phosphorylation is required for U50,488H-induced CPA in female but not male mice. Several G protein-biased KOR agonists produce antinociceptive and anti-scratch effects with reduced side effects, but non-identical side effect profiles, and some still induce CPA. PKC inhibition impairs novelty-induced hyperlocomotion and reduces motor incoordination, whereas mTOR inhibition, β-arrestin2 deletion, or KOR phosphorylation deficiency does not. Motor incoordination is reduced by PKC inhibition or β-arrestin2 deletion, but unchanged by mTOR inhibition.

That U50,488H increases serum corticosterone in mice is consistent with reports in mice, rats, and humans [1, 34–36]. Most previous studies were conducted in males; here we extend these findings to females. Basal corticosterone levels were higher in females, consistent with previous reports [37, 38]. This sex difference likely reflects reduced negative regulation of the hypothalamic-pituitary-adrenal axis [37, 38] and higher corticosteroid-binding globulin concentrations in females [39]. On the other hand, the net enhancements promoted by U50,488H showed no sex differences.

U50,488H and other KOR agonists increase urine output in rats, dogs, monkeys, and humans, and these effects are blocked by KOR antagonists [1, 34, 40–42][reviewed in [43]]. We extended these findings to mice and females, showing greater U50,488H-induced diuresis in males when normalized to body weight. In contrast, Craft, Ulibarri [19] reported similar responses in male and female rats. The reason for this discrepancy is unclear. KOR agonists have been explored as water diuretics for water retention associated with edema, elevated intracranial pressure, and liver cirrhosis [44–47]. Unlike traditional diuretics, they cause little or no loss of K⁺ and Na⁺ in urine. Potential sex differences in humans warrants further investigation.

For corticosterone measurements, blood samples were collected 1 h after U50,488H administration and urine output was measured from 10 to 70 min post-injection, based on prior studies of corticosterone, diuresis, and KOR phosphorylation. In rats, blood samples for corticosterone measurements were obtained 1 h after U50,488H administration [1, 18]. U50,488H-induced KOR phosphorylation peaks at 30–60 min [22].