CYX-5, a G-protein biassed MOP receptor agonist, DOP receptor antagonist and KOP receptor agonist, evokes constipation but not respiratory depression relative to morphine in rats

Background Strong opioid analgesics such as morphine alleviate moderate to severe acute nociceptive pain (e.g. post-surgical or post-trauma pain) as well as chronic cancer pain. However, they evoke many adverse effects and so there is an unmet need for opioid analgesics with improved tolerability. Recently, a prominent hypothesis has been that opioid-related adverse effects are mediated by β-arrestin2 recruitment at the µ-opioid (MOP) receptor and this stimulated research on discovery of G-protein biassed opioid analgesics. In other efforts, opioids with MOP agonist and δ-opioid (DOP) receptor antagonist profiles are promising for reducing side effects c.f. morphine. Herein, we report on the in vivo pharmacology of a novel opioid peptide (CYX-5) that is a G-protein biassed MOP receptor agonist, DOP receptor antagonist and kappa opioid (KOP) receptor agonist. Methods Male Sprague–Dawley received intracerebroventricular bolus doses of CYX-5 (3, 10, 20 nmol), morphine (100 nmol) or vehicle, and antinociception (tail flick) was assessed relative to constipation (charcoal meal and castor oil-induced diarrhoea tests) and respiratory depression (whole body plethysmography). Results CYX-5 evoked naloxone-sensitive, moderate antinociception, at the highest dose tested. Although CYX-5 did not inhibit gastrointestinal motility, it reduced stool output markedly in the castor oil-induced diarrhoea test. In contrast to morphine that evoked respiratory depression, CYX-5 increased tidal volume, thereby stimulating respiration. Conclusion Despite its lack of recruitment of β-arrestin2 at MOP, DOP and KOP receptors, CYX-5 evoked constipation, implicating a mechanism other than β-arrestin2 recruitment at MOP, DOP and KOP receptors, mediating constipation evoked by CYX-5 and potentially other opioid ligands.


Introduction
Historically, the pain-relieving and adverse effects evoked by strong opioid analgesics were thought to be mediated by the same intracellular mechanisms following µ-opioid (MOP) receptor activation [1][2][3][4]. This led to the belief that it would not be possible to separate the pain-relieving and adverse effects evoked by strong opioid analgesics. However, this notion was challenged by reports of increased antinociception and a reduction in some morphine-related adverse effects in β-arrestin2 knockout (KO) mice [5,6]. These findings in β-arrestin2 KO mice, led to development of the hypothesis that the antinociceptive effects of MOP receptor agonists are mediated by G-protein coupled intracellular signalling pathways whereas recruitment of β-arrestin2 at the MOP receptor transduces opioid-related adverse effects [5,6]. To test this hypothesis, a concerted effort in the past decade has been directed at discovery of novel G-protein biassed MOP receptor agonists including oliceridine (also called TRV130) and PZM21 [7]. Although intravenous oliceridine was approved by the US Food and Drug Administration in 2020 for use in hospitals when other treatment options are inadequate, it has a boxed warning on addiction and life-threatening respiratory depression [8]. This outcome has led to renewed research attention on other strategies for discovery of novel opioid analgesic ligands with superior adverse effect profiles relative to the 'gold standard' morphine.
Interestingly, co-administration of a MOP receptor agonist with a DOP receptor antagonist in rodent pain models, evoked antinociception with improved tolerability, less analgesic tolerance and reduced abuse liability compared with a MOP receptor agonist alone [16]. This suggested that ligands with MOP receptor agonist and DOP receptor antagonist profiles may retain analgesic efficacy with reduced propensity to evoke opioid-related adverse effects [14][15][16][17]. In our previous work, [18] we described the discovery of CYX-6 (also called compound 7) and CYX-5 (also called compound 9) as more stable analogues of the endogenous opioid peptide, endomorphin-2 (EM-2) in the context of a structure-activity-study involving replacement of one or more amino acids with unnatural amino acids. In brief, the tetrapeptide, CYX-6 (H-Dmt-Pro-Tmp-Tmp-NH 2 [18]; Dmt: 2,6-dimethyl-l-tyrosine; Tmp: 2,4,6-trimethyl-L-phenylalanine) was a G-protein biassed MOP receptor agonist, a DOP receptor antagonist and a KOP receptor antagonist in vitro [18]. In rats, single intracerebroventricular (icv) bolus doses of CYX-6 evoked potent, dose-dependent antinociception and it was devoid of constipation and respiratory depressant effects, in contrast to the constipating and respiratory depressant effects of icv bolus doses of the positive control, morphine [19].
Here we report on the antinociceptive, constipating and respiratory depressant effects evoked by single icv bolus doses of CYX-5 relative to morphine and vehicle, in male rats. The icv dosing route was chosen because direct administration of investigational agents into the cerebrospinal fluid of the lateral ventricle of the brain enables intrinsic efficacy and adverse effect profiles to be evaluated without the confounding effects of drug metabolism, due to the low capacity of the CNS to metabolise exogenous compounds [20]. Additionally, based on the hydrophilicity of CYX-5, it is unlikely to cross the blood-brain-barrier (BBB) and the icv dosing route bypasses BBB penetration. Furthermore, the icv dosing route avoids systemic pharmacokinetic effects that have the potential to confound data interpretation due to the likely poor ability of CYX-5 to cross the BBB and/ or the formation of the neuro-excitatory metabolite of the positive control, morphine, morphine-3-glucuronide, in the liver of rodents [21].

Dosing solutions
As CYX-5 is water soluble, icv dosing solutions were prepared in sterile water for injection as was morphine hydrochloride and the vehicle was sterile water for injection. In a preliminary study of the metabolic stability of CYX-5 using rat liver microsomes (unpublished), the half-life was 0.96 h which is much longer than that of the parent peptide, EM-2 (reviewed in Smith et al. [8]). The good metabolic stability of CYX-5 combined with the low capacity of the CNS to metabolise exogenous compounds [20], indicates that CYX-5 is suitably stable for icv dosing and data interpretation. The icv doses of CYX-5 were chosen to include the maximum tolerated dose (20 nmol) that was identified in a pilot study, and two lower doses (3 and 10 nmol).  [22] and with those of the Committee for Research and Ethical Issues of the International Association for the Study of Pain. Sprague-Dawley (SD) rats (male, albino) were from the Animal Resources Centre (Canning Vale, Western Australia, Australia) and they were housed in groups of two or three per cage with free access to food and water. The holding facility was maintained at a mean (± standard deviation) temperature of 23 (± 3) °C with relative humidity of 40-70% and a 12 h light/12 h dark cycle.

Animals
A brief description of the methods is given in the next several paragraphs and we have published the detailed methods elsewhere [19].
After acclimatisation, rats with the desired weight range (190-230 g), were anaesthetized to facilitate implantation of an icv guide cannula with a 9 mm cannula plug that remained in place except during injection of the test or control item. Rats were kept warm and monitored during post-surgical recovery and they were given a 5-7 days recovery period before icv dosing and experimentation.
Rats were allocated randomly to receive an icv bolus dose of CYX-5 at 3, 10 or 20 nmol, morphine at 100 nmol or vehicle. Dosing solutions were coded and randomized prior to administration for blinding of the investigator. Data were unblinded after completion of experimentation and data analyses. Experimentation was performed in the light phase (6:00 a.m. to 6:00 p.m.) by the same investigator (M.Z.I.).

Test item administration
Rats received a single icv injection (10 μl) of CYX-5, morphine or vehicle under a brief period of light anaesthesia. After completion of experimentation, rats were again lightly anaesthetized for administration of an icv injection of malachite green dye to verify correct cannula placement. Data exclusion criteria were: (i) incomplete injection of the test item because of backflow of the injected solution or (ii) an inadequate distribution of dye in the ventricles.

Warm-water tail flick test
The tail flick test was used to assess the extent and duration of antinociception evoked by bolus doses of icv CYX-5 (3, 10 and 20 nmol) relative to icv morphine (100 nmol) and vehicle [19,23]. To assess a role for opioid receptors in mediating antinociception, rats were dosed with naloxone (10 mg/kg; sc) 10 min prior icv dosing with CYX-5 at the highest dose tolerated (20 nmol).

Gastrointestinal (GI) motility assessed using the charcoal meal test
This test involved administration of a 1 mL oral dose of a 10% suspension of charcoal in 5% gum arabic in rats, to assess the effects of icv doses of CYX-5, morphine and vehicle on GI motility [19]. The charcoal meal was given 30 min after icv dosing of CYX-5, morphine or vehicle. One hour later, the rats were euthanised and the intestine was collected for measurement of the following: (i) the distance from the pylorus to the front edge of the charcoal meal; (ii) the distance between the pylorus and the distal edge of the ileum. The intestinal transit index was determined by dividing measurement (i) by measurement (ii) [19].

Opioid induced inhibition of diarrhoea induced by castor oil
After an overnight fast (16 h), the pre-dosing stool consistency of individual rats was recorded. This was followed by an icv dose of CYX-5, morphine or vehicle followed by an oral dose of castor oil (2.5 ml) at 5-min post-icv dosing by the blinded investigator. The weight and consistency of faeces were recorded pre-dose and hourly for 8 h post-dosing for individual animals. The scoring criteria for faeces consistency were as follows: 0 = normal faeces, 1 = well-shaped wet faeces, 2 = shapeless faeces, 3 = shapeless faeces with large amounts of liquid [19,23]. A score of ≥ 2 for both weight and consistency was defined as diarrhoea [19,23].

Respiratory function using whole body plethysmography
The effects of icv CYX-5, morphine and vehicle on respiratory parameters in awake, freely moving rats were measured using a Buxco small animal whole body plethysmography (WBP) system (Buxco Electronics, Troy, New York, USA) as described previously by our group [19,23]. After acclimatisation for 30-60 min, respiratory parameters were recorded for 45 min pre-dosing (three 15 min cycles) and for 2 h post-dosing (eight 15 min cycles). In each cycle, rats breathed normal air for 1 min, followed by a hypercapnic gas mixture (8% CO 2 , 21% O 2 and 71% N 2 ) for 5 min and then a 9 min recovery period breathing normal air [19,23]. Respiratory parameters recorded by the blinded tester (M.Z.I.) included minute ventilation (MV), respiratory frequency (RF) and tidal volume (Tv) using the BioSystem XA software [19,23]. The baseline was calculated as the mean difference in MV (ΔMV), RF (ΔRF) and Tv (ΔTv) during the pre-dosing exposure to normal air and the hypercapnic gas mixture [19,23]. Following icv dosing with CYX-5, morphine or vehicle in individual animals, ΔMV, ΔRF and ΔTv were calculated for each cycle of the 2 h postdosing period [19,23].

Statistical analyses
Data are presented as mean ± standard error of mean (SEM). One-way analysis of variance was used for analysing the following data: ΔLatency-AUC of warm water tail flick data

CYX-5 evokes moderate antinociception relative to morphine
For rats with a baseline tail flick latency in the range 2.0-3.5 s to a warm-water stimulus, single bolus doses of icv CYX-5 evoked moderate antinociception at the highest dose (20 nmol) tolerated ( Fig. 2A). By comparison, single icv bolus doses of the positive control, morphine (100 nmol), evoked pronounced antinociception characterised by a biphasic profile similar to previous work from our laboratory [23] (Fig. 2A). For icv bolus dose of CYX-5 at 3, 10 and 20 nmol, the mean (± SEM) peak tail flick latencies were 5.2 (± 0.6), 4.0 (± 0.3) and 5.5 (± 0.5) s respectively and the mean time to peak effect was 1 h post-dose ( Fig. 2A). By comparison, the first and second antinociceptive peak latencies for morphine were 8.9 (± 0.5) and 10.0 (± 0.0) s respectively and the corresponding mean times to peak effect were 0.75 and 1.5 h, respectively ( Fig. 2A). A preliminary experiment involving icv administration of CYX-5 at 30 nmol, resulted in pronounced neuro-excitatory side effects (including convulsions, barrel roll, wet-dog shakes, excessive alertness, etc.), so this and higher doses were not assessed. Clinical signs of neuroexcitation were mild, intermittent, and transient in ~ 50% of rats dosed with icv CYX-5 at 20 nmol. CYX-5 at 20 nmol also induced postural abnormalities including a flattened torso, splayed legs and impaired balance (ataxia) for up to 90 min.
Moreover, the % change in SIT relative to vehicle for the various CYX-5 doses were significantly different relative to that of morphine (100 nmol) (F 3, 26 = 34.62, p < 0.0001, one-way ANOVA) (Fig. 3B). The results demonstrate the minimal effect of CYX-5 on GI motility.
CYX-5 at the 3 and 10 nmol doses, demonstrated less inhibition of castor oil-induced stool output during the first 4 h of the experiment compared with the morphinetreated animals (Fig. 4C). However, both CYX-5 (20 nmol) and morphine (100 nmol) delayed the onset of diarrhoea and significantly reduced the amount of stool output (Fig. 4C, D). The cumulative stool weight difference between CYX-5 and vehicle was significantly different from that for morphine only for the 10 nmol dose of CYX-5 at 4 h (F 3, 32 = 3.57, p = 0.0247, one-way ANOVA) (Fig. 4D).
Additionally, the total amount of diarrhoeal stool compared with non-diarrhoeal stool and the period of time when the stool of varying consistency was discharged by rats following icv administration of CYX-5, gave a deeper insight. Figure 4E, F display the diarrhoeal and non-diarrhoeal stools discharged by 4 h and 8 h after castor oil gavage. After 4 h, CYX-5 at 3 nmol produced a cumulatively large amount of stool discharge with a consistency from normal to diarrhoeal in a manner similar to vehicle. Although the amount of diarrhoeal stool discharge was lower in rats given icv doses at 10 and 20 nmol after 4 h, the amount of non-diarhoeal stool was higher than that for the morphine-treated animals (Fig. 4E). Fig. 3 In rats, icv CYX-5 did not significantly alter intestinal motility in a manner similar to icv vehicle, and in contrast to icv morphine. A The intestinal transit of a charcoal meal given by gavage at 1 h prior to icv dosing in rats, did not differ significantly between animals given icv bolus doses of CYX-5 (3-20 nmol; n = 6/ dose) or vehicle (n = 13) (one-way ANOVA, ****p < 0.0001). As expected, the positive control, icv morphine (100 nmol; n = 12), significantly impaired intestinal transit relative to icv vehicle (one-way ANOVA, ****p < 0.0001). B Additionally, the % change in intestinal transit evoked by icv CYX-5 did not differ significantly from that evoked by icv saline in contrast to the significant reduction evoked by icv morphine in rats (one-way ANOVA, ****p < 0.0001). The icv morphine and icv vehicle data from Fig. 3 of Eur J Pharmacol 2020;871:172,918, have been re-used with permission. Note that the CYX-5 data were generated concurrently in a blinded manner [19] Fig. 4 Icv CYX-5 inhibited castor oil-induced diarrhoea (3 nmol, n = 7; 10 nmol, n = 6; 20 nmol, n = 6), as a measure of constipation, in a manner similar to icv morphine (n = 17). A % animals with diarrhoea after an oral dose of castor oil over an 8 h post-dosing period. B Area under the curve (AUC) of % rats with diarrhoea (% of vehicle control, one-way ANOVA, **p < 0.01, ***p < 0.001, ****p < 0.0001). C Cumulative stool output after oral administration of castor oil. D Cumulative stool weight in the first 4 h post-dose period was significantly reduced in rats given either icv CYX-5 or icv morphine, relative to that for rats given icv vehicle (one-way ANOVA, *p < 0.05, NS not significant). E, F Comparison of the nondiarrhoeal and diarrhoeal stool consistency observed during the first 4 h (E) and at 8 h (F) post-dosing. The icv morphine and icv vehicle data from Fig. 4, panels A-F, of Eur J Pharmacol 2020;871:172,918, have been re-used with permission. Note that the CYX-5 data were generated concurrently in a blinded manner [19] Together, these findings show that icv CYX-5 at 20 nmol evoked constipation without significantly inhibiting GI motility in contrast to the positive control, icv morphine at 100 nmol, that inhibited both GI motility and castor oilinduced diarrhoea.

Effect of CYX-5 on respiratory parameters in rats
When whole body plethysmography was used to assess the respiratory depressant effects of icv CYX-5 at the maximum tolerated dose (20 nmol) in awake, freely moving rats, there was a significant stimulatory effect on minute ventilation under hypercapnic conditions (F 2, 8, 15/120 = 27.18, 1.364 and 8.789, p < 0.0001, = 0.2656 and < 0.0001; two-way ANOVA with a Dunnett's post hoc multiple comparison test). The minute ventilation of the CYX-5 treated animals also differed significantly whilst breathing air at the 0.5 to 1. Icv CYX-5 stimulated respiratory function in rats. Icv CYX-5 at the highest tolerated dose (20 nmol (n = 6) did not depress minute ventilation or respiratory frequency significantly relative to vehicle (n = 6), for rats breathing either room air (A, E) or a hypercapnic gas mixture (B, F). By contrast, icv morphine at 100 nmol (n = 6) reduced minute ventilation at the time of peak effect (B). Morphine (100 nmol icv) but not CYX-5, reduced respiratory frequency significantly relative to icv vehicle for animals breathing the hypercapnic gas mixture (D). Tidal volume for rats given CYX-5 (20 nmol icv) was increased significantly relative to that for rats given icv vehicle and morphine irrespective of whether or not animals were breathing room air (E) or the hypercapnic gas mixture (F) (2-way ANOVA with a Dunnett's post hoc multiple comparison test; *p < 0.05. **p < 0.01). The icv morphine and icv vehicle data from Fig. 5, panels A-F, of Eur J Pharmacol 2020;871:172,918, have been re-used with permission. Note that the CYX-5 data were generated concurrently in a blinded manner [19] two-way ANOVA with a Dunnett's post hoc multiple comparison test) and hypercapnic gas exposure (Fig. 5A-D The AUC values of Δ tidal volume differed from that for the vehicle but it was not significant (Fig. 6). However, Δ minute ventilation AUC (F 2, 15 = 14.66, p = 0.0003), Δ respiratory frequency AUC (F 2, 15 = 4.054, p = 0.0391; one-way ANOVA) and Δ tidal volume AUC of CYX-5 (F 2, 15 = 9.834, p = 0.0019, one-way ANOVA) were all Fig. 6 Effects of icv CYX-5 on the extent and duration of the change (change in area under the curve, ΔAUC) in minute ventilation, respiratory frequency and tidal volume in rats breathing room air or a hypercapnic gas mixture (A-F). Icv CYX-5 at 20 nmol (n = 6) increased the Δ minute ventilation AUC but it did not affect the Δ respiratory frequency AUC under hypercapnic conditions relative to the corresponding parameters for vehicle (B, D). By contrast, icv morphine (n = 6), reduced the Δ minute ventilation AUC relative to that for icv vehicle (n = 6). Interestingly, CYX-5 increased the Δ tidal volume AUC for rats breathing either room air (E) or the hypercapnic gas mixture (F) whereas morphine reduced the Δ tidal volume AUC relative to that for vehicle, but these did not reach statistical significance. However, for CYX-5, Δ minute ventilation AUC, Δ respiratory frequency AUC and Δ tidal volume AUC of CYX-5 all differed significantly from that for morphine for rats breathing the hypercapnic gas mixture (B, D, F) (one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001). The icv morphine and icv vehicle data from Fig. 6, panels A-F, of Eur J Pharmacol 2020;871:172,918, have been re-used with permission. Note that the CYX-5 data were generated concurrently in a blinded manner [19] significantly different compared with morphine (p < 0.05) for rats exposed to the hypercapnic gas mixture (Fig. 5B, D, F).

Discussion
Herein, we show in rats that single icv bolus doses of the EM-2 analogue, CYX-5, a G-protein biassed MOP/ KOP receptor agonist and a DOP receptor antagonist [18] (Table 1), evoked moderate, naloxone-sensitive, antinociception at the highest dose tolerated (20 nmol) ( Table 1). In our previous work [19], we reported that another EM-2 analogue, CYX-6, that is a G-protein biassed MOP receptor agonist and a DOP/KOP receptor antagonist [18] (Table 1) also evoked antinociception albeit more potently than CYX-5 herein [19]. For both CYX-5 and CYX-6 [19], doses could not be further escalated due to prominent neuro-excitatory side effects evoked by icv doses larger than 20 nmol. These neuroexcitatory behaviours may be mediated, at least in part, by a mechanism in common with that of dynorphin A (2-17), a major metabolite of the endogenous KOP receptor agonist, dynorphin A (1-17) [24]. Another possibility is indirect activation of the N-methyl-d-Aspartate receptor in a manner similar to the neuroexcitatory effects of the major morphine metabolite, morphine-3-glucuronide that accounts for > 50% of each oral morphine dose [25].
Regarding constipation (Table 1), icv CYX-5 at 20 nmol did not inhibit GI motility, but it did attenuate castor oilinduced diarrhoea mimicking icv morphine (100 nmol) in this regard (Table 1). However, in our previous work [19], we reported that CYX-6 did not attuenuate castor oil-induced diarrhoea after icv dosing at 20 nmol in rats (Table 1) [19].A plausible explanation for these differences is that CYX-5 is a KOP receptor agonist (Table 1) whereas CYX-6 is a KOP receptor antagonist with both EM-2 analogues having MOP receptor agonist/DOP receptor antagonist activity (Table 1) [18]. In future work beyond the scope of that herein, MOP, DOP and KOP selective opioid antagonists and MOP, DOP and KOP receptor knockout mice will be used to gain additonal insight. In previous work by others, a MOP receptor agonist/DOP receptor antagonist had a lesser effect on GI transit and faecal output relative to that evoked by the peripherally selective MOP agonist, loperamide, following oral administration in mice [26]. In other work, sc 14-O-phenylpropyloxymorphone, a highly potent, mixed MOP/DOP/KOP receptor agonist, had a favourable GI transit profile relative to morphine in mice. Specifically, 14-O-phenylpropyloxymorphone was 2.4-fold less potent for inhibition of GI transit compared with antinociception whereas sc morphine was 1.8-fold more potent for GI transit inhibition than for antinociception [27]. In other work, biphalin, an octapeptide analogue of enkephalin and a mixed MOP/ DOP receptor agonist, at 5 mg/kg ip, decreased intestinal transit, colonic propulsion and delayed initiation of castor oil-induced diarrhoea [28]. Taken together, our findings and those of others highlight the potential importance of differential opioid receptor activity profiles (agonist/ antagonist) at each of MOP, DOP and KOP receptors with regard to the evoked constipation profile in rats.
As both CYX-5 and CYX-6 were G-protein biassed opioid ligands, the inhibitory effect of icv CYX-5 on stool hydration in the castor oil-induced diarrhoea test, was somewhat unexpected. Our findings on the inhibitory effects of CYX-5 on castor oil-induced diarrhoea suggest that β-arrestin2 does not have a major role in mediating this dimension of constipation, at least in rats. This notion is supported by previous work by others whereby the constipation Table 1 Comparison of the in vitro activity at MOP, DOP and KOP receptors and in vivo efficacy (antinociception, constipation and respiratory function) for icv CYX-5 relative to icv CYX-6 and icv morphine in Sprague-Dawley rats *In vitro inhibition of forskolin-stimulated cAMP formation using human embryonic kidney cells stably expressing MOP, DOP or KOP receptors **icv dosing route in male Sprague-Dawley rats # Assessed using whole body plethysmography under hypercapnic conditions in awake, freely moving rats a Cai et al. [18]; b Imam et al. [19]; c Kuo et al. [36]; d Kuo et al. [23] Opioid ligand In vitro agonist or antagonist activity* Antinociception** profiles of the biassed MOP receptor agonists, oliceridine (TRV130) and PZM21, in rodents [29,30] were less impressive than expected based upon extrapolation of earlier findings for morphine in MOP receptor-KO mice [31]. Specifically in mice, sc doses of the G-protein biassed MOP receptor agonist, oliceridine, evoked robust warm water tail flick antinociception that appeared to be resistant to tolerance with repeated administration [29]. However, in work by others, antinociceptive tolerance did develop in mice administered PZM21 at 40 mg/kg sc for 3 consecutive days [32]. Regarding constipation, oliceridine inhibited faecal output and colonic propulsion markedly, without development of tolerance to these effects after repeated administration [29], which is undesirable. PZM21, another biassed MOP receptor agonist that evoked promising antinociception (40 mg/kg sc), also reduced faecal output significantly as an indicator of constipation, at a dose (20 mg/kg sc) that was lower than the antinociceptive dose [30]. This latter effect is similar to morphine that was 1.8-fold more potent for GI transit inhibition than for evoking antinociception in mice [27].
Regarding impact on respiratory function herein, single icv doses of CYX-5 at the highest dose tolerated (20 nmol), increased minute ventilation in awake, freely moving rats exposed to 5-min epochs of a hypercapnic gas mixture in contrast to the respiratory depressant effects of the positive control, icv morphine (100 nmol) ( Table 1). This effect of icv CYX-5 in rats was underpinned by an increase in tidal volume but without a change in respiratory frequency. By comparison in our previous work, icv doses of CYX-6 at 20 nmol did not depress minute ventilation significantly under hypercapnic conditions in contrast to respiratory depression evoked by icv morphine (100 nmol) in awake freely moving rats (Table 1) [19]. Although CYX-6 like morphine, reduced respiratory frequency relative to vehicle, CYX-6 increased tidal volume relative to morphine and vehicle in these animals [19]. These differential effects of icv CYX-5 and CYX-6 on parameters of respiratory function in rats under hypercapnic conditions, despite both EM-2 analogues being G-protein biassed ligands at MOP, DOP and KOP receptors as well as MOP agonists and DOP antagonists, suggest that these differences may be underpinned by the KOP receptor agonist activity of CYX-5 in contrast to the KOP receptor antagonist activity of CYX-6 [18].
Close inspection of the effects of icv CYX-5 on respiratory parameters in rats breathing room air vis-à-vis 5 min epochs of a hypercapnic gas mixture (8% CO 2 in air), highlight the importance of the latter clinically relevant condition, to unmask the effects of CYX-5 and morphine respectively, on parameters of respiration (Figs. 5, 6). Additionally, comparison of the respiratory stimulant effects of the biassed MOP/KOP agonist-KOP antagonist, CYX-5, with those for the biassed MOP agonist, PZM21 (40 mg/kg sc), showed that although an initial report was that PZM21 did not induce respiratory depression in mice breathing room air [30], this was not the case in subsequent work by others [32]. Specifically, Hill and colleagues showed that PZM21 at 40 mg/kg sc evoked respiratory depression in mice breathing room air with peak effects observed at 10-15 min post-dosing. Additionally, when the mice were exposed to a hypercapnic gas mixture (5% CO 2 in air), PZM21 (10-40 mg/kg ip) evoked dose-dependent, naloxone-sensitive, respiratory depression characterized by reduced minute volume, but not respiratory rate, and these effects were sustained over a 60-min post-dosing period [32]. Although twice-daily injection of morphine (10 mg/kg) or PZM21 (40 mg/kg) for 5-consecutive days induced antinociceptive tolerance, tolerance did not develop to the respiratory depressant effects of PZM21 [32]. Factors potentially contributing to these between-study differences for PZM21 include differences in the dark-light cycle, differences in dosing route and between-vendor differences for mice used in these two studies [19,33]. Also, correlations between MOP receptor/β-arrestin2 recruitment and opioid-related adverse effects has been challenged by more recent data demanding a more critical assessment and interpretation of preclinical results [34][35][36]. In other work, the G-protein biassed MOP receptor agonist, oliceridine, produced less respiratory depression compared with an equiantinociceptive dose of morphine in mice [37]. However, based on the results of two Phase 3 clinical trials [38,39], FDA approval of intravenous oliceridine was contingent upon a black box warning on life-threatening respiratory depression [8]. Regarding the possibility that KOP receptor agonism may contribute to the respiratory stimulant effects of CYX-5, previous work by others found that KOP receptor agonists had variable responses on parameters of respiration [40]. For example, KOP receptor agonists decreased respiratory frequency and minute ventilation in pentobarbital-anesthetized rats [41]. They also decreased burst frequency in isolated rat brain-spinal cord neurons and had a minimal effect on breathing in conscious SD rats [42]. However, in other work, KOP receptor agonism in conscious turtles increased minute volume [40] and DOP receptor agonism had mixed excitatory-inhibitory actions [40]. This latter observation may be due to between-species differences as opioids such as morphine evoke CNS excitatory effects in some species and depressant effects in others [40,43].
In summary, icv bolus doses of CYX-5, a G-protein biassed MOP/KOP receptor agonist and a DOP receptor antagonist, evoked moderate antinociception at the highest dose tolerated (20 nmol) in rats. At the same dose, CYX-5 did not significantly impair GI motility in rats, but it inhibited castor oil-induced diarrhoea with the latter effect aligned with impaired GI motility evoked by the positive control, icv morphine. By comparison, we previously showed that icv dosing with the EM-2 analogue, CYX-6 at 20 nmol, did not inhibit either GI motility or castor oil-induced diarrhoea in rats (Table 1) [19]. This finding was somewhat unexpected as both CYX-5 and CYX-6 are G-protein biassed ligands at MOP, DOP and KOP receptors, and both are MOP receptor agonists and DOP receptor antagonists (Table 1). However, CYX-5 is a KOP receptor agonist whereas CYX-6 is a KOP receptor antagonist, and so this difference may explain their differential effects on castor oil-induced diarrhoea in rats (Table 1). Overall, our data suggest that for CYX-5, G-protein bias per se, does not abolish constipation, at least in rats. We also found differences between icv CYX-5 (20 nmol) and icv CYX-6 on parameters of respiratory function under hypercapnic conditions in rats. Specifically, icv CYX-5 (20 nmol), stimulated minute ventilation and tidal volume with a reduction in respiratory frequency in contrast to icv CYX-6 (20 nmol) that stimulated tidal volume with no changes in minute ventilation or respiratory frequency [19]. Consistent with expectations, the positive control, icv morphine at 100 nmol, evoked respiratory depression under hypercapnic conditions characterized by decreased minute ventilation, respiratory frequency and tidal volume, relative to animals dosed with icv vehicle (Fig. 5).
Author contributions MTS, AK and MZI designed the experiments. MZI performed all in vivo experiments. YC, YQ and TL synthesized CYX-5. MZI wrote the manuscript, analysed the data and prepared the graphs. All authors reviewed manuscript drafts and approved the final manuscript.
Funding Open Access funding enabled and organized by CAUL and its Member Institutions. Mohammad Zafar Imam was supported financially by a Research Training Program (RTP) Scholarship and a UQ Graduate School Scholarship. A.K., S.G. and this project were supported financially by an Australian Research Council Large Linkage grant in collaboration with Boehringer Ingelheim Pharma GmbH & Co. KG (#LP120200623). The authors acknowledge the Queensland Government Smart State Research Facilities Programme for supporting CIPDD research infrastructure. CIPDD is also supported by Therapeutic Innovation Australia (TIA). TIA is supported by the Australian Government through the National Collaborative Research Infrastructure Strategy (NCRIS) program. Synthesis of CYX-6 was supported financially in part by a grant from the National Natural Science Foundation of China (NSFC No. 81573280) to Tingyou Li.

Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Conflict of interest No relevant conflicts of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.