The HCN Channel Blocker ZD7288 Induces Emesis in the Least Shrew (Cryptotis parva)

Subtypes (1–4) of the hyperpolarization-activated cyclic nucleotide-gated (HCN) channels are widely expressed in the central and peripheral nervous systems, as well as the cells of smooth muscles in many organs. They mainly serve to regulate cellular excitability in these tissues. The HCN channel blocker ZD7288 has been shown to reduce apomorphine-induced conditioned taste aversion on saccharin preference in rats suggesting potential antinausea/antiemetic effects. Currently, in the least shew model of emesis we find that ZD7288 induces vomiting in a dose-dependent manner, with maximal efficacies of 100% at 1 mg/kg (i.p.) and 83.3% at 10 µg (i.c.v.). HCN channel subtype (1–4) expression was assessed using immunohistochemistry in the least shrew brainstem dorsal vagal complex (DVC) containing the emetic nuclei (area postrema (AP), nucleus tractus solitarius and dorsal motor nucleus of the vagus). Highly enriched HCN1 and HCN4 subtypes are present in the AP. A 1 mg/kg (i.p.) dose of ZD7288 strongly evoked c-Fos expression and ERK1/2 phosphorylation in the shrew brainstem DVC, but not in the in the enteric nervous system in the jejunum, suggesting a central contribution to the evoked vomiting. The ZD7288-evoked c-Fos expression exclusively occurred in tryptophan hydroxylase 2-positive serotonin neurons of the dorsal vagal complex, indicating activation of serotonin neurons may contribute to ZD7288-induced vomiting. To reveal its mechanism(s) of emetic action, we evaluated the efficacy of diverse antiemetics against ZD7288-evoked vomiting including the antagonists/inhibitors of: ERK1/2 (U0126), L-type Ca2+ channel (nifedipine); store-operated Ca2+ entry (MRS 1845); T-type Ca2+ channel (Z944), IP3R (2-APB), RyR receptor (dantrolene); the serotoninergic type 3 receptor (palonosetron); neurokinin 1 receptor (netupitant), dopamine type 2 receptor (sulpride), and the transient receptor potential vanilloid 1 receptor agonist, resiniferatoxin. All tested antiemetics except sulpride attenuated ZD7288-evoked vomiting to varying degrees. In sum, ZD7288 has emetic potential mainly via central mechanisms, a process which involves Ca2+ signaling and several emetic receptors. HCN channel blockers have been reported to have emetic potential in the clinic since they are currently used/investigated as therapeutic candidates for cancer therapy related- or unrelated-heart failure, pain, and cognitive impairment.


INTRODUCTION
Defensive processes such as nausea and vomiting protect vomitcompetent species avoid ingestion of toxic substances. While the act of forceful expulsion of gastrointestinal content through the mouth is called vomiting (or emesis), nausea is a painless unpleasant sensation that one may soon vomit. Vomiting may be preceded by retching behavior, where the gastrointestinal content is forced into the esophagus, but without the vomitus being expelled (Navari and Aapro, 2016;Adel, 2017). The anatomical sites involved in emesis include the dorsal vagal complex (DVC) emetic nuclei contained in the brainstem [the area postrema (AP), nucleus tractus solitarius (NTS) and dorsal motor nucleus of the vagus (DMNX)], the enteric nervous system (ENS) and enterochromaffin cells of the gastrointestinal tract, vagal afferents which carry emetic input from the gastrointestinal tract to the brainstem NTS, as well as vagal efferents which project motor signals from the DMNX to the gastrointestinal tract (Darmani and Ray, 2009;Babic and Browning, 2014). The AP and the NTS have loose capillaries, which allow diffusion of some circulating chemicals into the brainstem (Borison, 1989).
Hyperpolarization-activated cyclic nucleotide-gated (HCN) channels are a class of voltage-gated ion-channels permeable to Na + and K + and constitutively open at voltages near the resting membrane potential (Benarroch, 2013;McGovern et al., 2014). They consist of four subtypes 1-4, and are expressed on excitable cells in the central and peripheral nervous systems. The cation current (I h ) mediated by HCN channels elicits membrane depolarization toward threshold for action potential generation, which plays a pivotal role in controlling neuronal excitability (Gill et al., 2004;Benarroch, 2013). In the brainstem AP, up to 60% of neurons express HCN channels (Shinpo et al., 2012). ZD7288 is considered as a specific HCN channels blocker that acts nonselectively among the four known HCN channels (Chaplan et al., 2003). ZD7288 can depress the excitability of the AP since it blocks HCN channel activation (Shinpo et al., 2012). Moreover, ZD7288 at very low doses (10 −3 -10 −4 mg/kg, i. p.) can suppress apomorphine-induced conditioned taste aversion to saccharin preference and the apomorphine-evoked c-Fos expression in the rat area postrema (Shinpo et al., 2012). Based on these findings, the authors had suggested antinausea/antiemesis potential of ZD7288. However, they also found ZD7288 itself (>1 mg/kg) was the unconditioned stimulus for conditioned taste aversion learning (Shinpo et al., 2012), implying ZD7288 could be proemetic in emesis-competent species. In the clinic ivabradine was the first approved HCN4 blocker to lower heart rate and is currently used for the therapy of stable angina pectoris (Postea and Biel, 2011) and treatment of cancer chemotherapy-evoked left ventricular dysfunction (Sarocchi et al., 2018). Among patients receiving ivabradine, 3.3% have reported nausea (Chaturvedi et al., 2013).

Animals
A colony of adult least shrews from the Western University of Health Sciences Animal Facilities were housed in groups of 5-10 on a 14:10 light:dark cycle, and were fed and watered ad libitum. The experimental shrews were 45-60 days old and each weighed 4-6 g. Animal experiments were conducted in accordance with the principles and procedures of the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All protocols were approved by the Institutional Animal Care and Use Committee of Western University of Health Sciences. All efforts were made to minimize animals suffering and to reduce the number of animals used in the experiments.
ZD7288, ivabradine hydrochloride and palonosetron were dissolved in water. Nifedipine, U0126, MRS 1845, Z944, dantrolene and 2-APB were dissolved in 25% DMSO in water. Netupitant and RTX were dissolved in a mixture of ethanol/ Tween 80/saline at a volume ratio of 1:1:18. Sulpride was dissolved in distilled water with a 10 µL volume of 1/3 concentrated HCl which was then back titrated to pH 5 by the addition of NaOH. All drugs were administered at a volume of 0.1 ml/10 g of body weight.

Behavioral Emesis Studies
No gender differences between male and female least shrews have been detected in our previous studies. Thus, both males and female shrews were used in the current study. Prior to experimentation, least shrews were brought from the animal facility, separated into individual cages, and were allowed to adapt to the experimental conditions for up to 2 h (h). Daily food was restricted 2 h prior to the start of the experiment, but shrews were given 3-4 mealworms each prior to injection of an emetogen to aid identifying wet vomits as reported previously (Darmani, 1998). For systemic dose-response emesis studies, different groups of shrews were injected with varying doses of ZD7288 (0, 0.05, 0.25, and 1 mg/kg, i. p., n 8 shrews per group). In addition, different groups of shrews (n 6 per group) were injected intracerebroventricularly (i.c.v.) with either vehicle, 2.5, 5 or 10 µg ZD7288. Based on our published stereotaxic atlas of the least shrew brain (Ray and Darmani, 2007), the i. c.v. injection procedure in the least shrew has already been fully described by our laboratory (Darmani et al., 1994;Ray et al., 2009b). Likewise, the emetic effect of a structurally different HCN channel blocker ivabradine (0, 5, and 10 mg/kg, i. p., n 6 per group) was also examined.

Immunohistochemistry and Image Analysis
Immunohistochemistry of the least shrew brainstem (20 µm) and jejunal (25 µm) sections was carried out as previously reported (Zhong et al., 2016;Zhong et al., 2018;Zhong et al., 2019;Zhong and Darmani, 2020). The jejunal segment of the least shrew small intestine was dissected in accord with Ray et al. (2009a). The experimenter acquiring and analyzing the images were blind to experimental conditions.

Co-staining of c-Fos and TPH2
Another set of brainstem sections from vehicle-and ZD7288treated shrews (n 3 shrews per group) were incubated overnight with primary antibodies (rabbit anti-TPH2 antibody, 1:300, ab111828, Abcam; mouse anti-cFos antibody, 1:1o00, ab208942, Abcam) and incubated with the Alexa Fluorconjugated secondary antibodies (1:500, Alexa Fluor 488 and 594, Invitrogen). Nuclei of cells were stained with DAPI. The stained sections were examined as described above. For each animal, the number of c-Fos expressing cells among TPH2 positive cells of the selected areas, the AP, both sides of NTS and DMNX, from 3 sections at 90-μm intervals were counted. The mean average value was used for statistical analysis.

Statistical Analysis
The frequencies of vomits were analyzed using the Kruskal-Wallis non-parametric one-way analysis of variance (ANOVA) followed by Dunnett's post hoc test and presented as the mean ± SEM. The percentage of shrews vomiting across treatment groups at different doses was compared using the chi-square test. Statistical significance for differences between two treatment groups was tested by unpaired t-test. When more than two treatment groups were compared, a one-way ANOVA was used followed by Dunnett's post hoc test to determine statistical significance between experimental groups and control. p < 0.5 was considered statistically significant.

Expression of Hyperpolarization-Activated Cyclic Nucleotide-Gated Channel Subtypes in the Least Shrew Brainstem Dorsal Vagal Complex
The brainstem DVC constitutes the primary emetic nuclei including the AP, NTS and DMNX. We determined the expression of the HCN channel subtypes (1-4) by immunostaining 20 μm brainstem sections. In the least shrew brainstem DVC area, the most highly expressed HCN channel subtypes were HCN1 ( Figure 1) and HCN4 ( Figure 2). Both subtypes are highly enriched in the AP as compared with the NTS and DMNX ( Figures 1A, 2A). In the AP, both subunits appear to be localized on the membrane and dendrite regions of neurons ( Figures 1D, 2C). HCN1 and HCN4 immuno-positive puncta scattered in the NTS ( Figures 1G, 2D) and the DMNX ( Figures  1J, 2E) at a moderate and low level, respectively. HCN1 and HCN4 were also found to exhibit high level of expression in the hypoglossal nuclei (XII), an area below the DMNX ( Figures 1M,  2F). In contrast with the expression pattern of HCN1 and HCN4, expression of HCN2 and HCN3 subtypes were barely visualized in these three nuclei ( Figures 1B, 2B).

ZD7288 Causes Emesis and Activates the Brainstem Emetic Dorsal Vagal Complex Nuclei
The dose-response emesis data for ZD7288 are depicted in Figure 3. Intraperitoneal administration of ZD7288 increased the frequency of vomiting in the least shrew in a dose-dependent profile (KW (3, 28) 27.42, p < 0.0001). Dunn's multiple comparisons post hoc test showed that ZD7288 significantly increased the vomit frequency at its 0.25 (p 0.0364) and 1 mg/kg doses (p < 0.0001) ( Figure 3A). In addition, the chi-square test indicated that the percentage of animals vomiting in response to ZD7288 also increased dose-dependently (χ 2 (3, 28) 25.3, p < 0.0001). All shrews vomited at its 1 mg/kg dose (p < 0.0001), and 87.5% of shrews vomited at its 0.25 mg/kg (p 0.0004) ( Figure 3B). Intracerebroventricular injection of ZD7288 at 0, 2.5, 5 and 10 µg also increased the frequency of emesis in the least shrew in a dose-dependent manner (KW (3, 20) 11.54, p 0.0091). Dunn's multiple comparisons post hoc test showed that ZD7288 significantly increased the vomit frequency at its 10 µg dose (p 0.0066) ( Figure 3C). In addition, the chi-square test indicated that the percentage of animals exhibiting emesis in response to the centrally administered ZD7288 also increased in a dose-dependent fashion (χ 2 (3, 20) 10.49, p 0.0148). Indeed, 83.3% shrews vomited at its 10 µg dose (p 0.0034), and 50% of shrews vomited at its 5 µg (p 0.0455) ( Figure 3D). Intraperitoneal injection of HCN channel blocker ivabradine at 0, 1, 5 and 10 mg/kg also increased the frequency of emesis in the least shrew in a dose-dependent manner, with all tested shrews vomited at 10 mg/kg ( Figures 3E,F).
c-Fos induction is a recognized tool for evaluation of neuronal activation following peripheral stimulation with an agonist (Bullitt, 1990). Thus, we performed immunohistochemistry to examine c-Fos expression evoked by systemic administration of ZD7288. Relative to the control group treated with vehicle of ZD7288, a 1 mg/kg (i.p.) fully FIGURE 1 | Expression of HCN channel subtypes 1 and 2. HCN channel immunostaining with rabbit anti-HCN1 and mouse anti-HCN2 primary antibodies followed by Alexa Fluor 594 donkey anti-rabbit and 488 donkey anti-mouse secondary antibodies were performed on coronal brainstem sections (20 μm) prepared from naïve least shrews (n 3 shrews). Nuclei were stained with DAPI in blue. (A-C) Representative images (20x) show expression of HCN1 but not HCN2 observed in the least shrew brainstem dorsal vagal complex (DVC) area including the three emetic nuclei, the area postrema (AP), the nucleus tractus solitarius (NTS) and the dorsal motor nucleus of the vagus (DMNX). Scale bar, 100 µm. (D-O) Representative images (60x) show differential expression of HCN1 observed in the area postrema (AP), the nucleus tractus solitarius (NTS), the dorsal motor nucleus of the vagus (DMNX) and the hypoglossal nuclei (XII). Scale bar, 50 µm.
Frontiers in Pharmacology | www.frontiersin.org April 2021 | Volume 12 | Article 647021 5 emetic dose of ZD7288 caused marked increases in c-Fos expression in the brainstem DVC throughout the three emetic nuclei, the AP, NTS and DMNX ( Figures 4A-D). The numbers of c-Fos-positive cells in each region of interest are delineated in Figure 5. In vehicle-treated shrews, the average values for Fos-positive cells were 6.6 ± 0.6, 36.4 ± 2.1, and 17.2 ± 2 in the AP, NTS, and DMNX, respectively. Following vomiting induced by ZD7288, the average numbers of Fos-positive cells were increased to 43 ± 6.2 in the AP (p 0.0011 vs. Control), 133.3 ± 14.9 in NTS (p 0.0007), and 60.1 ± 6.3 in DMNX (p 0.0006). c-Fos expression in the ENS located in the jejunum was also examined. The ENS demonstrated a low c-Fos staining signal ( Figures 4E,G), with the mean number of c-Fos positive cells was 3.556 ± 0.7286 for ZD7288-treated group and 2.111 ± 0.9876 for vehicle-treated group ( Figure 5). The statistical analysis showed no significant difference for jejunal c-Fos expression between control and ZD7288-treated groups (p 0.3045) ( Figure 5). A further double staining with c-Fos and NeuN antibodies confirmed the jejunal c-Fos expression induced by ZD7288 occurred in NeuN-positive neurons ( Figure 6). A subsequent examination of c-Fos expression in TPH2-positive serotonergic neurons by double staining brainstem sections with c-Fos and TPH2 antibodies established that activation of serotonergic neurons in the brainstem DVC, composed of the AP, NTS and DMNX, induced by ZD7288 ( Figure 7). As shown in Supplementary Figure S1, in vehicle-treated shrews, the average values for c-Fos expressing TPH2 positive cells were 5.2 ± 0.3, 37.6 ± 6.3, and 12.1 ± 1.4 in the AP, NTS, and DMNX, respectively. Following vomiting induced by ZD7288, the average numbers of c-Fos expression in TPH2-positive cells were increased to 52.6 ± 7.8 in the AP (p 0.0037 vs. Control), 133.1 ± 9.2 in NTS (p 0.001), and 61.9 ± 4.9 in DMNX (p 0.0006).

Emesis Induced by ZD7288 Involves ERK1/2 Phosphorylation
To determine the participation of ERK1/2 in ZD7288-induced emesis, immunohistochemistry was performed to examine ERK1/2 phosphorylation evoked by ZD7288 (1 mg/kg, i. p.). Representative images from tile scanning of stained sections show that 15 min post ZD7288 administration, a significant increase in phosphorylation of ERK1/2 occurred in the brainstem DVC ( Figures 8A,C). Single field images acquired under ×20 objective show ZD7288 evoked significant ERK1/2 phosphorylation (p 0.0024) in brainstem dorsal vagal complex ( Figures 8B,D,E).

Ca 2+ Channel Modulators at the Cell Membrane Reduce ZD7288-Induced Emesis
Antiemetic Effect of L-Type Ca 2+ Channel Inhibitor Dose-dependent broad-spectrum antiemetic efficacy of the LTCC inhibitor nifedipine has previously been demonstrated in the least FIGURE 3 | The dose-response emetic effect of the HCN blockers in the lease shrew. Different groups of least shrews were given varying doses of ZD7288 (i.p., n 8 shrews per group or i. c.v., n 6) or ivabradine hydrochloride (i.p., n 6 shrews per group), and were observed for the next 30 min (A, C, E) The frequency of emesis was analyzed with Kruskal-Wallis non-parametric one-way ANOVA followed by Dunnett's post hoc test and presented as mean ± SEM. (B, D, F) Percentage of shrews vomiting was analyzed with chi-square test and presented as mean. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 vs. 0 mg/kg.  ZD7288 administration ( Figure 10). Indeed, relative to the control group pretreated with vehicle of nifedipine, nifedipine (2.5 and 5 mg/kg, s. c.) induced substantial decreases in both the mean vomit frequency in response to ZD7288 (p 0.0104 and p 0.0006, respectively) ( Figure 10A), as well as the number of shrews vomiting (p 0.1818 and p 0.0034, respectively) ( Figure 10B). Moreover, nifedipine at 10 mg/kg completely blocked ZD7288-evoked vomiting (p 0.0001 for frequency and p 0.0005 for percentage).

Anti-Emetic Effect of the TRPV1 Receptor Agonist
The TRPV1 agonist, resiniferatoxin (RTX) at low nanomolar nonemetic doses has the capacity to completely block vomiting caused by diverse emetogens in least shrews . Thus, the antiemetic efficacy of RTX was next tested against ZD7288 (1 mg/kg, i. p.)-induced vomiting. Significant reductions in a dose-dependent fashion were observed in both the mean vomit frequency [KW (4,34) 24.81, p < 0.0001] ( Figure 10C) and percentage of shrews vomiting [χ 2 (4, 34) 23.33; p 0.0001] ( Figure 10D). A significant reduction in the mean vomit frequency occurred from 1.0 μg/kg dose, whereas significance in the percentage  of shrews vomiting began from its 0.5 μg/kg dose. Moreover, complete suppression of both emetic parameters occurred at 2.5 μg/kg dose of RTX (p 0.0005 and p < 0.0001, respectively) ( Figures 10C,D).

Antiemetic Effect of SOCE Inhibitor
Store-operated Ca 2+ entry (SOCE) is an important route by which Ca 2+ mobilization occurs (Mizuta et al., 2008). As shown in Figures 10E,F
Frontiers in Pharmacology | www.frontiersin.org April 2021 | Volume 12 | Article 647021 FIGURE 9 | Effect of ERK1/2 inhibitor U0126 on the HCN channel blocker ZD7288-induced emesis. Different groups of shrews were given vehicle or varying doses of the ERK1/2 inhibitor U0126 (i.p.) (n 7 shrews per group), 30 min prior to ZD7288 (1 mg/kg, i. p.) administration. Shrews were observed the next 30 min. (A) The frequency of emesis was analyzed with Kruskal-Wallis non-parametric one-way ANOVA followed by Dunnett's post hoc test and presented as mean ± SEM. (B) Percentage of shrews vomiting was analyzed with chi-square test and presented as mean. ***p < 0.001 vs. 0 mg/kg. failed to significantly reduce the percentage of shrews vomiting in response to ZD7288 [χ 2 (3, 28) 6.4, p 0.0937] ( Figure 11D).

DISCUSSION
For the first time this study demonstrates that the HCN channel blocker ZD7288 evokes robust emetic behavior in least shrews when administered either centrally or peripherally. The emetic process is accompanied by activation of serotoninergic neurons in the brainstem emetic loci and is sensitive to Ca 2+ channel modulators as well as other emetic receptor antagonists.

Multifaceted Functions of Hyperpolarization-Activated Cyclic Nucleotide-Gated Channels
HCN channels are expressed in the heart and are known to have a critical role in controlling cardiac pacemaker activity (Rivolta et al., 2020). Ivabradine is the first clinically approved drug to efficiently treat heart failure (Postea and Biel, 2011) and cancer chemotherapy-evoked left ventricular dysfunction (Sarocchi et al., 2018). In agreement with clinical findings (Chaturvedi et al., 2013), in this study administration of ivabradine caused dose-dependent vomiting in all tested least shrews at 10 mg/kg (i.p.). In the nervous system, HCN channels are widely expressed in central and peripheral neurons where they play important roles in regulating excitability and neuronal firing (McGovern et al., 2014). They are known to be involved in epilepsies and neuropathic pain disorders, both of which are characterized by enhanced neuronal firing patterns (Postea and Biel, 2011). In addition, peripheral neuropathy evoked by oxaliplatin, a platinum-based chemotherapeutic agent, has been linked to activation of HCN channels-mediated currents (I h ), and ivabradine can attenuate this effect (Chang et al., 2020). Thus, HCN channels are considered as promising targets for anticonvulsant and analgesic drug discovery (Postea and Biel, 2011). Indeed, the HCN channel blocker ZD7288 can inhibit hippocampal synaptic plasticity in rats (Zhang et al., 2016) and reduces the generation of hippocampal epileptic discharges in rabbits (Kitayama et al., 2003). Furthermore, systemic, or local administration of ZD7288 suppresses neuropathic pain behavior as well as abnormal spontaneous firing in injured nerve fibers in rats, supporting the accumulating evidence that increased HCN channel activity is responsible for neuropathic pain (Chaplan et al., 2003;Lee et al., 2005;Luo et al., 2007;Ding et al., 2018). Moreover, HCN channel inhibition with ZD7288 provides protection for neural stem cells during radiotherapy, suggesting new therapeutic strategies against neurocognitive damage caused by chemotherapy and radiotherapy in cancer patients (Johard et al., 2020). The area postrema lacks a complete blood-brain-barrier and some circulating substances can easily enter the brainstem (Borison, 1989;Miller and Leslie, 1994). HCN channel activation in the area postrema (Milligan et al., 2006) should be important for emetic neuronal activity since a reduction in HCN channel-mediated current is associated with decreased neuronal excitability (Kitayama et al., 2003;Zhang et al., 2016). Thus, antiemetic effects should be expected when neurons in the area postrema are exposed to HCN channel blockers such as ZD7288 (Funahashi et al., 2003) or ivabradine (Funahashi et al., 2004). However, our study demonstrates the emetic effects of these two structurally different HCN blockers when administered systemically. Although ZD7288 can inhibit T-type Ca 2+ channels activity (Felix et al., 2003), the possibility that ZD7288 may exert an emetic effect through blocking T-type Ca 2+ channel can be excluded in this study. Indeed, ZD7288 is shown to be 20 times more selective for HCN channel than for T-type Ca 2+ channels (Sánchez-Alonso et al., 2008), and the selective and potent T-type Ca 2+ channel blocker Z944 (Casillas-Espinosa et al., 2015) lacks emetic activity in least shrews when administered intraperitoneally (data not shown). The current study is an extension of a published study in a vomit incompetent species (conditioned taste aversion test in rats), which had suggested a possible antiemetic/antinausea effect of ZD7288 at low doses vs. its probable pro-emetic potential at doses larger than 1 mg/kg (Shinpo et al., 2012). However, since ivabradine can evoke emesis in vomit competent species such least shrews and in angina patients at low doses (Chaturvedi et al., 2013), our notion that these structurally different HCN channel blockers exert emetic activity via blockade of HCN channels is a better FIGURE 12 | Efficacy of receptor-selective antiemetics against the HCN channel blocker ZD7288-induced emesis. Different groups of least shrews were given an injection of either the corresponding vehicles (0 mg/kg), or varying doses of 5-HT 3 R antagonist palonosetron (s.c.) (n 6 shrews per group) (A and B), NK 1 R antagonist netupitant (i.p.) (n 6) (C and D), or the D 2/3 R antagonist sulpride (s.c.) (n 6) (E and F), 30 min prior to ZD7288 administration (1 mg/kg, i. p.). Emetic parameters were recorded for the next 30 min (A, C) The frequency of emesis was analyzed with Kruskal-Wallis non-parametric one-way ANOVA followed by Dunnett's post hoc test and presented as mean ± SEM. (E) The frequency of emesis was analyzed with Unpaired t-test and presented as mean ± SEM. (B, D, F) Percentage of shrews vomiting was analyzed with chi-square test and presented as mean. *p < 0.05, **p < 0.01, ***p < 0.001 vs. 0 mg/kg. probability (Figure 3). In fact, both upregulation and downregulation of HCN channels have been associated with increased neuronal excitability (Doan et al., 2004;Postea and Biel, 2011).

Translational Significance
In this preclinical setting our study demonstrates the potential side-effects of HCN channel inhibitors/blockers in the process of clinical drug discovery. HCN channels are overexpressed in inflammatory and neuropathic pain states (Ramírez et al., 2018). In fact, HCN blockers reduce neuronal excitability and pain perception in animal studies and HCN channels are considered as pharmacological targets for pain management in patients (Ramírez et al., 2018). The HCN channel blocker used in the present study, ZD7288, is a widely used pharmacological tool to study both the function of these channels as well as its analgesic effect in various rodent pain models (Herrmann et al., 2015;Ramírez et al., 2018). Additional studies also demonstrate pain-attenuating efficacy of other HCN blockers including ivabradine and compound 12 m (Ramírez et al., 2018;He et al., 2019). The current study demonstrates that not only systemic (i.p.) administration of ZD7288 is proemetic in a dose-dependent manner, but other HCN blockers such as ivabradine and its analog cilobradine which ate are structurally different from ZD7288 (Bucchi et al., 2013), also evoke vomiting in least shrews. Furthermore, HCN channels have been implicated in the pathogenesis of epilepsy and the efficacy of HCN blockers used as antiseizure drugs in animal studies have recently been reviewed (Kharouf et al., 2020a;Kharouf et al., 2020b). Thus, in the above examples it seems quite clear that the emetic potential of new HCN blockers should be investigated in vomit-competent species prior to further clinical development. In fact, such a case already exists in the clinical development of phosphodiesterase inhibitors (Peng et al., 2020).
Our study provides the first evidence for involvement of HCN channels in the mediation of emesis in a laboratory animal model of vomiting. In fact, HCN channel blockers ivabradine and cilobradine are equipotent on HCN4 and HCN1 channels, while ZD7288 is more selective for HCN1 over HCN4 (Novella Romanelli et al., 2016). In the current study, ZD7288 at 1 mg/kg dose was a fully effective emetogen in all tested shrews. Ivabradine and cilobradine (DK-AH 269) also induced vomiting in all tested least shrews, but only at the 10 mg/kg dose. Thus, our findings not only demonstrate that HCN1 may be involved in vomiting induced by ZD7288, but also shows efficacy differences in the capacity of HCN blockers in evoking emesis. Of further interest, there are several well-known proemetic drugs with additional HCN channel-blocking activities: (1) Nicotine is a cholinergic agonist with proemetic effects in animals when given subcutaneously (Beleslin et al., 1981;Ueno et al., 1987;Yamamoto et al., 2004;Parker et al., 2009;Horn et al., 2014). Moreover, neuronal excitability induced by nicotine has been shown to be mediated via binding and inhibiting the HCN channels (Griguoli et al., 2010;Kodirov et al., 2014).
(2) Loperamide, an opiate receptor agonist, commonly used in the treatment of diarrhea, reliably induces emesis in the ferret when given subcutaneously (Bhandari et al., 1992;Zaman et al., 2000). Moreover, loperamide can block HCN1 more potently than HCN4 (Lee et al., 2008). (3) The inhalation anesthesthetic isoflurane causes emesis in the musk shrews (Horn et al., 2014). Isoflurane has been shown to inhibit the activity of HCN1 subunit containing HCN channels and increased the cortical neuronal excitability (Chen et al., 2009). Other anesthetics such as lidocaine and its metabolite, monoethylglycinexylidide, possess inhibitory actions on HCN channels (Meng et al., 2011). Therefore, the HCN channel subunit 1 (HCN1) may potentially represent an important mediator in nausea and vomiting.

ZD7288 Evokes Vomiting Through Central Activation of Serotonin Neurons
The expression pattern of HCN subtypes (1-4) vary in the central nervous system and at different levels in the brainstem (Shah, 2014). Furthermore, HCN1 and HCN3 channels were suggested to be the predominant forms in area postrema neurons (Monteggia et al., 2000), which is inconsistent with our present immunohistochemical staining showing besides HCN1, strong HCN4 immunoreactivity in the AP. Moreover, our study more specifically demonstrates their differential distribution among the DVC emetic nuclei of the brainstem. Although it remains unknown whether ZD7288 can cross the blood-brain barrier, circulating ZD7288 following its i. p. injection can act centrally on the brainstem area postrema which lacks a bloodbrain barrier (Borison, 1989;Miller and Leslie, 1994). In fact, systemic administration of small doses of ZD7288 (e.g., 1 mg/kg) evokes vomiting in all tested shrews which is accompanied by significant c-Fos expression in the brainstem emetic nuclei, including the AP, NTS and DMNX. On the other hand, i. p.administered ZD7288 only caused minimal c-Fos expression in a small number of neurons of the jejunal ENS, suggesting central emetic sites may play the principal role in ZD7288-induced emesis. The high levels of HCN1 and HCN4 in the AP area also support the notion of a central emetic site of action of bloodborne ZD7288 acting directly on HCN channels of the AP area. Although neurons residing in the ENS modulate intestinal motility and secretion (Furness, 2000;Sasselli et al., 2012), it appears that the peripheral emetic loci play a limited role in ZD7288-induced vomiting.
This above unlike the case of our previous studies where pronounced immunoreactivity in both c-Fos and ERK1/2 phosphorylation were observed in the jejunal ENS following emesis evoked by the Akt inhibitor MK-2206 (10 mg/kg, i. p.), suggesting the involvement of peripheral emetic loci in MK-2206evoked emesis . The jejunal ENS also exhibits significant c-Fos expression (relative to vehicle injection) following emesis induced by chemotherapeutic agent cisplatin as well as receptor selective emetogens, such as the serotonin 5-HT 3 receptor agonist 2-Methyl-5-HT, dopamine D 2 R agonist quinpirole and neurokinin NK 1 R agonist GR73632 (Ray et al., 2009a). Therefore, in this study, the strong c-Fos and phospho-ERK1/2 upregulation observed in the brainstem DVC, but not in the ENS in response to systemically administered ZD7288 could be considered as an evidence for central emetic action of ZD7288. The CNS neurons containing TPH2, the rate-limiting enzyme in 5-HT synthesis, are regarded as serotonergic cells (Cai et al., 2014). In the present study, it is noteworthy that following ZD7288 administration at its fully effective emetic dosage, c-Fos expression was exclusively increased in TPH2-expressing serotoninergic neurons of the dorsal vagal complex, indicating that circulating ZD7288 directly activate serotonergic neurons in the dorsal vagal complex, which probably leads to 5-HT release which would subsequently stimulate 5-HT 3 receptors to induce vomiting.

TRPV1R Agonist Resiniferatoxin
It is interesting to note that HCN channels exhibit structural similarities to TRPV1R with different gated mechanisms (Gill et al., 2004). Capsazepine, a well-known inhibitor of TRPV1R, possess HCN1 blocking activity (Biel et al., 2009). Our unpublished data show capsazepine (5 mg/kg, i. p.) is proemetic in the least shrews. Our previous studies have shown that the selective and ultra-potent TRPV1R agonist RTX has pro and antiemetic effects in the least shrew model of emesis (Darmani et al., 2014a;Darmani et al., 2020). A subcutaneous injection of RTX by itself induces vomiting in the least shrew at doses higher than 10 μg/kg. Lower (0.01-5.0 μg/kg) doses of RTX can suppress vomiting induced by diverse emetogens . Here we demonstrate that RTX at 2.5 μg/kg completely abolishes vomiting caused by ZD7288, which further supports the potent broad-spectrum antiemetic potential of RTX, which also has been demonstrated by several other groups Yamakuni et al., 2002;Darmani et al., 2014a;Rudd et al., 2015;Aghazadeh Tabrizi et al., 2017). The TRPV1R is expressed on neuronal membrane as well as in the membrane of intracellular organelles and plays a crucial role in maintaining intracellular Ca 2+ homeostasis (Lang et al., 2015;Hurt et al., 2016;Zhao and Tsang, 2017). Mechanisms of antiemetic effects of RTX are still unclear, but probably involves inhibition of voltage activated Ca 2+ channels  as well as desensitization of vagal afferents and its terminals in the NTS (Shiroshita et al., 1997).

SOCE Inhibitor MRS 1845
Store-operated Ca 2+ entry (SOCE) at the plasma membrane can be activated following Ca 2+ release from intracellular Ca 2+ stores through IP 3 R-and/or RyR-channels in both non-excitable and excitable cells (Putney and McKay, 1999;Putney, 2003;Parekh and Putney, 2005). In the current study, we found that relative to complete suppression of ZD7288-evoked vomiting by nifedipine at 10 mg/kg dose, pretreatment with MRS 1845, a potent and selective SOCE inhibitor, at the same dosage, significantly but partially reduced the frequency of the ZD7288-evoked vomiting without providing complete emesis protection in any of tested shrew (p > 0.05). These findings implicate a significant contribution of LTCC-mediated extracellular Ca 2+ influx and a minor role for SOCE in ZD7288-induced vomiting. MRS 1845 at the same dosage has also been shown to in part reduce the frequency of GR73632-evoked vomiting without providing complete emesis protection (p > 0.05) (Zhong et al., 2019). It also failed to exert a significant impact on thapsigargin-evoked vomiting (Zhong et al., 2016), which further supports a minor role for SOCE in vomiting.

T-type Ca 2+ Channel Inhibitor
Like the SOCE inhibitor MRS 1845, the potent and selective T-type Ca 2+ channel blocker Z944 at the same dose range, significantly attenuated the frequency of ZD7288-evoked Frontiers in Pharmacology | www.frontiersin.org April 2021 | Volume 12 | Article 647021 vomiting without providing complete emesis protection (p > 0.05), suggesting a minor role for T-type Ca 2+ channel in mechanisms underlying ZD7288-induced vomiting. Presynaptic HCN channels and T-type Ca 2+ channels are present on the same neuronal terminals and play a role in neurotransmitter release (Gill et al., 2004). Presynaptic HCN channels pharmacological blockade or genetic ablation of HCN1 channels, can lead to membrane hyperpolarization and enhanced Ca 2+ influx through T-type Ca 2+ channels, boosting spontaneous synaptic release (Huang et al., 2011;Shah, 2014). Therefore, it is possible that the minor antiemetic effect of T-type Ca 2+ channel blocker Z944 is due to limiting presynaptic neurotransmitter release evoked by the HCN channel blocker ZD7288 (Klar et al., 2003;Tokay et al., 2009). This hypothesis needs further investigation.

Intracellular Ca 2+ Channel Inhibitors
We have previously shown that intracellular sarcoplasmic/ endoplasmic reticulum luminal Ca 2+ release channels RyRs and IP 3 Rs can be differentially modulated by diverse emetogens including 2-Methyl-5-HT (Zhong et al., 2014), thapsigargin (Zhong et al., 2016) and the NK 1 receptor agonist GR73632 (Zhong et al., 2019). Here, we also demonstrate that significant reductions (60-85%) in the frequency of ZD7288-induced vomiting (but without full emesis protection) occur when shrews are pretreated with inhibitors of either RyRs (dantrolene at 5-10 mg/kg) or IP 3 Rs (2-APB at 10 mg/kg). Moreover, while a mixture of dantrolene (1 mg/kg) and 2-APB (1 mg/kg) did offer a slightly additional protection beyond what was afforded when each drug was administered alone, a mixture of the latter doses of 2-APB plus dantrolene combined with a 1 mg/kg dose of nifedipine, led to a further reduction in ZD7288-evoked vomiting (Data not shown). However, none of these formulations provided complete protection against ZD7288-induced vomiting (Data not shown). On the contrary, we have previously shown that combined pretreatment with low doses of nifidipine, dantrolene and 2-APB completely protects shrews from thapsigargin-evoked emesis (Zhong et al., 2016). Our present study provides in vivo evidence that the HCN blocker ZD7288 causes vomiting, a process that is partially dependent upon Ca 2+ mobilization which may involve Ca 2+ intracellular store release through IP 3 Rs and RyRs, as well as extracellular Ca 2+ entry via plasma membrane Ca 2+ channels (such as LTCC, SOCE, T-type Ca 2+ channel).

Effects of Diverse Emetic Receptor Antagonists
Substantial evidence indicates that cancer cytotoxic chemotherapeutic drugs including cisplatin cause vomiting via indirect stimulation of serotonin 5-HT 3 R, neurokinin NK 1 R and dopamine D 2 R, subsequent to release of corresponding monoamines in both the brainstem and gastrointestinal tract emetic loci Darmani and Ray, 2009). In the current study, we investigated the possibility of involvement of 5-HT 3 R, NK 1 R and D 2 R in vomiting evoked by the HCN channel blocker ZD7288. Indeed, the selective 5-HT 3 receptor antagonist palonosetron at 0.5 mg/kg significantly attenuated the vomit frequency evoked by ZD7288 by 50% but failed to completely protect shrews from vomiting up to a 5 mg/kg dose. Moreover, at the same dose-range, palonosetron has been shown not to affect vomiting in least shrews evoked either by the LTCC activator FPL64176 (Darmani et al., 2014b) or the intracellular Ca 2+ releaser, thapsigargin (Zhong et al., 2016). In fact, 5-HT 3 antagonists are considered as narrow-spectrum antiemetics (Sanger and Andrews, 2018). Netupitant and other neurokinin NK 1 R antagonists possess broad-spectrum antiemetic efficacy against several emetic challenges including cisplatin, apomorphine, morphine, ipecacuanha, copper sulfate and motion-induced emesis (Rudd and Andrews, 2004;Darmani et al., 2015;Rudd et al., 2016). In the present study, netupitant (10 mg/kg, i. p.) exerted 85% reduction in the frequency of ZD7288-induced emesis, but without providing complete emesis protection against the evoked vomiting. On the other hand, at this dose netupitant has been shown to provide a significant and near complete protection against vomiting evoked by thapsigargin (Zhong et al., 2016) and complete protection against vomiting evoked the NK 1 R agonist GR73632 (5 mg/kg, i. p.) in least shrews (Zhong et al., 2019). In comparison, the D 2/3 R antagonist sulpride at 8 mg/kg dose failed to affect ZD7288-induced emesis. This dose of sulpride can protect up to 80% of shrews from vomiting caused by D 2 R preferring agonist quinpirole, and its 2 mg/kg dose can fully protect shrews against vomiting caused by the nonselective dopamine receptor agonist apomorphine (Darmani et al., 1999).

CONCLUSION
Taken together, our findings demonstrate that the HCN channel blocker ZD7288, is a potent emetogen in least shrews. In fact, not only ZD7288 induces vomiting rapidly and in a potent manner, but except for nifedipine and resiniferatoxin, none of the other well-known tested antiemetics could provide total protection against the evoked vomiting. The induced emetic behavior was accompanied by a major central contribution as indicated by the ZD7288-evoked expression of c-Fos and ERK1/2 phosphorylation in the brainstem DVC emetic nuclei. Furthermore, ZD7288 induced c-Fos expression occurred in TPH2 positive neurons of the brainstem DVC emetic nuclei, suggesting activation of serotonin neurons may contribute to the evoked vomiting.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal study was reviewed and approved by the Western University's Protocol #R20IACUC018 -"Mechanisms of vomiting induced by chemotherapeutics, related emetics, and gastrointestinal disorders including cannabinoid-induced hyperemesis".