G-Protein Signaling Regulating Opioid-Induced Respiratory Rate Depression

G-protein Signaling Regulating Opioid-induced Respiratory Rate Depression

Jamil Danaf

A thesis submitted in conformity with the degree requirements for the degree of MSc Graduate Department of the Institute of Medical Science University of Toronto

G-protein Signaling Regulating Opioid-induced Respiratory Rate Depression

Jamil Danaf Master of Science Institute of Medical Science University of Toronto 2020

Abstract

Opioids are pain treatment mainstays, but present with potentially lethal respiratory depression. To develop safer opioid therapies, mechanisms underlying opioid-induced respiratory depression must be better understood. Opioids mainly bind to µ-opioid receptors (MORs), G- protein-coupled receptors that activate heterotrimeric G-proteins, including Gβγ. There is significant debate around whether G-proteins or β-arrestins, proteins involved in receptor internalization/desensitization, directly regulate respiratory circuit inhibition. While Gβγ activates

G-protein inwardly rectifying potassium (GIRK) channels, which are key mediators of respiratory rate depression, its direct role in opioid-induced respiratory depression is unknown. We investigated the contribution of Gβγ in opioid-induced respiratory rate depression and found Gβγ inhibition in respiratory circuits reversed opioid-induced respiratory rate depression, regulators of

G-protein signaling 4 (RGS4) are co-expressed with MORs in respiratory circuits, and RGS4 inhibition in vivo potentiates opioid-induced respiratory rate depression. Our data suggest G- protein signaling is key in opioid-induced respiratory depression and is not limited to β-arrestin recruitment. Acknowledgements

I would first like to thank my supervisor, Dr. Gaspard Montandon. The guidance and support you provided me over these past few years have been invaluable, and I am grateful to have had the opportunity to work with and learn from you during this time. I would also like to thank my committee members, Dr. Richard Horner and Dr. Beverley Orser, for their support and input throughout my thesis. Thank you to all my lab mates for their constant help and positivity, and for making the lab such an enjoyable and welcoming environment. Lastly, I would like to thank my family and friends, as their endless love, encouragement, and support have made completing my thesis possible.

Contributions I would like to thank Dr. Gaspard Montandon, Dr. Carolina da Silveira Scarpellini, and Hattie Liu for teaching and aided me in conducting in vivo experiments. Wendy Wang and Dr. da Silveira Scarpellini also helped run in situ hybridization assays. Dr. Caterina Di-Ciano Oliveira scanned all in situ hybridization images. Dr. Xiaofeng Lu trained me in various staining and tissue processing techniques.

Table of Contents G-protein Signaling Regulating Opioid-induced Respiratory Rate Depression ...... ii Abstract ...... ii Acknowledgements ...... iii Contributions...... iii Rationale ...... 1 Ventilation...... 2 Respiratory Rhythm Modulation ...... 6 Opioid Effects on Respiration ...... 8 Neural Mechanisms of Opioid-Induced Respiratory Depression ...... 8 PreBötzinger Complex ...... 9 Medullary Networks ...... 11 Pontine Networks ...... 12 Cortical/Subcortical Networks ...... 12 Neural Circuits of Nociception ...... 13 Opioid Analgesia ...... 14 Opioid Receptors and Opioid Pharmacology ...... 16 G-Protein Coupled Receptors ...... 17 G-Proteins ...... 19 Adenylyl Cyclase ...... 20 Calcium Channels ...... 20 GIRK Channels ...... 21 RGS Proteins ...... 22 Research Question ...... 24 Central Hypothesis ...... 24 Aims and Objectives ...... 24 Hypotheses ...... 25 Chapter 2: Methods ...... 26 Anesthetized Experiments in Rats ...... 26 Microperfusion of Drugs into Discrete Brainstem Regions ...... 28 Determination of Probe Site Location ...... 28 Data Analysis ...... 30 Correlation Maps ...... 31

In situ Hybridization ...... 32 Statistical Analysis ...... 33 Chapter 3: Results...... 35 Gallein Administration in the preBötC Reverses Respiratory Rate Depression by MOR Agonists ...... 35 Medullary region sensitive to DAMGO in gallein experiments ...... 40 RGS4 Expression in the preBötC ...... 42 RGS Inhibition Partially Potentiates Respiratory Rate Depression by Opioid Agonists ...... 44 Chapter 4: Discussion ...... 52 Role of Gβγ Subunit ...... 53 Role of RGS4 in Opioid-induced Respiratory Rate Depression ...... 55 Limitations ...... 58 Conclusion ...... 60 Future Studies ...... 61 References ...... 65

List of Abbreviations

βarr2 β-arrestin 2 ACC Anterior Cingulate Cortex aCSF Artificial Cerebrospinal Fluid AP Anterior-posterior ANOVA Analysis of Variance BötC Bötzinger Complex cAMP Cyclic Adenosine Monophosphate CCG 50014 4-[(4-Fluorophenyl) methyl]-2-(4-methylphenyl)-1,2,4-thiadiazolidine-3,5-dione CNS Central Nervous System

CO2 Carbon Dioxide

CTAP D-Phe-Cys-Tyr-D-Trp-Arg-Thr-Pen-Thr-NH2 cVRG Caudal ventral respiratory group Cy3 Cyanine dye 3 Cy5 Cyanine dye 5 DAMGO (D-Ala2-NMe-Phe4-, Gly-ol5)-enkaphalin dapB 4-hydroxy-tetrahydrodipicolinate reductase DAPI 4’6-diamidine-2’-phenylindole dihydrochloride Dbx1 Developing Brain Homeobox 1 ƪDia Diaphragm activity, rectified and time averaged DMSO Dimethyl Sulfoxide DRG Dorsal Root Ganglion DV Dorsal-ventral GABA γ-aminobutyric acid GAP GTPase Activating Protein GDP Guanosine Diphosphate ƪGG Genioglossus activity, rectified and time averaged GIRK G Protein-Coupled Inwardly Rectifying Potassium Gly183Ser Glycine to serine mutation at the amino acid 183 position GPCR G-Protein Coupled Receptor

GRN Gigantocellular Reticular Nucleus GTP Guanosine Triphosphate IO Inferior Olive KF Kölliker-Fuse MAPK Mitogen-activating Protein Kinase ML Medial-lateral MOR µ-Opioid Receptor MR Medullary Raphé mRNA Messenger RNA NA Nucleus Ambiguus NK-1R Neurokinin-1 Receptor NTS Nucleus Tractus Solitarius OPRM1 Opioid Receptor µ 1 gene PAG Periaqueductal Grey PBN Parabrachial Nucleus PBS Phosphate Buffered Saline

PCO2 Partial Pressure of Carbon Dioxide pF Parafacial nucleus PiCo Post-Inspiratory Complex

PIP2 Phosphatidylinositol 4,5-bisphosphate POLR2A RNA Polymerase II Subunit A PPIB Peptidylprolyl Isomerase B preBötC preBötzinger Complex RGS Regulators of G-Protein Signaling RGS4 Regulators of G-Protein Signaling 4 RGS6 Regulators of G-Protein Signaling 6 RGS7 Regulators of G-Protein Signaling 7 RGS8 Regulators of G-Protein Signaling 8 RGS9 Regulators of G-Protein Signaling 9 RM Repeated Measures RTN Retrotrapezoid Nucleus

vii

RVM Rostral Ventrolateral Medulla rVRG Rostral ventral respiratory group SEM Standard Error of the Mean S1 Primary Somatosensory Cortex S2 Secondary Somatosensory Cortex SST Somatostatin THA Thalamus UBC Ubiquitin C

viii

List of Figures

1. Figure 1: Brainstem respiratory centers that highly express MORs ...... 3 2. Figure 2: Neural networks involved in respiratory depression by opioids ...... 9 3. Figue 3: Neural circuits mediating nociception ...... 13 4. Figure 4: Molecular mechanisms of MOR-induced signaling...... 18 5. Figure 5: In vivo microperfusion of drugs into preBötC can reverse rspiratory rate depression by MOR agonists ...... 28 6. Figure 6: Low-dose gallein administration into preBötC partially reverses respiratory rate depression by MOR agonists in vivo...... 36 7. Figure 7: Administration of gallein (high-dose of 5 mM) into preBötC reverses respiratory rate depression by the MOR agonist DAMGO in vivo...... 37 8. Figure 8: Correlation between location of the perfusion site and intensity of respiratory rate depression induced by DAMGO...... 39 9. Figure 9: mRNAs for TACR1 (gene for NK-1R) and OPRM1 (gene for MOR) are co- expressed in preBötC neurons...... 41 10. Figure 10: RGS4 and OPRM1 are co-expressed in key brainstem respiratory regions. . 43 11. Figure 11: Functional role of RGS4 in modulating respiratory rate depression by the MOR agonist DAMGO...... 46 12. Figure 12: Respiratory rate depression by MOR agonists in the absence of RGS4 activity is only partially reversed by naloxone...... 49 13. Figure 13: Significant correlation between distances from perfusion sites to the preBötC and latency for respiratory rate to decrease by 10% in response to DAMGO...... 50 14. Supplemental Figure 1: Coronal brainstem sections of experiments with the probe placed further than 1.5 mm away from the preBötC ..…………………………………..63 15. Supplemental Figure 2: Preliminary data showing mRNAs for TAC1R (gene for NK- 1R) and OPRM1 (gene for MOR) are co-expressed in preBötC neurons …………...….64

List of Tables

1. Table 1: Opioid receptor subtypes ...... 16 2. Table 2: RGS proteins involved in MOR pathways...... 23 3. Table 3: Gallein correlation map data ...... 42 4. Table 4: RGS4 correlation map data ...... 51

Chapter 1: Introduction

Rationale Opioids are powerful analgesics that are frequently used in medicine today, but they also can present with various harmful side effects, including life-threatening respiratory depression (Dahan et al., 2010). Opioid drugs have been the mainstay of pain management for centuries due to their potent analgesic properties. These drugs are commonly prescribed to treat acute pain, and are also used to manage chronic pain, defined as pain that persists beyond 3 months (Chou et al., 2009). Evidence of the use of the opium poppy Papaver somniferum for medicinal purposes dates back to the Sumerians in 3000 BCE (Brownstein, 1993). In the 19th century, isolation of morphine as the active compound in the opium poppy allowed for the synthesis of novel opioid drugs, each with their own pharmacological properties (Brownstein, 1993). Aside from their analgesic qualities, opioids present various detrimental side effects, including constipation, sleep disturbances, an increased sensitivity to painful stimuli (hyperalgesia), addiction, tolerance, physical dependence, and respiratory depression (Alanmanou, 2006). These drugs are highly addictive, as they induce euphoria and can cause dysphoria if chronic use is terminated (Kolodny et al., 2015). Prolonged opioid use for chronic pain carries a high risk of developing addiction, and most patients who died of opioid overdose deaths while receiving opioid pain relievers showed evidence of addiction (Kolodny et al., 2015). Furthermore, 80% of current heroin users report that their opioid use began with prescription opioids (Kolodny et al., 2015). The high prevalence of medical opioid use, coupled with its highly addictive qualities, high potential for misuse, and potential for lethal overdose, has created a major health crisis today. The opioid crisis is reaching new heights in Canada and the United States. Opioid prescription rates have risen steadily since the mid-1990s (Rudd et al., 2016), leading to widespread misuse and addiction (Helmerhorst et al., 2017). Today, millions of Americans regularly use opioids to manage pain (Volkow & McLellan, 2016). From 1999 to 2014, prescription opioid sales have quadrupled, even though overall patient-reported pain stayed the same (Helmerhorst et al., 2017). Mortality due to opioid misuse has similarly increased over this time. By 2011 in the United States, there were 400,000 emergency department visits involving non-medical opioid prescription drugs (Helmerhorst et al., 2017), and over 90 have died daily of opioid overdose (Rudd et al., 2016). In 2016, over 42,000 people died of opioid overdose in the

United States alone (Volkow & Koroshetz, 2019). However, opioid-related deaths only partially explain the effects these drugs have on our society. Cost of hospitalizations, time lost from work, and poor quality of life after recovery from overdose all contribute to the burden opioids place on our society. Opioid overdoses cost the US healthcare system $115 billion in 2017 (Haffajee & Frank, 2018). The increase in opioid usage has outpaced that of other addictive illicit drugs, such as cocaine and methamphetamines (Rudd et al., 2016). The potent analgesic qualities of opioids make them a mainstay in the treatment of both acute and chronic pain, but they present various detrimental side effects (Jamison & Mao, 2015). In opioid overdose, patients show severe respiratory depression, which can lead to respiratory arrest, hypoxia and eventually death if not treated with the opioid receptor antagonist naloxone, or Narcan® (Alanmanou, 2006). Respiratory depression is characterized by decreased respiratory rate and reduced airflow, causing a reduction in tidal volume, CO2 output, and O2 intake (Macintyre et al., 2011). This reduction in ventilation leads to reduced blood oxygen levels (hypoxia) and increased blood CO2 (hypercapnia) (Montandon et al., 2016a). Although naloxone can be used to reverse the effects of respiratory depression, it is not an ideal drug because it blocks the analgesic effects of opioids when administered, and has a shorter half-life than most opioid analgesics, which leads to re-narcotization (Alanmanou, 2006). A therapy that could reverse or prevent opioid- induced respiratory depression without affecting analgesia is currently unavailable (Montandon & Slutsky, 2019), due in part to a lack of understanding of the molecular mechanisms regulating opioid inhibition and the impacts of opioid ligands on the neural control of breathing. This thesis investigates the G-protein-coupled receptor mechanisms regulating opioid-induced respiratory rate depression in anesthetized rats in vivo. An in-depth understanding of the current use of these drugs, their mechanisms of drug action, and the neural physiology of the affected neural circuits is needed to better understand respiratory depression by opioids and identify safe opioid therapies. Thus, opioid pharmacology, the neural physiology controlling both breathing and nociception, and the effects of opioids on these neural circuits are covered in detail in the introduction below.

Ventilation Breathing is a rhythmic behavior regulated by respiratory circuits in the medulla and pons, including the medullary raphé neurons (Severson et al., 2003), nucleus tractus solitarius (NTS) (Zoccal et al., 2014), parabrachial nucleus (PBN) / Köllker-Füse (KF) neurons (Barnett et al.,

2018), post-inspiratory complex (PiCo) (Anderson et al., 2016), the retrotrapezoid nucleus (RTN) (Feldman et al., 2003), the Bötzinger Complex (BötC) (Jiang & Lipski, 1990), and the preBötzinger Complex (preBötC) (Smith et al., 1991) (Figure 1). Neural circuits in the brainstem produce rhythmic motor activity, activate respiratory muscles and promote ventilation (Feldman & Del Negro, 2006). A respiratory cycle consists of two distinct phases: inspiration and expiration (Richter & Spyer, 2001). Inspiration is an active process, while expiration is normally a passive process involving relaxation of inspiratory muscles. Respiration is generated by three distinct phases of activity in respiratory neurons, each originating from distinct brainstem nuclei: inspiration, post-inspiration, and expiration (Anderson & Ramirez, 2017).

Figure 1: Brainstem respiratory networks involved in respiration and expression of µ-opioid receptors (MORs). Sagittal view of rat brain showing respiratory nuclei of interest. The nucleus

tractus solitarius (NTS), located in the dorsal-medial medulla, receives O2 and CO2 chemosensory input from the carotid bodies to modulate breathing. The medullary raphé (MR) are located near the brainstem midline and are involved in both nociception and the body’s ventilatory response to hypercapnia. The Kolliker-Fuse (KF) and parabrachial nucleus (PBN) control the inspiratory- expiratory phase transition during breathing. The Bötzinger Complex (BötC) inhibits the activation of inspiratory muscles via inhibition of the rostral ventral respiratory group (rVRG) and phrenic

nerve to produce passive expiration, while the lateral parafacial nucleus (pFL) mediates active expiration via activation of the caudal ventral respiratory group (cVRG). The preBötzinger Complex (preBötC) is located in the ventrolateral medulla and drives inspiratory rhythm by innervating the

rVRG, which in turn innervates the phrenic and intercostal nerves. Labelled in red, the NTS, MR, PBN, and KF are respiratory circuits expressing MORs and sensitive to opioid administration.

Inspiration in mammals is driven by the preBötC, a bilateral population of neurons located in the ventrolateral medulla (Smith et al., 1991). Pre-inspiratory drive is produced by activation of rhythmogenic propriobulbar preBötC neurons, which provide excitatory inputs to bulbospinal respiratory neurons in the rostral ventral respiratory group (rVRG), dorsal respiratory group, and premotor neurons in the lateral reticular formation (Smith et al., 2009). These groups of neurons then synapse onto motor neurons that activate the diaphragm, external intercostal muscles, and hypoglossal motor pool (Onimaru & Homma, 2006). The diaphragm and intercostal muscles act as pump muscles to expand the thoracic cavity, creating negative pressure that draws air into the lungs (Del Negro et al., 2018). The genioglossus muscle, which is innervated by the hypoglossal motor pool, dilates and/or stiffens the upper airways, allowing for airflow into the lungs (Kobayashi et al., 1996). PreBötC neurons also project onto other brainstem regions regulating breathing, including the BötC, PBN / KF nucleus, RTN, and the NTS (Tan et al., 2010; Yang & Feldman, 2018). The boundaries of the preBötC are not well defined, however the transcription factor Dbx1 is known to be expressed in neurons that mature into preBötC glutamatergic neurons and glia (Gray et al., 2010). Dbx1-/- mice form improper preBötC and die soon after birth due to an inability to breathe (Gray et al., 2010), and optogenetic photoinhibition in mice of archaerhodopsin- expressing Dbx1 preBötC neurons stops inspiratory rhythm in vitro and breathing in vivo (Koizumi et al., 2016; Vann et al., 2016), demonstrating the importance of these Dbx1-expressing precursor neurons in the development and rhythmicity of the preBötC. The mature preBötC is comprised of mixed populations of neurons expressing neurokinin-1 receptors (NK-1R), and the peptide somatostatin (SST) (Stornetta et al., 2003). NK-1R preBötC neurons produce inspiration in mammals (Gray et al., 1999). Indeed, gradual ablation of these neurons causes severe, graded breathing pathologies in rats that first manifest during sleep (Gray et al., 2001; McKay et al., 2005). SST neurons synapse onto preBötC NK-1R expressing neurons and can send excitatory or inhibitory signals to synchronize and regulate respiratory rhythmogenesis (Wei et al., 2012). Studies on the silencing of SST neurons have produced contradictory results. One group studied mice that expressed allostatin receptors on SST-expressing preBötC neurons (Tan et al., 2008).

When allostatin, an inhibitory neuropeptide, was administered to awake adult rats, persistent apnea occurred, suggesting these neurons are necessary for eupnic breathing (Tan et al., 2008). However, another study silenced VGLUT2 gene expression in SST-expressing preBötC cells in mice and did not observe a change in viability, suggesting they are not necessary for eupnic breathing in vivo (Tupal et al., 2014). Post-inspiration involves continued contraction of the diaphragm and adduction of the laryngeal muscles to slow expiratory airflow (Del Negro et al., 2018). Post- inspiration is controlled by the PiCo, which is located in the ventral respiratory group rostral to the preBötC and dorsal-medial to the nucleus ambiguus (Anderson et al., 2016). Post-inspiration involves continued contraction of the diaphragm and adduction of the laryngeal muscles to slow expiratory airflow (Del Negro et al., 2018), as well as coordination of breathing with other behaviors that involve pharyngeal muscles, such as swallowing, vocalization, and coughing (Anderson et al., 2016). In passive expiration, inhibitory neurons of the BötC synapse onto neurons in the spinal cord (Jiang & Lipski, 1990). Passive expiration involves relaxation of previously active respiratory pump muscles, causing the thoracic cavity to contract and force air out of the lungs due to elastic recoil (Del Negro et al., 2018). The BötC also contains neurons that inhibit bulbospinal neurons of the rVRG and dorsal respiratory group that are activated by the preBötC during inspiration (Smith et al., 2009). Active expiration, which involves active contraction of expiratory muscles, is mediated by the lateral parafacial nucleus (pFL) (Abdala et al., 2009). Disinhibition of these neurons leads to bursting in these cells concurrent with active expiration (Pagliardini et al., 2011); this inhibition is mediated by the preBötC to prevent simultaneous activation of inspiratory and expiratory muscles (Del Negro et al., 2018). The pFL innervates bulbospinal interneurons of the caudal ventral respiratory group (cVRG), which in turn synapse onto motor neurons that innervate the abdominal and internal intercostal muscles (Abdala et al., 2009). Innervation of these muscles forces lung volume below its resting level, which increases the tidal volume of the subsequent breath (Del Negro et al., 2018). Active expiration occurs during episodes of increased ventilatory demand, such as exercise (Aliverti et al., 1997). These respiratory networks generate respiratory rhythm, but also receive modulatory inputs from the brainstem and cortex that comprise a complex neural network that has yet to be fully understood.

Respiratory Rhythm Modulation Various brainstem structures play important roles in the modulation of respiratory rate. These structures receive homeostatic and nociceptive inputs and, via projections to the preBötC, can increase or decrease respiratory rate in response to metabolic demands (Zhang et al., 2007; Zhang et al., 2011; Levitt et al., 2015). The NTS plays a major role in the coordination of respiratory and sympathetic activities (Zoccal et al., 2014). It receives chemoreceptor inputs from the carotid bodies that relay blood

PCO2 and PO2 information (Zhang et al., 2011), while also receiving cardiovascular inputs from arterial baroreceptors and pulmonary stretch receptors (Machado, 2001). It integrates the sensory information and modulates specific neural networks to produce appropriate homeostatic respiratory responses (Zoccal et al., 2014). During episodes of hypoxia and/or hypercapnia, the NTS sends inputs to the preBötC to modulate respiratory activity (Zhang et al., 2011). Neurons of the medullary raphé are located in the ventral portion of the brainstem along the midline. This region is an important site for the modulation of respiration, specifically in mediating the body’s ventilatory response to hypercapnia (Severson et al., 2003). A subpopulation of these neurons in the caudal portion of the medullary raphé project to the preBötC and increase ventilatory output in response to increased blood CO2 levels (Severson et al., 2003). Furthermore, ventilation increases when acidosis is induced in vivo in serotonergic neurons of the medullary raphé of cats and goats by local administration of CO2, indicating these neurons play a role in chemoreception (Feldman et al., 2003). The medullary raphé are also known to play a role in pain modulation, as activation of these neurons modulates dorsal horn processing of nociceptive information via descending pain pathways (Mason, 2001). Currently, there is debate as to whether the medullary raphé or the RTN functions more as a central chemoreceptive area in the brainstem. Although there is significant in vitro and in vivo work suggesting serotonergic cells of the medullary raphé, the Guyenet group cites the low percentage of serotonergic neurons that responded to hypercapnia and excessively high levels of

CO2 used to stimulate serotonergic neurons in some studies as reasons to believe the data is insufficient in claiming these neurons as a key area of central chemoreception (Guyenet et al., 2005). They instead contend that cells of the RTN are activated by a change in pH that is unchanged by the presence of glutamate and purinergic agonists, and that RTN cells maintain their response to hypercapnia in vivo even after cessation of respiratory pattern by kynurenic acid

(Guyenet et al., 2005). While this data only shows that a subset of RTN cells respond to acidosis, we do know that the RTN detects blood PCO2 and pH levels and transmits signals to the preBötC and other brainstem respiratory sites to modulate respiratory rate (Feldman et al., 2003).

The parabrachial nucleus receives CO2 chemosensory input from the RTN and NTS (Herbert et al., 1990). During sleep, hypercapnic signaling from these areas to glutamatergic neurons of the lateral parabrachial nucleus, which includes the KF nucleus, mediates arousal from sleep (Kaur & Saper, 2019). The KF nucleus is a neuronal subpopulation of the parabrachial nucleus located in the dorsal-lateral pons that contributes to respiratory rhythm generation (Barnett et al., 2018). KF neurons are essential in mediating the transition from the inspiratory to expiratory phase of respiration as well as initiating the post-inspiratory phase (Dutschmann & Dick, 2012), and inhibition of KF neurons leads to prolonged inspiratory duration and abnormal breathing (Bautista & Dutschmann, 2014). Tongue and laryngeal muscular tones are also influenced by KF neurons. Excitation of KF neurons promotes laryngeal constriction and swallowing-induced resetting of the respiratory cycle (Bonis et al., 2013), as well as activation or suppression of hypoglossal motor neurons that in turn innervate tongue protrusor muscles (Kuna & Remmers, 1999). While the roles of these nuclei in the generation and modulation of respiratory rhythm have been well characterized, much less is known about their specific roles in respiratory depression by opioids, especially in humans.

G-protein Modulation of Respiratory Rhythm in the preBötC Respiratory rhythm can be modulated, in part, by activating various G-protein signaling pathways in the preBötC. NK-1Rs, present on opioid-sensitive preBötC neurons that drive inspiration (Gray et al., 1999), are G-protein coupled receptors (GPCRs) that, when bound to

Substance P, increase respiratory rate (Gray et al., 1999). SST2 receptors are GPCRs found on glutamatergic preBötC neurons (Gray et al., 1999; Stornetta et al., 2003), and when bound to SST, they decrease respiratory rate (Gray et al., 2010; Montandon et al., 2016b). SST-expressing preBötC neurons further modulate respiratory rhythmogenesis by projecting to other brainstem nuclei involved in the control of breathing to, including the BötC, rVRG, cVRG, RTN, pF, NTS, PBN/KF, and periaqueductal gray (Tan et al., 2010). MORs on NK-1R expressing neurons in the preBötC bind opioids to cause a pronounced decrease in respiratory rate (Gray et al., 1999). The G-protein signaling pathways that are activated by opioid binding in the preBötC and other opioid-

sensitive areas are responsible for the various effects that opioids have on respiration that contribute to respiratory depression.

Opioid Effects on Respiration

In humans, opioid drugs decrease ventilation by decreasing respiratory rate and tidal volume, as well as decreasing CO2 chemosensitivity. Opioid drugs such as morphine decrease respiratory rate in humans (Adriani, 1953), and decrease the hypercapnic ventilatory response (Weil et al., 1975). Chronic opioid users present increased numbers of central apneic episodes during sleep (Jungquist et al., 2012). Highly potent synthetic opioids such as fentanyl also cause rigidity of respiratory muscles in the chest wall, which leads to difficulty of breathing and decreased chest wall compliance (Glick et al., 1996). Chronic heroin use has been observed to cause bronchospasms in patients, leading to airway constriction (Cygan et al., 2000). Commonly used opioid drugs, such as fentanyl and morphine, display similar respiratory effects in rats as they do as humans (Santiago & Edelman, 1985; Dahan et al., 2005; Kuo et al., 2015), making rats a useful animal model to study opioid-induced respiratory depression. Morphine causes severe respiratory depression in rats, and at the highest doses, causes respiratory depression that lasts for 2 hours (Kuo et al., 2015). Fentanyl-induced respiratory depression is more potent and displays a faster onset than morphine (Kuo et al., 2015). Opioids reduce all parameters of respiration, including decreasing rate and depth of respiration, decreasing upper airway patency, and decreasing the ventilatory responses to hypoxia and hypercapnia (Lalley, 2008). Systemic naloxone administration reverses the effects of opioids (van den Hoogen & Colpaert, 1986). Interestingly, the severity of respiratory rate depression by fentanyl is associated with the sedative properties of opioids (Montandon & Horner, 2019b). Although the effects of opioids on breathing and ventilation are well documented, the neural and molecular mechanisms regulating respiratory rate depression are not well-known.

Neural Mechanisms of Opioid-Induced Respiratory Depression

While opioid analgesia has been investigated at length, respiratory depression by opioids has been far less studied. Nonetheless, several brainstem cell populations have been identified that highly express MORs and regulate respiratory rate depression due to systemic or local opioid

administration (Figure 2). However, it has yet to be determined which region, or regions, mediates respiratory depression in humans, as different species have exhibited differential responses to local opioid administration into discrete brainstem regions.

Figure 2: Neural networks involved in respiratory depression by opioids. Sagittal view of rat brain indicating brainstem areas contributing to respiratory depression by opioids. Opioid administration into the preBötzinger Complex (preBötC), Kölliker-Fuse (KF) nucleus, parabrachial nucleus (PBN), medullary raphé (MR), and rostroventrolateral medulla (RVM) induced respiratory depression in mammals. Opioid administration into the nucleus tractus solitarius (NTS) blunts the ventilatory responses to hypoxia and hypercapnia. Opioids in the periaqueductal gray (PAG), anterior cingulate cortex (ACC), and second somatosensory cortex (S2) inhibit the volitional aspects of breathing that can potentiate respiratory depression.

PreBötzinger Complex

The preBötC has been identified as a key area of interest in opioid research because of its ability to generate respiratory rhythm (Smith et al., 1991; Montandon & Horner, 2014). The preBötC contains NK-1R expressing neurons that also express MORs (Gray et al., 1999), and their activation causes severe respiratory rate depression both in vitro (Johnson et al., 1996; Mellen et al., 2003) and in vivo (Montandon et al., 2011).

Fentanyl administration into the preBötC slows the firing rate of respiratory rhythm- generating inspiratory neurons in both decerebrate cats (Lalley, 2005) and juvenile rats (Janczewski & Feldman, 2006). While systemic fentanyl administration into juvenile rats also induces “quantal slowing” of respiratory rate, an abnormal pattern of respiration that leads to skipped cycles of inspiratory output from the preBötC, possibly due to an opioid-induced imbalance of preBötC activity with rate-modulating inputs from opioid-insensitive pre-inspiratory neurons (Mellen et al., 2003), this finding has not been repeated by other investigators. Local bilateral administration of naloxone does not increase respiratory rate by itself, but reverses respiratory rate depression from systemic opioid administration in rodents (Montandon et al., 2011). DAMGO microinjection into the preBötC of dogs has the opposite effect as it does in rodents, instead producing tachypnea (Mustapic et al., 2010), although in this study the preBötC was defined as the region in the ventrolateral medulla that was most sensitive to glutamate-induced tachypnea, and therefore they may not have targeted NK-1R preBötC neurons. Microinjection of DAMGO at pharmacological concentrations (100 µM) into the preBötC of both young and adult rabbits induces robust respiratory rate depression (Stucke et al., 2015). Local bilateral injection of naloxone is insufficient in reversing respiratory rate depression in dogs (Mustapic et al., 2010) and rabbits (Stucke et al., 2015). Bilateral injection of DAMGO into the preBötC of awake adult goats does not produce respiratory rate depression (Krause et al., 2009). A recent study using mice with MOR-/- in the preBötC shows attenuated respiratory rate depression compared to wild type from low, anti-nociceptive doses of morphine (10 mg/kg) into the preBötC, but the MOR-/- mice show similar respiratory rate depression as wild type at high morphine doses (30 mg/kg), instead increasing the amount of apneas observed (Varga et al., 2019). This suggests that the preBötC may not be the only neuronal population responsible for opioid- induced respiratory rate depression, and that targeting multiple areas may be necessary to reverse respiratory rate depression in humans. Nevertheless, due to its crucial role in respiratory rhythm generation, its high expression of MORs, and the plethora of in vitro and in vivo work in multiple species showing targeting the preBötC alone is sufficient in inducing respiratory rate depression by opioids, the preBötC may constitute a potential target cell population to prevent opioid-induced respiratory depression in humans.

Medullary Networks

Opioids also cause depression of respiratory rate by acting on specific neural cell populations in the brain. While high concentrations (1 mM) of DAMGO microinjected into bulbospinal premotor neurons produces significant respiratory rate depression, local microinjection of remifentanil and morphine at clinically relevant concentrations are unable to induce respiratory rate depression, and local naloxone injection cannot reverse respiratory rate depression from intravenous remifentanil administration (Stucke et al., 2008), suggesting opioids act on upstream respiratory circuits to depress breathing. Systemic DAMGO injection into adult rats produces respiratory rate depression that is partially reversed by local administration of D-Phe-Cys-Tyr-D-Trp-Arg-Thr-Pen-Thr-NH2 (CTAP), an MOR antagonist, into the caudal region of the medullary raphé (Zhang et al., 2007), indicating that this cell population plays a limited role in opioid-induced respiratory depression, and cannot be targeted to fully reverse the respiratory effects of opioids. DAMGO microinjection into the rostral ventrolateral medulla (RVM), a large brainstem region comprised of the nucleus raphé magnus, the nucleus reticularis gigantocellularis pars alpha, and the nucleus paragigantocellularis lateralis (Renn & Dorsey, 2005), produces respiratory rate depression, possibly via the modulation of ‘ON’ cells (Phillips et al., 2012), which are neurons that show a burst of activity prior to a reflexive response to noxious stimuli, such as the tail-flick response (Fields et al., 1995). The RTN has been shown to be insensitive to opioids in juvenile rats (Janczewski & Feldman, 2006). Local DAMGO administration into the NTS in rats reduces the hypoxic ventilatory response, while local CTAP administration after systemic DAMGO partially reverses this reduction (Zhang et al., 2011). However, local DAMGO administration to the NTS fails to produce significant respiratory rate depression. Opioids also reduce motor neuron output of the hypoglossal nucleus to the genioglossus, a muscle in the tongue that controls upper airway tone (Hajiha et al., 2009). Furthermore, opioids decrease the frequency of respiratory motor neuron discharge from cervical neurons in adult rat brainstem slices (Murakoshi et al., 1985). While these areas themselves do not seem to mediate respiratory rate depression by opioids, their neural connections to the preBötC, while not yet fully understood, play an important role in modulating the effects of opioids on respiration.

Pontine Networks

DAMGO administration into the PBN of both young and adult rabbits leads to naloxone- reversible respiratory rate depression, but intravenous remifentanil-induced respiratory rate depression is only partially reversed by local naloxone administration (Miller et al., 2017). Local DAMGO injection into KF neurons, which control the opening of the upper airways and mediates the inspiratory/expiratory phase transition (Levitt et al., 2015), of decerebrate dogs produces bradypnea, and local naloxone injection partially reverses respiratory rate depression from systemic opioid injection (Prkic et al., 2012). Furthermore, brain slices containing KF neurons become hyperpolarized upon MOR activation, and the subsequent outward current in these neurons is reversed by administering barium chloride, suggesting potassium channels may be responsible for these effects (Levitt et al., 2015). In mice with MOR-/- in KF neurons, respiratory rate depression by systemic morphine is attenuated at all doses (Varga et al., 2019), suggesting the KF nucleus could constitute another key target in respiratory depression by opioids.

Cortical/Subcortical Networks

Several cortical and subcortical regions are involved in the volitional control of breathing. The periaqueductal gray (PAG) is the main behavioral modulator of breathing. Various subpopulations of neurons in the PAG modulate breathing in response to dangerous or stressful stimuli via their connections to respiratory networks in the brainstem (Subramanian et al., 2008). Using fMRI, remifentanil was shown to depress the neural activity in the anterior cingulate cortex (ACC), periaqueductal grey PAG, secondary somatosensory cortex, and the left dorsal-lateral prefrontal cortex associated with voluntary modulation of respiratory muscle activity (Pattinson et al., 2009a). These effects on higher brain regions to modulate respiration potentiates the depressive effects of opioids on brainstem-mediated control of breathing. For example, these areas have been implicated in awareness of internal body states, especially the experience of dyspnea (von Leupoldt & Dahme, 2005), and are partially activated during hypercapnia (Pattinson et al., 2009b). Together, they drive the body’s volitional urge to breathe, and the depressive effects of opioids in these cortical and subcortical structures potentiates their effects on respiration by diminishing the urge to breathe that would normally accompany the hypercapnia and bradypnea caused by opioids effects on brainstem respiratory circuits. However, because of the challenges of studying volitional

control of breathing in animal models, and because human studies have been limited mostly to fMRI studies, our understanding of the cortical and subcortical mechanisms of opioids that influence respiratory depression is relatively scarce.

Neural Circuits of Nociception

Nociception occurs when the free nerve endings of afferent sensory nerves, or nociceptors, transduce noxious chemical, mechanical, or thermal stimuli in peripheral tissue into electric signals to be transmitted throughout the CNS (Willis, 1985). Nociceptive fibers have their cell bodies located in the dorsal root ganglion (DRG) and synapse onto second order projection neurons in the dorsal horn of the spinal cord (Steeds, 2009). After nociceptive signaling is received by second order neurons in the dorsal horn, it is transmitted rostrally along the spinal cord to higher brain centers via ascending spinal cord tracts (Figure 3) (Willis, 1985). After integration from various ascending pathways in the thalamus (Renn & Dorsey, 2005), the nociceptive information reaches the cerebral cortex to undergo cognitive and emotional processing via a complex network of interconnected structures including the somatosensory S1 and S2 areas (Casey et al., 1996), and the anterior cingulate cortex (Derbyshire et al., 1997). Pain perception undergoes further modulation from higher brain areas that exert modulatory control onto spinal nociceptive neurons (Hagbarth & Kerr, 1954). This process, known as descending modulation of pain, is predominantly mediated by the PAG (Mantyh, 1983) and RVM (Heinricher et al., 2009).

Figure 3: Neural circuits mediating nociception. Sagittal view of rat brain indicating key brain networks involved in ascending (blue) and descending (black) nociception pathways. The thalamus (THA) receives nociceptive input from the spinal cord via ascending pain pathways and sends that information to various cortical brain areas for emotional and cognitive processing, including the primary somatosensory cortex (S1), secondary somatosensory cortex (S2), and the anterior cingulate cortex (ACC). Descending pain modulation is mediated mostly by the periaqueductal grey (PAG), which modulates the excitability of spinal cord nociceptors via its connections with neurons in the rostral ventromedial medulla (RVM).

A major portion of the neurons projecting to the spinal dorsal horn originate in the RVM (Willis, 1988). The RVM exerts both pro-nociceptive effects through the activity of ‘ON cells’ that increase firing onto spinal dorsal horn nociceptive neurons in response to noxious stimuli and anti-nociceptive effects through the activity of ‘OFF cells’ that directly inhibit spinal neurons (Heinricher et al., 2009). “OFF” cells are GABAergic neurons that inhibit pain by projecting onto and inhibiting nociceptive neurons in the spinal cord (Heinricher et al., 2009). The PAG is a midbrain structure composed of densely packed heterogeneous neurons surrounding the central aqueduct (Bandler & Shipley, 1994). Few PAG neurons project directly to the spinal cord dorsal horn (Mantyh & Peschanski, 1982); the PAG predominantly exerts its descending pain modulatory effects indirectly through its connections to various brainstem structures, primarily the RVM (Mantyh, 1983), by either inhibiting pro-nociceptive pathways or exciting inhibitory GABAergic pain pathways (Fields, 1999).

Opioid Analgesia

Opioids exert their analgesic effects by acting on MORs in both spinal and supraspinal structures to suppress neurotransmitter and neuropeptide release and inhibit the activity of nociceptive neural networks and transmission of nociceptive information (Stein, 2018). RVM neurons highly express MORs (Gutstein et al., 1998). DAMGO microinjection into the RVM is sufficient to produce analgesia (Rossi et al., 1994). RVM “ON cells are directly inhibited by MOR administration (Barbaro et al., 1986), while “OFF cells”, so called because they display a pause in activity prior to reflexive responses to noxious stimuli (Fields et al., 1983), are activated by MORs, likely in an indirect manner (i.e. inhibition of interneurons that themselves inhibit OFF cells)

(Fields et al., 1983). Systemic morphine administration causes “OFF” cells to become continuously active (Nicoll et al., 1980). The PAG is an essential structure in the neural analgesia pathway (Basbaum & Fields, 1978). Local morphine administration in the PAG produces profound analgesia (Rossi et al., 1994). Approximately half of all PAG neurons projecting to the RVM express MORs (Commons et al., 2000). While the neural circuits mediating opioid analgesia have been studied, the specific molecular mechanisms underlying analgesia have yet to be fully determined.

Opioid Sedation Opioid drugs also display potent sedative effects which can potentiate their effects on respiration. Sedation by opioids is characterized by reduced motor activity, loss of attention, changes in electroencephalogram activity, and depressed arousal (Young-McCaughan & Miaskowski, 2001; Wang & Teichtahl, 2007). Rodent studies have demonstrated respiratory depression by opioids is more pronounced in states of reduced arousal, as microperfusion of DAMGO into the preBötC of freely-behaving rats was more pronounced during non-REM sleep than wakefulness (Montandon et al., 2011). Thus, systemic opioid administration could lead to greater respiratory rate depression than local administration of opioids into brainstem respiratory networks, as the respiratory depression caused by opioids may be potentiated by their sedative effects as well. Accordingly, opioid-induced respiratory depression is increased during sedation in patients in perioperative settings (Lee et al., 2015). Further studies show opioid-induced changes in electrocortical activity that underlie sedation are tightly associated with respiratory rate depression severity (Montandon et al., 2016a; Montandon & Horner, 2019b), suggesting opioid- induced sedation can potentiate opioid-induced respiratory depression in humans, possibly by acting on overlapping neural structures. The PAG is a prime possibility, as it plays a role in opioid analgesia and respiratory depression, and is known to play a key role in sedation, possibly through opioid-mediated inhibition of GABAergic neurons (Brown et al., 2011). The convergence of opioid analgesic, sedative and respiratory depression pathways on the PAG make it a key area of

interest in opioid research, but the specific neural mechanisms involved in the interaction between sedation and respiratory depression remain unknown.

Opioid Receptors and Opioid Pharmacology

Opioid drugs act on several subtypes of opioid receptors, including κ-, δ-, and µ-opioid receptors (MORs) (Waldhoer et al., 2004). Table 1 displays the various opioid receptors and which endogenous opioid and opioid drug binds to each one. Morphine, a common analgesic used to treat moderate to severe pain (Trescot et al., 2008), is highly selective to MORs, with a binding afinity (Ki) value of 5.7, compared to 456 and and 120 in δ- and κ-opioid receptors, respectively (Vecchietti et al., 1992; Shreder et al., 1998; Dosa & Amin, 2016). Ki is a measure of binding affinity where lower values indicate greater binding affinity a ligand displays to the receptor. DAMGO, which is also highly selective to MORs, has a higher affinity to MORs compared to morphine, with a Ki of 3.8 (Mathew et al., 2009). Fentanyl, another commonly used opioid that is highly selective to MORs, has even higher affinity to the MOR, with a Ki of 1.6 (Poulain et al., 2001). Naloxone, an opioid antagonist, has a Ki of to the MOR (Majumdar et al., 2011). Naloxone also antagonizes opioid binding at δ- and κ- opioid receptors, with a Ki of 42.6 and 9.2 respectively (Sweetnam et al., 1993; Le Bourdonnec et al., 2008).

Table 1: Opioid Receptor Subtypes. Table of the three endogenous opioid receptors. Includes which endogenous opioid ligand has the highest affinity to each receptor and which commonly used opioid drugs preferentially bind to each receptor. Opioid Receptor µ-opioid δ -opioid κ-opioid Subtype Endogenous ligands β-endorphin Met-enkaphalin/ Dynorphin A/B leu-enkaphalin

Opioid agonists Morphine, fentanyl, Meperidine, codeine Butorphanol, meperidine, enadoline, DAMGO, heroin, pentacozine, hydrocodone, buprenorphine oxycodone, codeine, buprenorphine Opioid Antagonists Naloxone Naloxone Naloxone

G-Protein Coupled Receptors

Most opioid analgesics, such as oxycodone, fentanyl, and morphine, are highly selective for MORs (Pasternak & Pan, 2011). MORs are GPCRs that, when activated, inhibit neuronal activity by activating various downstream signaling pathways (Figure 4) (Pasternak & Pan, 2011). These signaling pathways lead to several physiological effects of opioids including analgesia, respiratory depression, and sedation (Waldhoer et al., 2004). Opioids exert their inhibitory effects by decreasing neuronal excitability and inhibiting neurotransmitter release (Sicuteri et al., 1983; Vaughan et al., 1997; Zhu & Pan, 2005). These inhibitory effects are mediated through activation of MORs on either pre-synaptic neurons, largely via the inhibition of N-type calcium channels (Heinke et al., 2011), or post-synaptic neurons, mainly via the inhibition of adenylyl cyclase or the activation of potassium channels (Collier & Roy, 1974; Torrecilla et al., 2002) (Figure 4). The key G-protein mediated signaling mechanisms involved in opioid analgesia and respiratory depression have yet to be fully elucidated.

Figure 4: Molecular mechanisms of MOR-induced signaling. Binding of opioid to extracellular surface of MOR causes it to activate intracellular G-protein. Activated Gα undergoes conformational shift to exchange GDP with GTP, causing dissociation from Gβγ subunit. Gβγ can then bind to intracellular tail of GIRK channels, activating them and allowing potassium ions out of the cell. RGS proteins are GTPase-activating proteins, which accelerate the exchange of GTP on the Gα back to GDP, causing it to re-associate with the Gβγ subunit and to inhibit their activities in their respective signaling pathways. Gβγ can also bind to and inhibit N-type calcium channels. MOR activation can also lead to inhibition of adenyl cyclase, the enzyme responsible for cyclic AMP (cAMP) production. β-arrestins can also bind to MORs and internalize the receptor, leading to a buildup of tolerance.

G-Proteins

Once opioid ligands bind to MORs, a conformational change occurs in the three switch region of the intracellular G-protein Gα subunit that surrounds the GDP binding pocket (Bohm et al., 1997), thus facilitating the switch from the Gα subunit GDP-bound inactive state to its GTP- bound active state and dissociation from the Gβγ subunit complex (Pasternak & Pan, 2011). The Gα and Gβγ complexes then diffuse through the cytoplasm to various target molecules in order to activate their secondary messenger pathways (Pasternak & Pan, 2011). The Gβγ subunit functions in various signaling pathways. Gβγ activation occurs when Gα dissociates and uncovers effector interaction sites on Gβγ (Camps et al., 1992). Unlike the Gα subunit, Gβγ does not have any catalytic domains, and thus signals via direct protein-protein interactions (Khan et al., 2013). There are 5 β and 12 γ subtypes, each of which have different affinities for specific GPCRs (Hurowitz et al., 2000). Gβ1-4 share 90% sequence homology among themselves, while Gβ5 only shares 52% similarity (Khan et al., 2013). Hydrophobic molecules are post-transcriptionally added to the Gγ subunit to anchor it to the plasma membrane, which aids in coupling it to specific GPCRs (Wedegaertner et al., 1995). Various Gβ and Gγ subtypes play a role in opioid receptor-activated cell signaling pathways. Gβ2 and Gβ4 are known to mediate the coupling of nociceptive receptors to presynaptic

N-type calcium channels (Mahmoud et al., 2012). Gβ5γ2 is a potent regulator of GIRK channels and N-type calcium channels (Yoshikawa et al., 2000). Inhibition of Gγ2, the most abundant Gγ subunit in the brain (Betty et al., 1998), decreases opioid-induced analgesia in mice (Varga et al., -/- 2005). Gγ3 mice display resistance to morphine treatment (Schwindinger et al., 2009). -/- Interestingly, Gγ3 mice also demonstrate a lack of Gαi/o3β1/2γ3 protein complex expression, suggesting this specific complex plays a key role in opioid analgesia (Schwindinger et al., 2009). Currently, the molecules that regulate Gβγ subunit activity in the opioid-induced respiratory depression pathway, and the specific Gβ and Gγ isoforms that couple to MORs in this pathway, are unknown.

Adenylyl Cyclase

MOR activation also inhibits cAMP production (Yu et al., 1990). The Gαi subunit interacts with and acutely inhibits adenylyl cyclase, an enzyme that catalyzes cAMP production (Ingram & Williams, 1994). Downregulation of cAMP decreases intracellular calcium levels by reducing calcium influx and neuronal excitability (Al-Hasani & Bruchas, 2011). Adenylyl cyclase is known to play an important role in the opioid analgesic pathway, as inhibiting adenylyl cyclase produces analgesic effects in mice (Wang et al., 2011). Mice lacking the adenylyl cyclase 5 isoform (AC5-/-) show attenuated analgesic responses to morphine compared to wild-type (Kim et al., 2006).

Naloxonazine, an MOR antagonist that selectively blocks µ1 opioid receptors, antagonizes morphine-induced analgesia but has no effect on respiratory depression (Ling et al., 1985). Furthermore, fentanyl-induced respiratory depression in decerebrate rats is reversed by stimulation of adenylyl cylase in respiratory neurons in the ventral respiratory group (Ballanyi et al., 1997), although no further studies have implicated adenylyl cyclase in respiratory depression by opioids.

Calcium Channels

N-type calcium channels are voltage-gated channels that open to allow calcium ions influx into the cell (Al-Hasani & Bruchas, 2011). The Gβγ subunit binds to and inhibits these channels, causing a decrease in intracellular calcium and inhibition of calcium-dependent signaling pathways (Bourinet et al., 1996). One such pathway is presynaptic neurotransmitter release, which is significantly attenuated due to a lack of calcium influx into the neuron (Patil et al., 1996). MOR activation decreases calcium channel current on both pre- and post-synaptic neurons, while subsequent naloxone administration increases calcium channel current (Berecki et al., 2016). Attenuation of intracellular calcium levels decreases the excitability of these cells, thus inhibiting their nociceptive signaling (Wang et al., 2010). Drugs inhibiting voltage-gated calcium channels are currently used in chronic pain patients, although their efficacy is somewhat limited (Perret & Luo, 2009). However, there is no evidence that calcium channel inhibition mediates opioid- induced respiratory depression.

GIRK Channels

A third pathway regulating MOR inhibition is the activation of GIRK channels via the Gβγ subunit. Gβγ’s ability to modulate GIRK channel activation make it a key target in opioid research. The dissociated Gβγ subunit of intracellular G proteins binds to the intracellular C-terminal domain of GIRK channels to induce their activation (Dascal & Kahanovitch, 2015). However, unlike Gβ1-

4 -containing Gβγ subunits, Gβ5-containing subunit binding to GIRK channels inactivates the channel (Mirshahi et al., 2002). Multiple Gβγ binding sites exist on GIRK channels, and an increase in Gβγ subunits occupying these sites leads to increased GIRK channel current (Sadja et al., 2002). There are four different subunits of GIRK proteins (GIRK1-4), but only GIRK1-3 subunits are found in the brain, with the GIRK1 and GIRK2 subunits being the most common (Luscher & Slesinger, 2010). GIRK subunits form a tetrameric structure (Luscher & Slesinger, 2010). GIRK2 subunits also contain domains that bind to postsynaptic density protein 95 and synapse-associated protein 97, both of which tether GIRK channels to the same microdomain as MORs (Hibino et al., 2000). This allows for increased MOR coupling to GIRK channels, creating a rapid and specific response to MOR activation (Hibino et al., 2000). GIRK channels are described as inward rectifiers because their outward current is smaller than their inward current at the membrane reversal potential (Luscher & Slesinger, 2010). In the absence of GPCR-mediated activation, GIRK channels display basal activity that shifts the resting membrane potential by approximately -10 mV (Lüscher et al., 1997). Activation of GIRK channels by G-proteins leads to prolonged channel opening and augmented activity (Lüscher et al., 1997). When open, these channels cause potassium ions to flow out of the cell (Figure 4), hyperpolarizing the neuron (Dascal & Kahanovitch, 2015), which would decrease its firing rate. Along with Gβγ binding, the Gα subunit of the MOR can also bind to the cytosolic tail of the GIRK channel and enhance or inhibit its activation (Leal-Pinto et al., 2010). Binding of PIP2 and sodium ions to the GIRK channel is necessary to stabilize the channel’s open conformation (Luscher & Slesinger, 2010). Mice knockout studies show GIRK2 and GIRK3 subunits are involved in opioid analgesia (Al-Hasani & Bruchas, 2011), and that GIRK2 plays a significant role in respiratory depression by opioids (Montandon et al., 2016c). Along with respiratory depression and analgesia, GIRK channel activation regulates other opioid side effects, including some of the compulsive behavior seen in addiction by increasing the excitability of dopaminergic neurons in the ventral tegmental area (Bonci & Williams, 1997).

GIRK channels have been found in the preBötC, and GIRK2-/- mice exhibit reduced opioid- induced respiratory rate depression than wild-type mice (Montandon et al., 2016c). While it is known that GIRK channels mediate respiratory depression by opioids, the upstream mechanisms that regulate GIRK channel activation in respiratory depression have yet to be elucidated. However, GIRK channels are deactivated 100 times faster in vivo than predicted based on Gα’s intrinsic GTP hydrolysis properties, suggesting the presence of GTPase activating proteins (GAPs) that accelerate the GTP hydrolysis process (Breitwieser & Szabo, 1988).

RGS Proteins

Regulators of G-protein signaling (RGS) proteins are a family of 26 proteins (De Vries & Gist Farquhar, 1999). They bind to the Gα subunit and function as GAPs, which exchange the GTP bound to the Gα subunit of the activated MOR with GDP (De Vries et al., 2000). This exchange causes inactivation of the Gαi/o subunit, leading to its re-association with the Gβγ subunit complex and deactivation of both G protein subunits (De Vries et al., 2000) (Figure 4). The Gαi/o and Gβγ subunits are no longer able to carry out the signal initiated by ligand binding to the GPCR, such as activation of GIRK channels by Gβγ (Doupnik et al., 1997). Gα Gly183Ser point mutations lead to insensitivity of the Gα subunit to RGS protein binding without affecting other Gα activity (Lan et al., 1998). R4 family RGS proteins, which includes RGS1-5 and RGS8, act only as GAPs, and only act on Gαi/o and Gαq subunits (Soundararajan et al., 2008). Other RGS proteins, such as the R7 family that contains RGS6, 7, and 9, and 11, play roles in the mitogen-activated protein kinase (MAPK) pathway as well (Gold et al., 1997). RGS proteins act on specific GPCRs to differing effects. The N-terminus of RGS proteins determines the potency of GAP activity that the specific RGS protein has on a specific receptor (Xu et al., 1999). The N-terminus also regulates the rate of degradation of the RGS protein (Xu et al., 1999). Various RGS proteins have been implicated in opioid signal pathways, including RGS4, RGS6, RGS7 and RGS9 (Table 2). RGS4 is known to be highly expressed in the amygdala, locus coeruleus, nucleus accumbens, and throughout the spinal cord (Gold et al., 1997). Morphine administration leads to increased RGS4 mRNA in the nucleus accumbens, but decreased RGS4 mRNA in the locus coeruleus (Bishop et al., 2002). RGS4 expression is also increased in patients with chronic pain (Garnier et al., 2003). The modulation of GIRK channels by RGS proteins has been well documented (De Vries et al., 2000). In PAG neurons, transgenic mice with Gα subunits

that are insensitive to RGS binding exhibit decreased GIRK channel activation from DAMGO and fentanyl administration compared to wild types (McPherson et al., 2018). GIRK channel activation is also attenuated in the presence of Gαo1 inhibitors, and this attenuation of GIRK channels only decreases spinal antinociceptive signaling and not supraspinal signaling (McPherson et al., 2018). While this study has shown that RGS proteins can modulate opioid-induced GIRK channel activation in antinociceptive circuits, the RGS protein subtypes involved in this pathway are yet to be determined, and their roles specifically in the respiratory depression pathway are unknown.

Table 2: RGS proteins involved in MOR pathways. Various RGS protein subtypes and their known roles in opioid signaling pathways. RGS proteins, including RGS4, have been implicated in opioid reward, tolerance, and analgesic pathways, but no RGS subtype has yet been shown to play a role in respiratory depression.

RGS Subtype Opioid Pathway

RGS4 Reward (Kim et al., 2018), analgesia (Yoon et al., 2015)

RGS6 Tolerance (Garzon et al., 2003), (Stewart et al., 2015)

RGS7 Reward (Sutton et al., 2016)

RGS8 Tolerance (Saitoh et al., 2001)

RGS9 Reward (Gaspari et al., 2014), tolerance (Garzon et al., 2001), analgesia (Psifogeorgou et al., 2011)

Research Question

To demonstrate whether G-protein signaling regulates respiratory rate depression by opioids, we aim to modulate G-protein activity in key brainstem respiratory neurons regulating respiratory rate depression by opioids. The key question is whether MOR inhibition of respiratory circuits is mediated by G-protein signaling mechanisms. It has previously been shown that GIRK channel activity, which itself is activated by Gβγ proteins (Dascal & Kahanovitch, 2015), mediates opioid-induced respiratory depression in the preBötC (Montandon et al., 2016c). Here, we propose the G-protein signaling mechanisms play a significant role in mediating respiratory rate depression by opioids.

Central Hypothesis

We hypothesize that neuronal inhibition of respiration by MORs is regulated by Gβγ protein activity and regulators of G-protein signaling in the preBötC, two key second messengers regulating G-protein signaling. To investigate our hypothesis, we have defined three objectives:

Aims and Objectives 1) To determine the role of Gβγ activity in MOR-mediated respiratory rate depression in the preBötC.

2) To assess co-expression of RGS4, a prototypical RGS protein that functions solely as a GAP, and MOR receptors in brainstem respiratory regions, including the preBötC.

3) To identify the functional role of RGS4, a protein that indirectly inhibits Gβγ activity, in regulating respiratory rate depression by MORs.

Hypotheses

1) Microperfusion of gallein, a Gβγ inhibitor, reverses respiratory rate depression induced by micrcoperfusion of the MOR agonist DAMGO into the preBötC.

2) RGS4 mRNA is expressed and co-expressed with the µ-opioid receptor 1 (OPRM1) mRNA in the preBötC, the NTS, and the gigantocellular reticular nucleus, three structures expressing MOR.

3) Microperfusion of the RGS4 inhibitor CCG 50014 into the preBötC potentiates the respiratory effects of the MOR agonist DAMGO perfused to the preBötC.

By inhibiting Gβγ activity in the preBötC in the presence of MOR ligands, we aim to investigate its role in respiratory rate depression by MOR activation. While GIRK channels are known to play a significant role in respiratory rate depression by opioids (Montandon et al., 2016c), they also play significant roles in other opioid signaling pathways, including analgesia (Al-Hasani & Bruchas, 2011). Thus, upstream mechanisms of GIRK channel activation must be targeted to uncover distinct opioid-induced respiratory depression machinery. We first aim to demonstrate direct inhibition of the Gβγ subunit can reverse respiratory rate depression by MOR agonists. We then stimulate Gβγ activity by inhibiting RGS4, an endogenous and prototypical RGS protein that deactivates Gβγ via its GAP activity. We propose that RGS4 can modulate respiratory inhibition by MOR activation, as RGS proteins are known to inhibit Gβγ activity. This finding would demonstrate RGS proteins exhibit functional roles in the opioid-induced respiratory rate depression pathway and suggest RGS proteins can be further investigated to uncover specific subtypes that play a role in respiratory depression by opioids without affecting opioid analgesia.

Chapter 2: Methods

All procedures were performed in accordance with the recommendations of the Canadian Council on Animal Care and were approved by the St. Michael’s Hospital Animal Care Committee. All studies were performed in adult male Wistar rats with body weight between 200- 300g from Charles River Laboratories. Male rats were exclusively used because female rats have been shown to have differential skull anatomy (Hughes et al., 1978). As the stereotaxic coordinates of the preBötC are defined in male Wistar rats only (Paxinos & Watson, 2014), it is increasingly difficult to target the preBötCin females. We solely performed experiments in male Wistar rats because of these sex-specific anatomical differences between males and females, and the lack of clear stereotaxic coordinates for the preBötC in female rats.

Anesthetized Experiments in Rats

To determine the role of various G-proteins on respiratory rate depression by MOR agonists, we used reverse microdialysis to continuously perfuse agonists, antagonists or blockers into the preBötC of male Wistar rats while recording respiratory muscle activity as described previously (Montandon et al., 2011). 187 total rats were anesthetized with isoflurane (3-4%) mixed with 50% oxygen and then tracheostomized. Isoflurane was then maintained at 1.5-2.5%. Body temperature was monitored using a rectal probe (TC-1000 Temperature Controller, CWE Inc., Ardmore, PA, USA), and kept at 37˚C. Two stainless steel needle electrodes were inserted into the tongue to measure genioglossus muscle activity. Two electrodes made of stainless-steel braided wire were inserted inside the abdomen under the right costal diaphragm to measure diaphragm muscle activity.

The rat was placed on a stereotaxic frame (model SAS-4100, ASI Instruments Inc., Warren, MI, USA) in the prone position and ear bars and a mouthpiece were used to stabilize the skull. A microdialysis probe (CX-I-12-01, 200µm diameter, 1 mm length of diffusing membrane, Eicom USA, San Diego, CA, USA) was inserted in the region of the preBötC, but not directly into the preBötC so inspiratory neurons are not damaged by the probe. preBötC coordinates were determined according to previous studies, which used a combination of stereotaxic coordinates and NK-1R expression in the ventrolateral medulla to locate the preBötC (Montandon et al., 2011; Montandon & Horner, 2013; Montandon et al., 2016c). The probe was unilaterally inserted into 26

the brainstem 12.2 mm posterior, 10.5 mm ventral, and 2.0 mm lateral to the bregma. Drugs were continuously perfused at 3µL/min. Genioglossus muscle activity was closely monitored, as previous studies show it decreases by about 30% when the probe is lowered near the preBötC (Montandon et al., 2011; Montandon et al., 2016c), likely due to the probe interfering with premotor neurons that project to the hypoglossal motor nucleus. Diaphragm muscle activity, genioglossus muscle activity, and breathing rate were recorded throughout. The raw signals were amplified, filtered, and recorded using the Spike2 software (Cambridge Electronic Design, Cambridge, UK). The probe was inserted close to the preBötC to allow drugs perfuse to the preBötC in the allotted time frame. To determine whether the probe was located in the vicinity of the preBötC, we determined whether microperfusion of DAMGO (5 µM) decreased respiratory rate by 10% in less than 30 minutes. 30 minutes was used as a strict cutoff interval based upon previous studies (Montandon et al., 2011; Montandon et al., 2016c) and preliminary data. Although increasing the time interval would potentially lead to more successful experiments, it would also cause the drugs to perfuse to areas further than the preBötC, and thus decrease the spatial specificity of this approach. Furthermore, the probe could not be placed directly in or too close to the preBötC, as it would damage preBötC neurons and cause abnormal breathing.

Of the 187 experiments performed in total, 137 did not display decreases in genioglossus muscle and respiratory rate from DAMGO administration due to inaccurate probe placement as determined by post-mortem histological analysis (Supplemental Figure 1). 17 animals died during surgery, either due to improper tracheostomies, isoflurane overdose, diaphragmatic rupture, or probe damage to preBötC neurons (Figure 5). A total of 23 experiments were successful for the RGS4 study and 10 for the gallein study.

Figure 5: Flow chart of rats used in study. Flow chart demonstrating how many rats were used in each experiment.

Microperfusion of Drugs into Discrete Brainstem Regions To manipulate MORs, RGS4, and Gβγ, combinations of drugs were perfused into the preBötC of anesthetized rats. All drugs were purchased from Tocris Bioscience. The MOR agonist DAMGO was microperfused at a concentration of 5 µM which was significant to decrease respiratory rate by 10% or more but not high enough to completely stop breathing (Montandon et al., 2011). CCG 50014 (20 µM) is a selective RGS4 inhibitor. Naloxone (20 µM) is an MOR antagonist. Gallein (1 and 5 mM) is a Gβγ inhibitor. As gallein is insoluble in water, dimethyl sulfoxide (DMSO) was used as a vehicle to dissolve it in artificial cerebrospinal fluid (aCSF). aCSF is perfused for all baseline recordings. The composition of aCSF is as follows: 125 mM

NaCl, 3 mM KCl, 1 mM KH2PO4, 2 mM CaCl2, 1 mM MgSO4, 25 mM NaHCO3, and 30 mM glucose. The pH of the solution was adjusted down to 7.4 by bubbling CO2 into the aCSF.

Determination of Probe Site Location Histology was performed post-mortem to determine the location of the probe. After rats were overdosed with 5% isoflurane, they were trans-cardially perfused with saline solution followed by formalin (10%) to fixate tissue. The brain was removed and stored in formalin (10%), then transferred to 30% sucrose solution and stored in the refrigerator. After 24h in sucrose, brainstem tissue was frozen in a cryostat (Leica Biosystems, Weltzar, Germany) at -20˚, cut coronally into 50 µm thick sections, and mounted onto slides. Hematoxylin and eosin were used

to stain tissue sections. Slides were visualized under 4X magnification of an Olympus upright BX50 light microscope (Olympus, Tokyo, Japan). Using anterior-posterior, dorsal-ventral, and medial-lateral coordinates in standard brain maps (Paxinos & Watson, 2014) as well as anatomical markers such as the NA, facial nucleus, and inferior olive, the stereotaxic coordinates of the probe sites and the preBötC were determined (Montandon et al., 2011; Montandon & Horner, 2013).

Animals with probes placed more than 1.5 mm away from the center of the preBötC, as determined by post-mortem histological analysis, were not considered for analysis (Supplemental Figure 1). In these experiments, drugs could not properly diffuse to the preBötC within the allotted time interval, and thus the animal did not exhibit respiratory rate depression by DAMGO. The distance of 1.5mm was determined using mathematical models estimating drug diffusion through brain tissue. According to this equation (i) DAMGO would diffuse roughly 1-2 mm2 from the probe site if continually perfused for 30 minutes:

퐶 푙푛( ) 푥 = − 퐶푂 (i) √푘푒 퐷∗

where x denotes distance (in mm) of drug perfusion from probe site, C denotes drug concentration at the probe site outside the microdialysis probe, CO denotes drug concentration at a distance x from the perfusion site, ke denotes the efflux half-life of DAMGO, and D* denotes the effective brain coefficient (Wolak & Thorne, 2013). C for DAMGO was equal to 0.5 µM, which is 10% of the concentration inside the probe. CO was set at 0.05 µM, ke and D* were estimated to be 1.9 and

0.85 based on ke and D* values of molecules with similar sizes and chemical properties (Wolak & Thorne, 2013). Using this formula (ii), the distance x was estimated at:

0.05 푙푛( ) 2.3 푥 = − 0.5 = = 1.53mm (ii) 1.9 2.2 √ 0.85

According to this formula, at a distance of 1.03 mm from the probe, the concentration of DAMGO was equal to 0.05 µM or 10x less than the concentration at the probe site. According to the pharmacological properties of DAMGO, a concentration of 0.05 µM or 50 nM is sufficient to have 100% of MORs bind with DAMGO (Raynor et al., 1994).

In vivo Experimental Procedures

(1) We first determined whether Gβγ inhibition could modulate respiratory rate depression by DAMGO. For baseline recordings, we microperfused aCSF for 30 minutes into the preBötC, followed by microperfusion of DAMGO for 30 minutes, and subsequent microperfusion of gallein (1 mM) for 40 minutes concurrently with 5 µM DAMGO. In a separate set of experiments, we investigated if a higher dose of gallein (5 mM) could fully reverse respiratory rate depression. At the end of each experiment, naloxone (20 µM) was perfused to determine the full reversal of respiratory rate depression by DAMGO. As gallein was dissolved in DMSO, an equivalent amount of DMSO, which equated to 87 µL, was added to the naloxone+DAMGO drug combination in the high-dose gallein experiments.

(2) To determine the role of RGS4 in regulating respiratory inhibition by MOR activation, we first microperfused aCSF for baseline recordings, followed by either CCG 50014 in treatment animals or aCSF in controls, each for 40 minutes. We then microperfused either the RGS4 inhibitor CCG 50014 to the preBötC concurrently with 5 µM DAMGO, or 5 µM DAMGO alone, for 30 minutes. In a separate set of experiments, we microperfused aCSF for 30 minutes, followed by either 5 µM DAMGO for controls or 20 µM CCG 50014 concurrently with 5 µM DAMGO for treatment animals, each for 30 minutes. We then followed each experiment with administration of 20 µM naloxone concurrent with DAMGO in controls and DAMGO and CCG 50014 in treatment.

Data Analysis

For each drug condition, respiratory rate, genioglossus muscle amplitude, and diaphragm muscle amplitude were reported as average values over the final 10 minutes of the drug perfusion interval. To normalize data across experiments, respiratory rate, amplitudes of diaphragm muscle activity and genioglossus muscle amplitude were normalized as percentage of baseline activity.

Correlation Maps

To determine the relationship between probe position compared to the preBötC and the functional effect of DAMGO, we related the distance of the probe perfusion site from the center of the preBötC with the latency of DAMGO to depress respiratory rate. The probe perfusion site was determined using post-mortem histological analysis. As the probe dialysis membrane was a 1 mm long section at the probe tip, the point 0.5 mm above the bottom tip of the probe lesion site was identified as the probe perfusion site. The center of the preBötC was determined using a combination of anatomical markers such as the facial nucleus, inferior olive, and nucleus ambiguus from post-mortem histological analysis and stereotaxic coordinates (Paxinos & Watson, 2014). Drug latency was defined as the time for DAMGO administration to decrease respiratory rate by 10%. The reasoning behind constructing these maps is that, if a specific area in the brainstem is responsible for mediating opioids effects on respiratory rate, then the time needed for the drug to diffuse through the brainstem tissue to the site of interest and gradually decrease respiratory rate is dependent on the distance of the probe from the site of interest.

After determination of the probe site relative to the preBötC, as described above using post- mortem histological samples, drug latencies for DAMGO were calculated. Latency was determined as the time, in minutes, necessary for DAMGO administration to decrease respiratory rate by 10%. As respiratory rate can become less stable under anesthesia and lead to transient, short lived changes (Rocco & Zin, 2002), respiratory rate must decrease by 10% of baseline and remain at or below that threshold for 10 minutes to qualify as respiratory rate depression due to drug activity. These criteria have been used previously to define respiratory rate depression in rodents (Montandon et al., 2011; Montandon & Horner, 2013; Montandon et al., 2016c). These values (distances of the probe from the center of the preBötC and drug latencies) were collected for all experiments. A cutoff for the distances from the preBötC to the probe sites larger than 1.5 mm was used and experiments with distances larger than 1.5 mm were excluded. The reasoning is that with distances beyond 1.5 mm, DAMGO was not able to diffuse to the preBötC during the 30 minute time interval, as shown by previous calculations, and thus respiratory rate depression would only be due to degradation of breathing over time in anesthetized experiments, therefore not reflecting the effect of DAMGO diffusion through brainstem tissue.

Correlation maps were then constructed to demonstrate whether DAMGO specifically acts on preBötC neurons or whether it may diffuse to other areas of the brainstem to depress respiratory rate. For every 50 µm area in a 3-dimensional grid spanning from 11.0-13.0 mm caudal the bregma, we calculated the distance of the probe site from each corresponding coordinate and correlated the value with drug latency. Similar to previous studies (Montandon et al., 2011; Montandon & Horner, 2013; Montandon et al., 2016c), we used an exponential fit because it most adequately mimics the spatial diffusion of a drug through neural tissue in 3-dimensional space, along with the observation that distances above 1.5 mm significantly increase the drug latency time. The correlation values (0< r2 <1) for each coordinate were then calculated and plotted onto standard brain maps. Blue corresponded to lower values and red to higher values. The maps and correlation values were generated using customized Matlab scripts (Mathworks, Natick, MA, USA). In other words, these correlation maps highlighted hot-spots indicating areas of the brainstem where DAMGO quickly depresses respiratory rate.

In situ Hybridization To determine the expression of the RGS4 and OPRM1, an RNA-based in situ hybridization assay was used to determine the co-expression patterns of various proteins in brainstem respiratory networks (RNAscope ACD Bio, Newark, CA, USA). The preBötC, NTS, and gigantocellular reticular nucleus all contain neurons that express MORs (Ding et al., 1996; Gray et al., 1999). Rats were euthanized with a 5% isoflurane overdose and transcardially perfused with phosphate buffered saline, followed by formalin, and brains were placed in formalin, 10%, 20%, and then 30% sucrose solutions overnight. Brainstem tissue was frozen in a cryostat at -20˚ (Leica Biosystems, Weltzar, Germany) and cut coronally into 25 µm thick sections. Target retrieval reagent (ACD Bio, Newark, CA, USA) was applied directly to slides. Protease solution (ACD Bio, Newark, CA, USA) was added onto slides, followed by application of the OPRM1 and RGS4 mRNA probe. Slides were then incubated in multiplex fluorescent v2 AMP1 and AMP2 solutions (ACD Bio, Newark, CA, USA) to amplify signal. TSA Cy3 fluorescent dye solution (Perkin Elmer, Waltham, MA, USA) was added for OPRM1 signal, and Cy5 fluorescent dye was added for RGS4 signal. DAPI counterstain was added to sections, slides were mounted with fluorsave (Millipore Sigma, Burlington, MA, USA), coverslipped, and dried overnight. Sections were scanned using

Zeiss Axioscanner (Carl Zeiss AG, Oberkochen, Germany) to visualize staining patterns. Various landmarks, including the nucleus ambiguus, inferior olive, and XII nucleus, along with molecular markers, such as OPRM1 expression, were used to locate the preBötC, NTS, and gigantocellular reticular nucleus, and the brainstem section that most clearly defined the region was used to count cells. Borders of the preBötC, the NTS, and the gigantocellular reticular nucleus were determined according to rat brain with stereotaxic coordinates for Wistar rats (Paxinos & Watson, 2014). Additionally, NK-1R expression was used to further determine the preBötC boundaries (Gray et al., 1999).

The total number of cells inside these areas was determined by counting the number of nuclei stained with DAPI. Expression of RGS4, OPRM1, or both mRNAs in a cell was determined by fluorescent signal being expressed within the area of the cell, as defined by the DAPI-stained nucleus. If fluorescent signaling for both mRNAs were present above threshold levels, the cell was deemed to co-express both RGS4 and OPRM1. Negative controls were used to set a threshold of fluorescent expression, of which the signal had to exceed in order to be counted. Negative control probes targeted the bacterial dapB gene, which is not expressed in rat neural tissue. Positive control probes included POLR2A, PPIB, and UBC genes, which are common housekeeping genes. The number of cells expressing each mRNA are reported as proportions of the number of total DAPI- stained cells.

Statistical Analysis Data was presented as mean values (± standard error of the means). N values for each group were indicated in the figure captions. For studies with gallein, a one-way RM ANOVA was used, with the repeated factor being conditions (aCSF or drugs). For studies with CCG 50014, a two- way RM ANOVA was used, where the repeated factor was condition (baseline, DAMGO, naloxone) and the non-repeated factor intervention (control vs RGS4 inhibitor). Normality was tested with the Shapiro-Wilk test and homogeneity of variances were tested using the equal variance test or Brown-Forsythe test. When tests were statistically significant with P<0.05, Holm- Sidak post-hoc tests were performed to uncover the differences in the two factors. When data was non-normal, one- or two-way RM ANOVAs on ranks were ran. All statistical analyses and figures were performed and constructed using SigmaPlot 14 (Systat Software Inc., San Jose, CA, USA).

Graphs were then further edited with Adobe Illustrator CC (Adobe Inc., San Jose, CA, USA) for improved styling and to create multi-panel figures. MOR signaling diagrams were created using BioRender (BioRender, Toronto, ON, Canada).

Chapter 3: Results

Gallein Administration in the preBötC Reverses Respiratory Rate Depression by MOR Agonists To determine the role of Gβγ proteins in regulating respiratory rate depression by activation of MORs, we microperfused the MOR agonist DAMGO (5 µM), followed by a combination of DAMGO and the Gβγ inhibitor gallein (1 mM, Figure 6) into the preBötC of anesthetized rats. There was a significant effect of drug microperfusion into the preBötC on the respiratory rate compared to baseline when measured as both absolute values (F(4,14)=105.90, P<0.001, 1-way RM

ANOVA) and percentage of baseline (F(2,14)=92.43, P<0.001, 1-way RM ANOVA). There was a significant effect of drug microperfusion into the preBötC on diaphragm amplitude when reported as both absolute values (F(2,14)=7.67, P=0.014, 1-way RM ANOVA) and percentage of baseline

(F(2,14)=8.19, P=0.012, 1-way RM ANOVA). There was no significant effect of drug microperfusion on genioglossus muscle amplitude when measured as absolute values

(F(2,14)=0.066, P=0.937, 1-way RM ANOVA) or percentage of baseline (F(2,14)=0.756, P=0.501, 1-way RM ANOVA). Post-hoc Holm-Sidak tests showed that DAMGO led to a sustained decrease in respiratory rate from 49.6±1.7 breaths/min to 41.9±1.8 breaths/min (P=<0.001, n=5, Figure 7A), equating to a decrease to 84.4±1.2% of baseline (P<0.001, n=5, Figure 7A), while inducing a significant decrease in diaphragm muscle amplitude of 1.7±0.2 mV (P=0.017, n=5, Figure 7B), which equates to a decrease to 89.0±2.9% (P=0.013, n=5, Figure 7B). Gallein partially reversed respiratory rate depression by DAMGO to 44.6±1.8 breaths/min (P=0.033, n=5, Figure 7A), or 89.9±1.0% of baseline (P=0.003, n=5, Figure 7A), but was unable to significantly change diaphragm amplitude (Figure 7B) compared to DAMGO condition when measured as both absolute values (P=0.530, n=5, Figure 7B) or percentage of baseline (P=0.574, n=5, Figure 7B).

Figure 6: In vivo microperfusion of drugs into preBötC to measure respiratory rate depression. Schematic of in vivo experiments demonstrating unilateral insertion of the microdialysis probe into the preBötC of anesthetized rats. Examples of genioglossus and diaphragm muscle tracings, as well as respiratory rate, are shown.

Figure 7: Low-dose gallein administration into preBötC partially reverses respiratory rate depression by MOR agonists in vivo. A) Group mean data (n=5) for the effects of DAMGO (5 µM) and low-dose gallein (1 mM) administration on respiratory rate, diaphragm muscle activity, and genioglossus muscle activity. B) Data of DAMGO (5 µM) and low-dose gallein (1 mM) from section A normalized according to percent of baseline measurement for respiratory rate, diaphragm muscle, and genioglossus muscle activity. Data reported as breaths/min, amplitude, or % of baseline value. Error bars are SEM. Empty circles represent individual data points for each drug condition. * indicates significant difference where p<0.05 by post-hoc Holm-Sidak tests.

In a separate set of experiments, gallein concentration was increased to 5 mM, and naloxone was microperfused after gallein to see if it could significantly change respiratory rate. Statistical analysis uncovered a significant effect of drug microperfusion on respiratory rate when measured as both absolute values (F(4,19)=20.92, P<0.001, 1-way RM ANOVA) and percentage of baseline (F(4,19)=23.19, P<0.001, 1-way RM ANOVA). There was no significant effect of drug microperfusion on diaphragm muscle amplitude when measured as absolute value (F(4,19)=2.99,

P=0.073, 1-way RM ANOVA) or percentage of baseline (F(4,19)=2.96, P=0.075, 1-way RM ANOVA), or on genioglossus muscle amplitude when measured as both absolute values (P=0.058, 1-way RM ANOVA on ranks) and percentage of baseline (P=0.058, 1-way RM ANOVA on ranks). Post-hoc Holm-Sidak tests revealed that a higher concentration of gallein (5 mM) significantly reversed respiratory rate depression by DAMGO from 36.1±2.6 breaths/min to 42.7±2.9 breaths/min (P<0.001, n=5, Figure 8A), equating to an increase from 79.7±1.7% of baseline to 94.4±2.8% of baseline (P<0.001, n=5, Figure 8A). Reversal of respiratory rate depression by gallein was not significantly different from baseline values when measured as absolute values (P=0.066, n=5, Figure 8A) and percentage of baseline (P=0.053, n=5, Figure 8A). As expected (Montandon et al., 2011), naloxone fully reversed respiratory rate depression induced by DAMGO (36.1±2.3 breaths/min to 44.0±1.7 breaths/min, P<0.001, n=5, Figure 8A), equating to a reversal from 79.7±1.4% of baseline to 96.8±2.1% of baseline (P<0.001, n=5, Figure 8A). Importantly, the reversal of respiratory rate by naloxone was not significantly higher than baseline condition when measured as both absolute values (P=0.349, n=5, Figure 8A) and percentage of baseline (P=0.234, n=5, Figure 8A), therefore suggesting that gallein by itself did not increase respiratory rate instead of blocking the inhibition effect of DAMGO.

Figure 8: Administration of gallein (high-dose of 5 mM) into preBötC reverses respiratory rate depression by the MOR agonist DAMGO in vivo. A) Group mean data (n=5) for the effects of DAMGO (5 µM), gallein (5 mM), and naloxone (20 µM) on respiratory rate, diaphragm muscle amplitude, and genioglossus muscle amplitude. B) Data of DAMGO (5 µM), high-dose gallein (5 mM), and naloxone (20 µM) from section A normalized according to percent of baseline measurement for respiratory rate, diaphragm muscle, and genioglossus muscle activity. Data reported as breaths/min, amplitude, or % of baseline value. Error bars are SEM. Empty circles represent individual data points for each drug condition. * indicates significant difference where p<0.05 by post-hoc Holm-Sidak tests.

Medullary region sensitive to DAMGO in gallein experiments To determine whether the location of perfusion site was related to the effect of DAMGO, we correlated the distances from each probe site to the middle of the preBötC with the latencies for DAMGO to depress respiratory rate by 10%. There was a statistically significant, exponential effect of distances of the probe site from the preBötC on the latencies of DAMGO effect (R2=0.69, P<0.001, n=5, Figure 9D). We demonstrated that diffusing DAMGO closer to the preBötC led to faster depression of respiratory rate than DAMGO perfused further away from the preBötC. It is however possible that diffusion of DAMGO to other brainstem sites may also have significant depressive effect on respiratory rate. We therefore calculated coefficient correlations for every possible set of coordinates in the medullary section containing the preBötC, i.e. we applied the same correlation using all possible set of coordinates as reference coordinates. The rationale for these hotspot maps was to confirm that the preBötC was the area of the brainstem most sensitive to respiratory rate depression by DAMGO in the gallein experiments. The data for probe location, distance from preBötC, and drug latency are shown in Table 3.

Figure 9: Correlation between location of the perfusion site and intensity of respiratory rate depression induced by DAMGO in the gallein experiments. A) Post-mortem histological sample showing site of microdialysis probe relative to preBötC. Distance from preBötC, in mm from bregma, (AP: -12.40, DV: -10.00, ML: 2.10) was measured and plotted against drug latency. Inferior olive, nucleus ambiguus, and facial nucleus were used as landmarks to locate the preBötC. B) Drug latency was defined as time after DAMGO administration until respiratory rate decreased by 10%. C) Data points were plotted and an exponential regression was performed to show the correlation between drug latency and distance from the preBötC. D) Correlation maps were then constructed to show the area of the brainstem most sensitive to respiratory rate depression by MOR agonists. Red indicates areas of high correlation, while blue indicates low correlation. The scale indicates R2 values.

Table 3: Data for correlation and correlation map relating probe locations and DAMGO latencies for the gallein experiments. Data used to construct correlation maps for Gβγ inhibition data. Anterior- posterior (AP), dorsal-ventral (DV), and medial-lateral (ML) probe location coordinates were calculated and compared to known preBötC coordinates to find distance, in mm, of probe site from preBötC, and data was compared to drug latency, defined as the time in minutes after DAMGO administration for the respiratory rate to decrease by 10%, to determine correlation values for gallein data.

Experiment Distance from preBötC coordinates Probe coordinates Latency Number preBötC AP DV ML AP DV ML (mm) (min) 1 -12.40 -10.00 2.10 -11.25 1.10 -10.80 1.3 30 2 -12.40 -10.00 2.10 -12.10 1.70 -10.70 0.4 1 3 -12.40 -10.00 2.10 -12.25 2.10 -10.00 1.0 4 4 -12.40 -10.00 2.10 -11.45 2.60 -10.10 1.3 20 5 -12.40 -10.00 2.10 -12.50 2.80 -10.30 1.1 7 6 -12.40 -10.00 2.10 -12.50 2.60 -10.20 1.0 7

RGS4 Expression in the preBötC To determine if RGS4, a prototypical RGS protein, is involved in respiratory rate depression by opioids, we first investigated whether it was present in MOR-expressing cells of the preBötC. To this aim, in situ hybridization assays were run on brainstem tissue samples from male Wistar rats (n=3) to visualize expression profiles of RGS4 and OPRM1. In a representative rat, RGS4 (red) and OPRM1 (green) expressions are shown in the NTS, the preBötC, and the gigantocelluar reticular nucleus (Figure 10B), three cell populations known to express MORs (Gray et al., 1999). We found that, of the brainstem nuclei studied, RGS4 (90.2±5.1% of cells, n=3, Figure 10C) and OPRM1 (84.7±7.0%, n=3, Figure 10C) were most highly expressed in the NTS with a large proportion of these cells co-expressing OPRM1 and RGS4 (82.9±7.3%, n=3, Figure 10C). In the preBötC, RGS4 (81.4±4.4%, n=3, Figure 10C) and OPRM1 (74.8±7.5%, n=3, Figure 10C) expressions were less pronounced compared to NTS, however, a majority of

cells co-expressed RGS4 and OPRM1 (65.3±6.1%, n=3, Figure 10C). Similarly in the gigantocellular reticular nucleus, RGS4 (78.0±8.3%, n=3, Figure 10C) and OPRM1 (72.0±11.6%, n=3, Figure 10C) showed lower expressions than in the NTS, but showed a substantial co- expression as well (61.9±11.9%, n=3, Figure 10C).

Figure 10: RGS4 and OPRM1 are co-expressed in key brainstem respiratory regions. A) Fluorescent brainstem section images of in situ hybridization assay run to visualize brainstem respiratory neuron populations that expressed OPRM1 mRNA (pictured in green). B) Zoomed in images show RGS4 (red) and OPRM1 (green) expression in the NTS, preBötC, and gigantocelluar reticular nucleus (GRN). Yellow areas denote co-expression of both mRNAs. Cell nuclei were stained with DAPI (blue). C) Quantification of RGS4/ OPRM1 expression in NTS, preBötC, and gigantocellular reticular nucleus. Quantification measured in percent of total number of cells (as identified with DAPI) in respiratory nucleus expressing or co-expressing each mRNA.

RGS Inhibition Partially Potentiates Respiratory Rate Depression by Opioid Agonists To determine if RGS4 play a significant role in regulating respiratory rate decrease by DAMGO, two different groups of experiments were conducted (control and CCG50014 groups). CCG50014 is a RGS4 blocker. In the control group (n=6), baseline was recorded while aCSF was perfused for 30 minutes followed by a control time period of 40 min during which only aCSF was perfused. Then DAMGO was applied for 30 minutes. In the CCG50014 group (n=7), baseline was recorded for 30 minutes followed by CCG50014 (20 µM) for 40 minutes. Following CCG50014, DAMGO was perfused for 30 minutes. The use of two groups was necessary to allow CCG50015 to ensure sufficient perfusion of the preBötC, and to determine whether CCG50014 by itself depressed respiratory rate.

There was a statistically significant interaction between RGS4 inhibition and drug microperfusion into the preBötC on the respiratory rate when measured as absolute values

(F(2,38)=3.74, P=0.004, 2-way RM ANOVA), but not percent of baseline (F(2,38)=3.40, P=0.052, 2-way RM ANOVA), as well as on diaphragm muscle amplitude when measured as both absolute values (F(2,38)=4.88, P=0.018, 2-way RM ANOVA) and percent of baseline (F(2,38)=4.99, P=0.016, 2-way RM ANOVA). RGS4 inhibition and drug microperfusion did not display a statistically significant interaction on genioglossus muscle amplitude when measured as absolute value

(F(2,38)=0.91, P=0.418, 2-way RM ANOVA) or percentage of baseline (F(2,38)=0.22, P=0.804, 2- way RM ANOVA).

Post-hoc Holm-Sidak tests showed that RGS4 inhibition did not significantly potentiate respiratory rate depression by DAMGO (P=0.051, Figure 11B). In the control group, DAMGO

alone lowered breathing rate (from 46.8±1.9 to 38.2±3.3 breaths/min, P=0.019, n=6, Figure 11B). In the CCG 50014 group, CCG 50014 and DAMGO significantly potentiated respiratory rate depression from 45.2±1.8 to 27.9±3.7 breaths/min, P<0.001, n=7, Figure 11B). RGS4 inhibition significantly changed diaphragm response to DAMGO when measured as both absolute values (P=0.008, Figure 11B) and percentage of baseline (P=0.002, Figure 11C). Controls showed a statistically significant decrease in diaphragm muscle amplitude from DAMGO when measured as absolute value from 0.55±0.04 to 0.50±0.04 mV (P=0.014, n=6, Figure 11B), but not when measured as percentage of baseline (P=0.034, n=6, Figure 11C). In the CCG50014 group, animals showed a potentiated decrease in diaphragm muscle amplitude from 0.48±0.01 to 0.35±0.03 (P<0.001, n=7, Figure 11B), or to 73.2±6.1% of baseline (P<0.001, n=7, Figure 11C).

Figure 11: Functional role of RGS4 in modulating respiratory rate depression by the MOR agonist DAMGO. A) Representative tracings of respiratory rate measurements in control animals (black) vs intervention animals (red). B) Unilateral perfusion of 20 µM CCG 50014 into the preBötC potentiated respiratory rate depression by 5 µM DAMGO (n=7) as compared to DAMGO alone (n=6). Decrease in genioglossus muscle amplitude by DAMGO administration was similar in both groups. Diaphragm muscle amplitude was also decreased from DAMGO administration in both groups. C) Data was shown in percentage of baseline as well. D) Data was normalized to drug condition and reported as percent of baseline to clearly define the change in respiratory rate from DAMGO administration on both groups. E) Graph showed effect of RGS4 inhibition alone on respiratory rate in the absence of MOR activation. Separate set of experiments were performed that added 20 µM naloxone to each group after DAMGO treatment to investigate if naloxone can fully reverse respiratory rate depression in intervention group (n=4) vs control (n=3). Error bars are SEM. Empty circles represent individual data points for each drug condition. * indicates significant difference with p<0.05 by post-hoc Holm-Sidak tests.

Because administration of CCG50014 alone decreases respiratory rate (45.2±1.5 to 32.3±2.7 breaths/min, P<0.001, n=10, Figure 11E), we normalized the data according to the conditions before DAMGO was perfused to compare the slope in control and CCG50014 groups to more clearly demonstrate the interaction between RGS4 inhibition and DAMGO-induced respiratory rate depression. When data was normalized to conditions before DAMGO, there was no statistically significant interaction between RGS4 inhibition and DAMGO microperfusion on respiratory rate (P=0.915, Figure 11D) or genioglossus amplitude (P<0.689, Figure 11D). However, there was a significant effect of interaction on diaphragm amplitude (P=0.039, Figure 11D).

Post-hoc Holm-Sidak tests revealed that DAMGO administration in control groups produced a non-significant decrease in diaphragm muscle amplitude (P<0.394, n=6, Figure 11D). In the CCG 50014 group, DAMGO administration led to a decrease of diaphragm muscle amplitude to 85.5±4.4% of baseline (P=0.001, n=7, Figure 11D).

In a separate set of experiments, CCG50014 and DAMGO were microperfused into the preBötC followed by naloxone to investigate whether naloxone, an MOR antagonist, can fully reverse respiratory rate depression by DAMGO. There was a statistically significant interaction between RGS4 inhibition and drug microperfusion into the preBötC on respiratory rate when measured as absolute value (F(2,20)=11.74, P=0.002, 2-way RM ANOVA) and percentage of baseline (F(2,20)=12.92, P=0.002, 2-way RM ANOVA). There was no significant interaction between RGS4 inhibition and drug microperfusion into the preBötC on diaphragm amplitude when measured as both absolute value (F(2,20)=11.74, P=0.002, 2-way RM ANOVA) and percentage of baseline (F(2,20)=11.74, P=0.002, 2-way RM ANOVA), or genioglossus muscle when measured as absolute value (F(2,20)=11.74, P=0.002, 2-way RM ANOVA) or percentage of baseline

(F(2,20)=11.74, P=0.002, 2-way RM ANOVA).

RGS4 inhibition again potentiated DAMGO-induced respiratory rate depression when measured as absolute values (P=0.012, Figure 12A) and percentage of baseline (P<0.001, Figure 12B). CCG50014 animals exhibited a decrease in respiratory rate from 49.6±2.6 to 30.9±1.9 breaths/min (P<0.001, n=4, Figure 12A), equating to 62.2±1.4% (P<0.001, n=4, Figure 12B). When only DAMGO was administered, it induced a decrease in respiratory rate from 49.9±2.0 to 43.1±1.6 breaths/min (P=0.004, n=3, Figure 12A), or to 86.0±3.3% (P=0.003, n=3, Figure 12B). RGS4 inhibition also had a significant interaction on naloxone administration when measured as percentage of baseline (P=0.003, Figure 12B) but not absolute values (P=0.078, Figure 12A). Upon administration of naloxone, breathing rate increased back to baseline in the control group (94.8±4.5% of baseline, P=0.032, n=3, Figure 12B) while the CCG50014 group only displayed a partial reversal in respiratory rate (80.0±4.0% of baseline, P<0.001, n=3, Figure 12B).

Figure 12: Respiratory rate depression by MOR agonists in the absence of RGS4 activity is only partially reversed by naloxone. A) Separate set of experiments were performed that added 20 µM naloxone to each group after DAMGO treatment to investigate if naloxone could fully reverse respiratory rate depression in intervention group (n=4) vs control (n=3). Diaphragm and genioglossus muscle activity showed similar trends. B) Data shown in section A is normalized to percent of baseline. Error bars are SEM. Empty circles represent individual data points for each drug condition. * indicates significant difference with p<0.05 by post-hoc Holm-Sidak tests.

To determine if the area of the brainstem the most sensitive to DAMGO for the set of experiments using CCG50014, we created a map that correlated the microperfusion sites with the

effects of DAMGO. We used a correlation relating DAMGO effects and locations of perfusion sites to determine whether microperfusion of DAMGO had the most effect at the level of preBötC as done previously (Montandon et al., 2011; Montandon & Horner, 2013). There was a positive correlation between proximity to the preBötC and drug latency (R2=0.48, P<0.001, n=6, Figure 13B). Data indicating probe locations, distances of perfusion sites from preBötC, and drug latencies are shown in Table 4. The red area on the map indicates high correlation coefficients between locations and drug latencies. This “hotspot” overlapped with the preBötC, indicating that this area is highly sensitive to DAMGO (Figure 13A). Importantly, the highly correlated area did not overlap with other regions of the brainstem with high expression of µ-opioid receptors such as the nucleus tractus solitarius.

Figure 13: Significant correlation between distances from perfusion sites to the preBötC and latency for respiratory rate to decrease by 10% in response to DAMGO in the CCG50014 experiments. A) Correlation between distance of probe from the preBötC (AP: -12.30, DV: -10.00, ML: 2.10) vs. the drug latency (R2=0.48). B) Brainstem section image shows hotspot map, where red indicates a high correlation value and blue indicates low values, and locations of brainstem respiratory neuronal populations shown to contain high MOR expression.

Table 4: Data for correlation map relating probe locations and DAMGO latencies for CCG50014 experiments. Data used to construct correlation maps for RGS4 inhibition data. Anterior-posterior (AP), dorsal-ventral (DV), and medial-lateral (ML) probe location coordinates were calculated and compared to known preBötC coordinates to find distance, in mm, of probe site from preBötC, and data was compared to drug latency, defined as the time in minutes after DAMGO administration for the respiratory rate to decrease by 10%, to determine correlation values for RGS4 experiments.

Experiment Distance from preBötC coordinates Probe coordinates Latency Number preBötC AP DV ML AP DV ML (mm) (min) 1 -12.30 -10.00 2.10 -12.55 -9.50 1.80 1.3 1 2 -12.30 -10.00 2.10 -12.65 -9.50 2.10 1.2 5 3 -12.30 -10.00 2.10 -12.30 -9.90 2.80 0.9 8 4 -12.30 -10.00 2.10 -12.35 -9.50 1.70 1.0 17 5 -12.30 -10.00 2.10 -12.05 -10.85 1.10 1.1 3 6 -12.30 -10.00 2.10 -12.25 -8.60 1.20 1.4 19

Chapter 4: Discussion

Opioid use, both recreationally and clinically as prescription painkillers, is highly prevalent in our society (Volkow & Koroshetz, 2019). Yet despite this, very little is known about the mechanisms that govern respiratory depression, a common and potentially lethal side effect of opioids (Alanmanou, 2006). Thus, understanding the neural and molecular mechanisms that underlie opioid-induced respiratory depression, and how those mechanisms may differ from opioid-induced analgesia, is of critical importance, and could potentially lead to the development of new, safer opioid therapies.

Currently, there is controversy surrounding whether opioid effects on breathing are mediated by G-protein signaling pathways (Kliewer et al., 2019) or β-arrestin recruitment (Raehal et al., 2005), and many still believe that opioid side effects, i.e. those independent of analgesia, are mediated solely by β-arrestin recruitment (Manglik et al., 2016; van Gastel et al., 2018), specifically the side effect of respiratory depression (Yang et al., 2011; Schmid et al., 2017). β- arrestins inhibit MOR signaling by causing receptor desensitization and/or internalization, as shown in Figure 4 (Bohn et al., 2000). Following GPCR kinase-mediated MOR phosphorylation, β-arrestins bind to the intracellular tail of MORs and prevent them from activating the coupled G- protein, even in the continued presence of opioids (Oakley et al., 1999). GIRK channel activation can also be modulated by the activity of β-arrestins (Dascal & Kahanovitch, 2015).

Knockout mice for β-arrestin2 (βarr2-/- ) display altered analgesic responses to opioids (Bohn et al., 1999; Raehal et al., 2005; Kliewer et al., 2019), including significantly reduced respiratory depression as compared to wild-type (Raehal et al., 2005). Furthermore, PZM21, a biased MOR ligand that potently activates MOR-associated G-proteins and displays minimal βarr2 recruitment, initially showed increased analgesic properties compared to morphine without inducing respiratory depression (Manglik et al., 2016). These findings initially suggested βarr2 plays a significant role in opioid-induced respiratory depression. However, recent studies have presented contradictory findings. Knock-in mice with mutated MORs that are unable to recruit βarr2 display similar respiratory depression to opioids as wild-type (Kliewer et al., 2019), suggesting βarr2 activity may not have influence on respiratory depression by opioids. A new study showed that PZM21 presents significant respiratory depression (Hill et al., 2018), suggesting that

respiratory depression is instead mediated by G-protein mechanisms. Currently, it is unclear whether G-protein or β-arrestin signaling mechanisms mediates respiratory depression by opioid drugs. Considering the current debate within the scientific community as to which pathway mediates respiratory depression, a clear understanding of the contribution of G-protein signaling to respiratory depression by opioid analgesics is necessary.

Here we demonstrate that modulation of the G-protein βγ subunit reversed respiratory rate depression by MOR agonists. Although this finding does not disprove a role of β-arrestin recruitment in respiratory rate depression by opioids, the complete reversal of respiratory rate by Gβγ inhibition, along with the previous finding that GIRK channel activation is involved in respiratory depression by opioids (Montandon et al., 2016c), indicates a significant role of Gβγ signaling in opioid-induced respiratory rate depression. We also determined that, while the RGS4 protein was co-expressed with MORs in preBötC neurons, it is only moderately involved in regulating respiratory rate depression by MOR ligands, as the potentiation of respiratory rate depression observed from RGS4 inhibition was significantly less than the potentiation of opioid- induced analgesia by RGS4 inhibition (Yoon et al., 2015) and reward. Our study further lends evidence that respiratory rate depression by opioids is mediated by G-protein mechanisms in the preBötC.

Role of Gβγ Subunit The present study is the first to show that modulation of Gβγ subunit activity in the preBötC modulates respiratory rate depression by MOR ligands in vivo. Here we showed that pharmacological inhibition of Gβγ subunit activity in preBötC neurons in vivo reversed respiratory rate depression by MOR agonist DAMGO. Naloxone, a MOR antagonist, is able to fully reverse respiratory rate depression by DAMGO, but is not a respiratory stimulant itself (Montandon et al., 2011). Naloxone microperfusion directly after perfusion of gallein, a Gβγ inhibitor, did not significantly change respiratory rate, and respiratory rate in both conditions were similar to baseline. Furthermore, intraperitoneal injection of gallein alone does not have a statistically significant effect on respiratory rate compared to saline injection in mice (Hoot et al., 2013). Taken together, this information suggests gallein acts only to reverse respiratory rate depression by DAMGO, and is not a respiratory stimulant itself.

It is possible for respiratory rate to decay over time independent of MOR activation due to prolonged exposure to isoflurane and trauma during surgery, specifically from probe placement into medullary neural tissue. To demonstrate the stability of this experimental procedure over the time frame used in our study, we microperfused naloxone at the end of our experiments. Previous studies have shown that naloxone microperfusion into the preBötC fully reverses respiratory rate depression from DAMGO back to basleine levels, but does not act as a respiratory stimulant itself (Montandon et al., 2011). As naloxone perfusion reversed respiratory rate back to baseline levels in our studies as well (Figure 8), it demonstrated that the respiratory rate was stable throughout our experiments and that respiratory rate depression was due solely to DAMGO activity.

Using correlation maps, we identified the region of the medulla where DAMGO and the Gβγ subunit inhibitor gallein were perfused in these studies (Figure 9). Preliminary data (Supplemental Figure 2) shows co-expression of MORs and NK-1R, a marker of the preBötC region (Gray et al., 1999), in this region as well. We also showed a reduction in genioglossus muscle amplitude with DAMGO microperfusion. This decrease may be due to the proximity of the preBötC to the parahypoglossal premotor neurons (Tan et al., 2010), a brainstem region located caudal, medial, and dorsal to the preBötC which controls hypoglossal motor neurons and genioglossus muscle amplitude (Montandon et al., 2011).

The Gβγ subunit is known to activate various MOR-induced signaling pathways, including inhibiting N-type calcium channels (Law et al., 2000) and activation of GIRK channels (Torrecilla et al., 2002). The role of calcium channel inhibition in respiratory depression by opioids is currently unknown, but our data suggest that it could be of interest to study these channels in the context of respiratory depression by opioids. However, previous research has demonstrated that GIRK channel activation plays a significant role in mediating respiratory depression by MOR activation (Montandon et al., 2016c). Thus, the reversal of respiratory rate depression observed in our studies is most likely due to the Gβγ subunit’s ability to activate GIRK channels, especially since we used similar DAMGO concentration in our studies that it was used in previous studies (Montandon et al., 2011; Montandon et al., 2016c).

MORs may inhibit neuronal activity and may lead to respiratory depression by inhibiting adenylyl cyclase through Gα subunit (Ballanyi et al., 1997). Elevation of cAMP levels via the in vitro application of forskolin onto medullary neurons of the ventral respiratory group reversed

fentanyl-induced respiratory depression (Ballanyi et al., 1997). Although this study showed that adenylyl cyclase activity can modulate fictive breathing in vitro, it has yet to be shown that in vivo modulation of adenylyl cyclase in the preBötC is able to modulate respiratory depression by opioids. While these previous findings implicate molecular pathways that partially overlap with GIRK channel activation in respiratory depression by opioids, the reversal of respiratory rate depression by gallein administration strongly suggests that Gβγ subunits and GIRK channels are involved in MOR-induced suppression of neuronal activity in respiratory networks. Gallein inhibits Gβγ activity by competitively binding to the protein complex’s active site and blocking the binding of other molecules (Surve et al., 2014), which suggests that respiratory inhibition by MORs may not involve Gα signaling activity and adenylyl cyclase inhibition.

Role of RGS4 in Opioid-induced Respiratory Rate Depression Understanding the distinct molecular pathways that underlie respiratory depression by MOR activation is of great importance, as opioids are widely used to manage pain today (Volkow & McLellan, 2016). While RGS proteins, including RGS4, have been shown to modulate opioid analgesia (Yoon et al., 2015), reward (Kim et al., 2018), and tolerance pathways (Stewart et al., 2015), RGS role in respiratory depression has yet to be characterized. Using fluorescent mRNA in situ hybridization assays, we first showed that MOR mRNA is expressed in the gigantocellular reticular nucleus, NTS, and preBötC, the latter two of which play key roles in the control of respiration (Smith et al., 1991; Severson et al., 2003; Zoccal et al., 2014). We then investigated the amount of co-expression of RGS4 with MORs in neurons in these medullary structures. Certain RGS subtypes are highly selective for specific opioid receptor subtypes; for example, RGS12 selectively acts on κ-opioid receptor signaling over other opioid receptor signaling pathways (Gross et al., 2019). As RGS4 is able to modulate the analgesic effects of opioids (Yoon et al., 2015), and analgesia is largely mediated by MOR signaling (Kieffer, 1999), it is therefore plausible that the RGS4 subtype functions in MOR-induced respiratory depression pathway. Indeed, our data show that, in each nucleus, more than half of the cell populations co-expressed both RGS4 and MORs, including about 70% of cells in the preBötC. While these findings do not establish that RGS4 demonstrates selectivity towards MORs, it further suggests that RGS4 in the preBötC may be regulating respiratory depression by MORs.

To demonstrate the role of RGS4 in respiratory depression by MOR activation, we performed unilateral inhibition of RGS4 in the area of the preBötC in vivo. RGS4 inhibition did not significantly potentiate the respiratory rate depression induced by DAMGO. Administration of the RGS4 inhibitor CCG 50014 was able to strongly potentiate the analgesic effects of opioids in rats in vivo, as was able to decrease the response to nociceptive stimuli by greater than 50% (Yoon et al., 2015). In contrast, we found that CCG50014 administration did not significantly potentiate the respiratory effects of opioids, (Figure 11D). Furthermore, the effect of DAMGO on genioglossus muscle activity was not different in the presence of CCG50014 (Figure 11). However, the decrease in diaphragm muscle amplitude observed in the CCG50014+DAMGO condition was greater than DAMGO alone. CCG50014 by itself was able to decrease diaphragm amplitude in the absence of DAMGO (Figure 11C). This suggests that the decrease observed with CCG50014+DAMGO was due to an additive effect of CCG50014 to the decrease of diaphragm muscle amplitude with DAMGO (Figure 11D). As RGS4 inhibition did not potentiate DAMGO- induced respiratory rate depression, and did not significantly change the response of diaphragm and genioglossus muscle amplitude to DAMGO, we conclude that RGS4 inhibition does not significantly affect respiratory rate depression by opioids. While these findings suggest that RGS4 may not be a potential target to reverse or prevent respiratory rate depression, they suggest that RGS4 may be targeted to potentiate analgesia, without altering breathing. If blocking RGS4 greatly enhances the analgesic effects of opioids (Yoon et al., 2015) without significantly changing their effects on respiration, lower doses of opioids may be used to achieve equianalgesic effects without respiratory side-effects. Furthermore, these data did not disqualify other RGS protein subtypes from playing more significant roles in the modulation of respiratory rate depression by opioids (Gaspari et al., 2014; Wang et al., 2017; Gross et al., 2019), and their roles could be explored in future studies.

RGS4 inhibition alone, in the absence of opioids, caused a significant decrease in respiratory rate. RGS4 is endogenously expressed in the preBötC (Figure 10B) and functions as a GAP in other neuronal populations (Soundararajan et al., 2008). RGS4 also inhibits SST signaling in vitro (Huang et al., 1997) and SST modulates respiratory rate in vivo (Montandon et al., 2016b). SST and its cognate receptors are expressed in the preBötC (Stornetta et al., 2003), and are necessary for generating rhythmic breathing (Gray et al., 2001; Tan et al., 2008). SST is known to decrease rhythmic breathing by acting on SST2 receptors (Gray et al., 2010), through GIRK

channels (Montandon et al., 2016b), therefore suggesting that RGS4 inhibition may potentiate this decrease in rhythmic breathing by modulating the effects of endogenous SST on GIRK channels.

G-protein versus β-arrestin Signaling One of the current controversies related to the mechanisms regulating opioid-induced respiratory depression is the prevailing idea that respiratory depression by opioids are mainly regulated by the β-arrestin signaling pathway (Montandon & Slutsky, 2019). This idea was first proposed by the Dr. Laura Bohn in a study showing that β-arrestin-2 knockout (βarr2-/-) mice did not show respiratory depression by opioids and showed increase opioid analgesia (Bohn et al., 1999). Mice lacking βarr2 exhibited dramatically reduced respiratory depression from morphine, with wild-type mice showing as much as a sevenfold increase in respiratory depression compared to βarr2-/- mice (Raehal et al., 2005). This study led to the concept of biased opioid ligands defined as ligands acting preferentially through the G-protein pathway instead of the β-arrestin pathway. Consistent with concept, the MOR agonist PZM21, was developed to activate G-protein mechanisms with minimal β-arrestin recruitment, and therefore exhibited no respiratory depression (Manglik et al., 2016). However, the respiratory measurements performed in these studies only relied on plethysmography recordings where the animal’s behavior and sleep-wake states were not considered. It is, however, well described that sleep-wake states and behaviors can substantially modulate respiratory activity in freely-behaving rodents (Montandon & Horner, 2013). Importantly, arousal states directly impact on MOR inhibition of respiratory networks by DAMGO (Montandon et al., 2011) and on the severity of respiratory depression by the opioid fentanyl (Montandon & Horner, 2019a). These studies suggest that studies quantifying respiratory depression in rodents in vivo should always consider arousal and behavioral states, because states of decreased arousal (i.e. sleep and sedation) can increase the severity of respiratory depression by opioids (Montandon et al., 2016a). A state-dependent respiratory depression could explain the respiratory effects observed with the biased-opioid ligands (Manglik et al., 2016; Hill et al., 2018). Recently, another study showed that PZM21 also induced respiratory depression (Hill et al., 2018), therefore raising questions about the validity of biased agonists to induce analgesia without respiratory depression.

Consistent with our hypothesis that respiratory rate depression is regulated by G-protein signaling pathway rather than β-arrestin, a recent study showed that transgenic mice with mutated MORs that were not able to recruit βarr2 still displayed respiratory depression by opioids (Kliewer et al., 2019). Recently, a study with newly generated βarr2-/- mice showing similar respiratory depression by opioids than wild-type mice (Kliewer et al., 2020). These data further suggest opioid-induced respiratory depression is independent of β-arrestin-2 and refuting previous studies (Raehal et al., 2005).

Taken together there is strong evidence that G-protein mechanisms mediate opioid-induced respiratory depression, as respiratory depression persisted despite a lack of β-arrestin recruitment. To support this theory, previous studies have shown that GIRK channels, which depend on Gβγ (Torrecilla et al., 2002), significantly contribute to respiratory depression by opioids in the preBötC (Montandon et al., 2016c). Here, in agreement with recent studies (Hill et al., 2018; Kliewer et al., 2019; Kliewer et al., 2020), we show that modulation of G-protein signaling directly affects respiratory rate depression by MOR agonists. Our data further support the concept that respiratory rate depression by MOR ligands depends on a G-protein-signaling mechanism rather than internalization and desensitization of MORs involving β-arrestin.

Limitations Our studies present several limitations. First, co-expression profiles of RGS4 and OPRM1 exhibit significant co-expression in respiratory nuclei, but it did not confirm a functional link between these two proteins. In addition, we did not perform in situ hybridization assays on other RGS subtypes, suggesting that RGS4 could show relatively low co-expression with MORs compared to other subtypes. As RGS4 has been shown to play a role in many different GPCR signaling pathways (Cifelli et al., 2008; Madigan et al., 2018; Michaelides et al., 2018), it may have relatively little activity in the MOR pathway, even though it is present in the same cells. Another limitation is that NK-1R expression in preBötC cells that co-expressed OPRM1 and RGS4 was not determined. Because NK-1R expressing cells are preferentially inhibited by opioids in the preBötC (Gray et al., 1999; Montandon et al., 2011), MOR and RGS4 co-expression in those specific cells would need to be assessed to draw more concrete conclusions about RGS4 role in opioid-induced respiratory rate depression.

Furthermore, RGS proteins are known to exhibit functional redundancy (Zhang & Mende, 2011), which means that the inhibition of one RGS subtype could cause other subtypes to increase their activity in order to compensate, leading to compensatory changes in behavior. While the difference seen with RGS4 inhibition was not substantially different than controls, this could be due to the ability of other RGS proteins to compensate for the lack of RGS4 activity. Another method of inhibiting RGS activity is to mutate the Gβγ at the site of RGS binding, so that RGS proteins cannot bind to the Gβγ subunit, functionally silencing its activity (Lan et al., 1998). While this method is an effective way to avoid the issue of functional redundancy and can be done in preBötC cells to avoid disrupting RGS activity in other tissues, it is not specific to RGS4 and it would also prevent the binding of other RGS proteins.

The non-specificity of gallein also presents a limitation to this study. Gallein is a non- specific Gβγ inhibitor that acts on various GPCR-coupled G-proteins. As NK-1R and SST are GPCRs (Gray et al., 1999), expressed on neurons that are present in the preBötC (Stornetta et al., 2003), and known to modulate respiratory rhythm (Gray et al., 1999; Gray et al., 2010), the reversal of respiratory rate seen when perfusing gallein could have been due to its effects on NK- 1R- and/or SST-associated G-proteins. While perfusing naloxone after gallein administration did not significantly change respiratory rate, suggesting gallein alone does not increase respiratory rate, another set of experiments where gallein is perfused alone should be done to determine gallein’s effect on respiratory rate independent of MOR activation.

Another major limitation of this study is the experimental method used to specifically target the preBötC. The probe must be placed close enough to the preBötC for the drug to diffuse to and act on preBötC neurons within the 30-min interval of drug microperfusion, but not directly in the preBötC, as damaging these neurons can lead to abnormal breathing (McKay et al., 2005). Therefore, there is a small area, determined by mathematical models of drug diffusion (Wolak & Thorne, 2013) and previous experiments, that the probe can be placed in to allow for accurate and successful experiments. Perfusing drugs for longer than 30 minutes would also allow drug diffusion beyond the preBötC, thus limiting the specificity of this method. Therefore, there is a very limited window of time and probe location needed to successfully perform these in vivo experiments. Other methods, such as using mice with RGS4 knockouts in Dbx1-expressing preBötC cells, one of the marker of preBötC cells (Hayes et al., 2017) or chemogenetic approaches

involving expression of designer receptors in the preBötC solely activated by exogenous ligands, would lack anatomical specificity. Therefore, the approach used in our studies is the best available method to test the role of Gβγ and RGS proteins in the preBötC under in vivo conditions.

Due to the experimental method’s high degree of difficulty, it was difficult to obtain results from an adequate number of animals even though we performed 187 experiments. This limitation is especially noticeable in genioglossus and diaphragm amplitude data, as the effect of DAMGO on both measurements was not consistent between data sets or with previous studies. For example, genioglossus muscle amplitude did not decrease after DAMGO administration in low-dose gallein experiments (Figure 7A) but did decrease in the high dose gallein experiments (Figure 8A), despite no differences in experimental design. Previous studies reported a decrease in genioglossus muscle amplitude after DAMGO perfusion into the preBötC (Montandon et al., 2011; Montandon et al., 2016c), and the lack of effect seen in the low-dose gallein experiments could be due in part to the large error bars, which themselves are a result of the low power of the study. Similarly, diaphragm muscle amplitude decreased in low-dose gallein experiments (Figure 7A) but not in any other sets of experiments, and previous studies showed no decrease in diaphragm muscle amplitude after DAMGO administration into the preBötC (Montandon et al., 2011; Montandon et al., 2016c).

Lastly, opioid-induced analgesia is mediated by G-protein signaling mechanisms as well (Varga et al., 2005; Schwindinger et al., 2009). Therefore, it would be increasingly difficult to separate G-protein signaling pathways that are specific to analgesia to those specific to respiratory depression. As we are directly targeting the preBötC in our study, we would not expect changes in nociception, as the preBötC is not involved in analgesia. However, using a drug that inhibits all Gβγ activity would not be effective when given systemically to prevent respiratory rate depression because it would also prevent analgesia. Interestingly, previous mice knockout studies have shown that specific G-protein subunit subtypes are involved in analgesia (Schwindinger et al., 2009), suggesting that targeting specific G-protein complex combinations could induce opioid analgesia with decreased respiratory effects, although further studies must be conducted.

Conclusion In conclusion, this study shows: 60

1) Chemical inhibition of Gβγ subunit activity in the preBötC inhibits respiratory rate depression by the MOR ligand DAMGO in rats.

2) While the RGS4 protein has been shown to play a key role in regulating opioid analgesia (Yoon et al., 2015) and reward (Kim et al., 2018), it does not significantly affect opioid-induced respiratory rate depression. This suggests RGS4 could be targeted to enhance opioid analgesia without potentiating respiratory depression, although further studies are needed to investigate this relationship.

3) These findings further support the theory that G-protein mechanisms significantly contribute to opioid-induced respiratory rate depression, as opposed to it solely being mediated by β-arrestin recruitment as previously thought.

Future Studies Uncovering specific second messengers involved in G-protein signaling and regulating opioid-induced respiratory depression is of considerable interest to identify pharmacological targets for drug development. While modulation of the Gβγ subunit showed promise, it may not be directly used in pain management therapies, as Gβγ-mediated GPCR signaling is found all throughout the body in various signaling pathways. However, there are multiple Gα, Gβ, and Gγ subtypes, and different heterotrimers displaying selectivity for different types of GPCRs (Hurowitz et al., 2000). Certain heterotrimeric G-protein complexes are more important in opioid analgesia that other opioid functions (Schwindinger et al., 2009). Finding if specific G-protein heterotrimers preferentially couple to MORs and if they mediate specific opioid pathways, would be of great use in understanding the molecular pathways regulating respiratory depression by opioids.

Future studies could also utilize freely-behaving experiments to further understand RGS and Gβγ inhibition in opioid-induced respiratory depression. Sedation is known to potentiate opioid-induced respiratory depression (Lee et al., 2015; Montandon et al., 2016a), therefore using isoflurane to anesthetize the animal changes the respiratory response to opioids. Freely-behaving experiments in plethysmography chambers would allow us to administer these drugs and measure respiratory response without the use of anesthetics. Furthermore, sleep-wake states are known to affect opioid response, as opioid-induced respiratory depression is potentiated in states of

decreased arousal (i.e. sleep) (Montandon et al., 2011). Freely-behaving experiments can be used to study the effects of RGS and Gβγ inhibition during sleep as well, making it a useful tool for future studies.

Supplemental Figures

Supplemental Figure 1: Coronal brainstem sections of experiments with the probe placed further than 1.5 mm away from the preBötC. A) Probe coordinates are AP:-12.00, DV: -9.00, ML: 2.80. B) Probe coordinates are AP:-12.84, DV: -8.50, ML: 1.30. C) Probe coordinates are AP:-11.90, DV: -10.70, ML: 1.20. D) Probe coordinates are AP:-11.82, DV: -10.60, ML: 1.80. The hypoglossal nucleus (7N), facial nucleus (12N), inferior olive (IO), nucleus ambiguus (NA), and preBötzinger Complex (preBötC) are labelled.

Supplemental Figure 2: Preliminary data showing mRNAs for TAC1R (gene for NK-1R) and OPRM1 (gene for MOR) are co-expressed in preBötC neurons. A) Coronal brainstem (AP:-12.50, DV: -10.5, ML: 2.10) of fluorescent in situ hybridization showing expression of MOR (green) and NK- 1R (red) mRNA. preBötC is identified by high expression of NK-1R and by sing nucleus ambiguus (NA) as landmark. Cell nuclei (blue) stained with DAPI. B) Zoomed in image of NA and preBötC showing NK-1R and OPRM1 expression. N=1.

References

Abdala AP, Rybak IA, Smith JC & Paton JF. (2009). Abdominal expiratory activity in the rat brainstem-spinal cord in situ: patterns, origins and implications for respiratory rhythm generation. The Journal of physiology 587, 3539-3559.

Adriani J. (1953). Combating narcotic-induced respiratory depression in surgery. The Merck report 62, 27-29.

Al-Hasani R & Bruchas MR. (2011). Molecular mechanisms of opioid receptor-dependent signaling and behavior. Anesthesiology 115, 1363-1381.

Alanmanou E. (2006). Opioids. Decisions in Pain Management, 2.

Aliverti A, Cala SJ, Duranti R, Ferrigno G, Kenyon CM, Pedotti A, Scano G, Sliwinski P, Macklem PT & Yan S. (1997). Human respiratory muscle actions and control during exercise. Journal of applied physiology (Bethesda, Md : 1985) 83, 1256-1269.

Anderson TM, Garcia AJ, 3rd, Baertsch NA, Pollak J, Bloom JC, Wei AD, Rai KG & Ramirez JM. (2016). A novel excitatory network for the control of breathing. Nature 536, 76-80.

Anderson TM & Ramirez JM. (2017). Respiratory rhythm generation: triple oscillator hypothesis. F1000Research 6, 139.

Ballanyi K, Lalley PM, Hoch B & Richter DW. (1997). cAMP-dependent reversal of opioid- and prostaglandin-mediated depression of the isolated respiratory network in newborn rats. The Journal of physiology 504 ( Pt 1), 127-134.

Bandler R & Shipley MT. (1994). Columnar organization in the midbrain periaqueductal gray: modules for emotional expression? Trends in neurosciences 17, 379-389.

Barbaro NM, Heinricher MM & Fields HL. (1986). Putative pain modulating neurons in the rostral ventral medulla: reflex-related activity predicts effects of morphine. Brain research 366, 203-210.

Barnett WH, Jenkin SEM, Milsom WK, Paton JFR, Abdala AP, Molkov YI & Zoccal DB. (2018). The Kolliker-Fuse nucleus orchestrates the timing of expiratory abdominal nerve bursting. Journal of neurophysiology 119, 401-412.

Basbaum AI & Fields HL. (1978). Endogenous pain control mechanisms: review and hypothesis. Annals of neurology 4, 451-462.

Bautista TG & Dutschmann M. (2014). Inhibition of the pontine Kolliker-Fuse nucleus abolishes eupneic inspiratory hypoglossal motor discharge in rat. Neuroscience 267, 22-29.

Berecki G, Motin L & Adams DJ. (2016). Voltage-Gated R-Type Calcium Channel Inhibition via Human mu-, delta-, and kappa-opioid Receptors Is Voltage-Independently Mediated by Gbetagamma Protein Subunits. Molecular pharmacology 89, 187-196.

Betty M, Harnish SW, Rhodes KJ & Cockett MI. (1998). Distribution of heterotrimeric G- protein beta and gamma subunits in the rat brain. Neuroscience 85, 475-486.

Bishop GB, Cullinan WE, Curran E & Gutstein HB. (2002). Abused drugs modulate RGS4 mRNA levels in rat brain: comparison between acute drug treatment and a drug challenge after chronic treatment. Neurobiology of disease 10, 334-343.

Bohm A, Gaudet R & Sigler PB. (1997). Structural aspects of heterotrimeric G-protein signaling. Current opinion in biotechnology 8, 480-487.

Bohn LM, Gainetdinov RR, Lin FT, Lefkowitz RJ & Caron MG. (2000). Mu-opioid receptor desensitization by beta-arrestin-2 determines morphine tolerance but not dependence. Nature 408, 720-723.

Bohn LM, Lefkowitz RJ, Gainetdinov RR, Peppel K, Caron MG & Lin FT. (1999). Enhanced morphine analgesia in mice lacking beta-arrestin 2. Science (New York, NY) 286, 2495- 2498.

Bonci A & Williams JT. (1997). Increased probability of GABA release during withdrawal from morphine. The Journal of neuroscience : the official journal of the Society for Neuroscience 17, 796-803.

Bonis JM, Neumueller SE, Krause KL, Pan LG, Hodges MR & Forster HV. (2013). Contributions of the Kolliker-Fuse nucleus to coordination of breathing and swallowing. Respiratory physiology & neurobiology 189, 10-21.

Bourinet E, Soong TW, Stea A & Snutch TP. (1996). Determinants of the G protein-dependent opioid modulation of neuronal calcium channels. Proceedings of the National Academy of Sciences of the United States of America 93, 1486-1491.

Breitwieser GE & Szabo G. (1988). Mechanism of muscarinic receptor-induced K+ channel activation as revealed by hydrolysis-resistant GTP analogues. The Journal of general physiology 91, 469-493.

Brown EN, Purdon PL & Van Dort CJ. (2011). General anesthesia and altered states of arousal: a systems neuroscience analysis. Annual review of neuroscience 34, 601-628.

Brownstein MJ. (1993). A brief history of opiates, opioid peptides, and opioid receptors. Proceedings of the National Academy of Sciences of the United States of America 90, 5391-5393.

Camps M, Carozzi A, Schnabel P, Scheer A, Parker PJ & Gierschik P. (1992). Isozyme-selective stimulation of phospholipase C-beta 2 by G protein beta gamma-subunits. Nature 360, 684-686.

Casey KL, Minoshima S, Morrow TJ & Koeppe RA. (1996). Comparison of human cerebral activation pattern during cutaneous warmth, heat pain, and deep cold pain. Journal of neurophysiology 76, 571-581.

Chou R, Fanciullo GJ, Fine PG, Adler JA, Ballantyne JC, Davies P, Donovan MI, Fishbain DA, Foley KM, Fudin J, Gilson AM, Kelter A, Mauskop A, O'Connor PG, Passik SD, Pasternak GW, Portenoy RK, Rich BA, Roberts RG, Todd KH, Miaskowski C & American Pain Society-American Academy of Pain Medicine Opioids Guidelines P. (2009). Clinical guidelines for the use of chronic opioid therapy in chronic noncancer pain. J Pain 10, 113-130.

Cifelli C, Rose RA, Zhang H, Voigtlaender-Bolz J, Bolz SS, Backx PH & Heximer SP. (2008). RGS4 regulates parasympathetic signaling and heart rate control in the sinoatrial node. Circulation research 103, 527-535.

Collier HO & Roy AC. (1974). Morphine-like drugs inhibit the stimulation of E prostaglandins of cyclic AMP formation by rat brain homogenate. Nature 248, 24-27.

Commons KG, Aicher SA, Kow LM & Pfaff DW. (2000). Presynaptic and postsynaptic relations of mu-opioid receptors to gamma-aminobutyric acid-immunoreactive and medullary- projecting periaqueductal gray neurons. The Journal of comparative neurology 419, 532- 542.

Cygan J, Trunsky M & Corbridge T. (2000). Inhaled heroin-induced status asthmaticus: five cases and a review of the literature. Chest 117, 272-275.

Dahan A, Aarts L & Smith TW. (2010). Incidence, Reversal, and Prevention of Opioid-induced Respiratory Depression. Anesthesiology 112, 226-238.

Dahan A, Yassen A, Bijl H, Romberg R, Sarton E, Teppema L, Olofsen E & Danhof M. (2005). Comparison of the respiratory effects of intravenous buprenorphine and fentanyl in humans and rats. British journal of anaesthesia 94, 825-834.

Dascal N & Kahanovitch U. (2015). The Roles of Gbetagamma and Galpha in Gating and Regulation of GIRK Channels. International review of neurobiology 123, 27-85.

De Vries L & Gist Farquhar M. (1999). RGS proteins: more than just GAPs for heterotrimeric G proteins. Trends in cell biology 9, 138-144.

De Vries L, Zheng B, Fischer T, Elenko E & Farquhar MG. (2000). The regulator of G protein signaling family. Annual review of pharmacology and toxicology 40, 235-271.

Del Negro CA, Funk GD & Feldman JL. (2018). Breathing matters. Nature reviews Neuroscience 19, 351-367.

Derbyshire SW, Jones AK, Gyulai F, Clark S, Townsend D & Firestone LL. (1997). Pain processing during three levels of noxious stimulation produces differential patterns of central activity. Pain 73, 431-445.

Ding YQ, Kaneko T, Nomura S & Mizuno N. (1996). Immunohistochemical localization of mu- opioid receptors in the central nervous system of the rat. The Journal of comparative neurology 367, 375-402.

Dosa PI & Amin EA. (2016). Tactical Approaches to Interconverting GPCR Agonists and Antagonists. Journal of medicinal chemistry 59, 810-840.

Doupnik CA, Davidson N, Lester HA & Kofuji P. (1997). RGS proteins reconstitute the rapid gating kinetics of gbetagamma-activated inwardly rectifying K+ channels. Proceedings of the National Academy of Sciences of the United States of America 94, 10461-10466.

Dutschmann M & Dick TE. (2012). Pontine mechanisms of respiratory control. Comprehensive Physiology 2, 2443-2469.

Feldman JL & Del Negro CA. (2006). Looking for inspiration: new perspectives on respiratory rhythm. NatRevNeurosci 7, 232-242.

Feldman JL, Mitchell GS & Nattie EE. (2003). Breathing: rhythmicity, plasticity, chemosensitivity. Annual review of neuroscience 26, 239-266.

Fields HL. (1999). Central nervous system mechanisms of pain modulation. Textbook of Pain.

Fields HL, Malick A & Burstein R. (1995). Dorsal horn projection targets of ON and OFF cells in the rostral ventromedial medulla. Journal of neurophysiology 74, 1742-1759.

Fields HL, Vanegas H, Hentall ID & Zorman G. (1983). Evidence that disinhibition of brain stem neurones contributes to morphine analgesia. Nature 306, 684-686.

Garnier M, Zaratin PF, Ficalora G, Valente M, Fontanella L, Rhee MH, Blumer KJ & Scheideler MA. (2003). Up-regulation of regulator of G protein signaling 4 expression in a model of neuropathic pain and insensitivity to morphine. The Journal of pharmacology and experimental therapeutics 304, 1299-1306.

Garzon J, Lopez-Fando A & Sanchez-Blazquez P. (2003). The R7 subfamily of RGS proteins assists tachyphylaxis and acute tolerance at mu-opioid receptors. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 28, 1983-1990.

Garzon J, Rodriguez-Diaz M, Lopez-Fando A & Sanchez-Blazquez P. (2001). RGS9 proteins facilitate acute tolerance to mu-opioid effects. The European journal of neuroscience 13, 801-811.

Gaspari S, Papachatzaki MM, Koo JW, Carr FB, Tsimpanouli ME, Stergiou E, Bagot RC, Ferguson D, Mouzon E, Chakravarty S, Deisseroth K, Lobo MK & Zachariou V. (2014). Nucleus accumbens-specific interventions in RGS9-2 activity modulate responses to morphine. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 39, 1968-1977.

Glick C, Evans OB & Parks BR. (1996). Muscle rigidity due to fentanyl infusion in the pediatric patient. Southern medical journal 89, 1119-1120.

Gold SJ, Ni YG, Dohlman HG & Nestler EJ. (1997). Regulators of G-protein signaling (RGS) proteins: region-specific expression of nine subtypes in rat brain. The Journal of neuroscience : the official journal of the Society for Neuroscience 17, 8024-8037.

Gray PA, Hayes JA, Ling GY, Llona I, Tupal S, Picardo MC, Ross SE, Hirata T, Corbin JG, Eugenin J & Del Negro CA. (2010). Developmental origin of preBotzinger complex

respiratory neurons. The Journal of neuroscience : the official journal of the Society for Neuroscience 30, 14883-14895.

Gray PA, Janczewski WA, Mellen N, McCrimmon DR & Feldman JL. (2001). Normal breathing requires preBotzinger complex neurokinin-1 receptor-expressing neurons. Nature neuroscience 4, 927-930.

Gray PA, Rekling JC, Bocchiaro CM & Feldman JL. (1999). Modulation of respiratory frequency by peptidergic input to rhythmogenic neurons in the preBotzinger complex. Science (New York, NY) 286, 1566-1568.

Gross JD, Kaski SW, Schmidt KT, Cogan ES, Boyt KM, Wix K, Schroer AB, McElligott ZA, Siderovski DP & Setola V. (2019). Role of RGS12 in the differential regulation of kappa opioid receptor-dependent signaling and behavior. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 44, 1728-1741.

Gutstein HB, Mansour A, Watson SJ, Akil H & Fields HL. (1998). Mu and kappa opioid receptors in periaqueductal gray and rostral ventromedial medulla. Neuroreport 9, 1777- 1781.

Guyenet PG, Stornetta RL, Bayliss DA & Mulkey DK. (2005). Re: Homing in on the specific phenotype(s) of central respiratory chemoreceptors. Experimental Physiology 90, 266- 268.

Haffajee RL & Frank RG. (2018). Making the Opioid Public Health Emergency Effective. JAMA psychiatry 75, 767-768.

Hagbarth KE & Kerr DI. (1954). Central influences on spinal afferent conduction. Journal of neurophysiology 17, 295-307.

Hajiha M, DuBord MA, Liu H & Horner RL. (2009). Opioid receptor mechanisms at the hypoglossal motor pool and effects on tongue muscle activity in vivo. The Journal of physiology 587, 2677-2692.

Hayes JA, Kottick A, Picardo MCD, Halleran AD, Smith RD, Smith GD, Saha MS & Del Negro CA. (2017). Transcriptome of neonatal preBotzinger complex neurones in Dbx1 reporter mice. Scientific reports 7, 8669.

Heinke B, Gingl E & Sandkuhler J. (2011). Multiple targets of mu-opioid receptor-mediated presynaptic inhibition at primary afferent Adelta- and C-fibers. The Journal of neuroscience : the official journal of the Society for Neuroscience 31, 1313-1322.

Heinricher MM, Tavares I, Leith JL & Lumb BM. (2009). Descending control of nociception: Specificity, recruitment and plasticity. Brain research reviews 60, 214-225.

Helmerhorst GT, Teunis T, Janssen SJ & Ring D. (2017). An epidemic of the use, misuse and overdose of opioids and deaths due to overdose, in the United States and Canada: is Europe next? The bone & joint journal 99-b, 856-864.

Herbert H, Moga MM & Saper CB. (1990). Connections of the parabrachial nucleus with the nucleus of the solitary tract and the medullary reticular formation in the rat. The Journal of comparative neurology 293, 540-580.

Hibino H, Inanobe A, Tanemoto M, Fujita A, Doi K, Kubo T, Hata Y, Takai Y & Kurachi Y. (2000). Anchoring proteins confer G protein sensitivity to an inward-rectifier K(+) channel through the GK domain. The EMBO journal 19, 78-83.

Hill R, Disney A, Conibear A, Sutcliffe K, Dewey W, Husbands S, Bailey C, Kelly E & Henderson G. (2018). The novel mu-opioid receptor agonist PZM21 depresses respiration and induces tolerance to antinociception. British journal of pharmacology 175, 2653- 2661.

Hoot MR, Sypek EI, Reilley KJ, Carey AN, Bidlack JM & McLaughlin JP. (2013). Inhibition of Gβγ-subunit signaling potentiates morphine-induced antinociception but not respiratory depression, constipation, locomotion, and reward. Behavioural pharmacology 24, 144- 152.

Huang C, Hepler JR, Gilman AG & Mumby SM. (1997). Attenuation of Gi- and Gq-mediated signaling by expression of RGS4 or GAIP in mammalian cells. Proceedings of the National Academy of Sciences of the United States of America 94, 6159-6163.

Hughes PC, Tanner JM & Williams JP. (1978). A longitudinal radiographic study of the growth of the rat skull. Journal of anatomy 127, 83-91.

Hurowitz EH, Melnyk JM, Chen YJ, Kouros-Mehr H, Simon MI & Shizuya H. (2000). Genomic characterization of the human heterotrimeric G protein alpha, beta, and gamma subunit genes. DNA research : an international journal for rapid publication of reports on genes and genomes 7, 111-120.

Ingram SL & Williams JT. (1994). Opioid inhibition of Ih via adenylyl cyclase. Neuron 13, 179- 186.

Jamison RN & Mao J. (2015). Opioid Analgesics. Mayo Clinic proceedings 90, 957-968.

Janczewski WA & Feldman JL. (2006). Distinct rhythm generators for inspiration and expiration in the juvenile rat. The Journal of physiology 570, 407-420.

Jiang C & Lipski J. (1990). Extensive monosynaptic inhibition of ventral respiratory group neurons by augmenting neurons in the Botzinger complex in the cat. Experimental brain research 81, 639-648.

Johnson SM, Smith JC & Feldman JL. (1996). Modulation of respiratory rhythm in vitro: role of Gi/o protein-mediated mechanisms. Journal of applied physiology (Bethesda, Md : 1985) 80, 2120-2133.

Jungquist CR, Flannery M, Perlis ML & Grace JT. (2012). Relationship of chronic pain and opioid use with respiratory disturbance during sleep. Pain management nursing : official journal of the American Society of Pain Management Nurses 13, 70-79.

Kaur S & Saper CB. (2019). Neural Circuitry Underlying Waking Up to Hypercapnia. Frontiers in neuroscience 13, 401.

Khan SM, Sleno R, Gora S, Zylbergold P, Laverdure JP, Labbe JC, Miller GJ & Hebert TE. (2013). The expanding roles of Gbetagamma subunits in G protein-coupled receptor signaling and drug action. Pharmacological reviews 65, 545-577.

Kieffer BL. (1999). Opioids: first lessons from knockout mice. Trends in pharmacological sciences 20, 19-26.

Kim J, Lee S, Kang S, Jeon TI, Kang MJ, Lee TH, Kim YS, Kim KS, Im HI & Moon C. (2018). Regulator of G-Protein Signaling 4 (RGS4) Controls Morphine Reward by Glutamate Receptor Activation in the Nucleus Accumbens of Mouse Brain. Molecules and cells 41, 454-464.

Kim KS, Lee KW, Lee KW, Im JY, Yoo JY, Kim SW, Lee JK, Nestler EJ & Han PL. (2006). Adenylyl cyclase type 5 (AC5) is an essential mediator of morphine action. Proceedings of the National Academy of Sciences of the United States of America 103, 3908-3913.

Kliewer A, Gillis A, Hill R, Schmidel F, Bailey C, Kelly E, Henderson G, Christie MJ & Schulz S. (2020). Morphine-induced respiratory depression is independent of beta-arrestin2 signalling. Br J Pharmacol.

Kliewer A, Schmiedel F, Sianati S, Bailey A, Bateman JT, Levitt ES, Williams JT, Christie MJ & Schulz S. (2019). Phosphorylation-deficient G-protein-biased mu-opioid receptors improve analgesia and diminish tolerance but worsen opioid side effects. Nature communications 10, 367.

Kobayashi I, Perry A, Rhymer J, Wuyam B, Hughes P, Murphy K, Innes JA, McIvor J, Cheesman AD & Guz A. (1996). Inspiratory coactivation of the genioglossus enlarges retroglossal space in laryngectomized humans. Journal of applied physiology (Bethesda, Md : 1985) 80, 1595-1604.

Koizumi H, Mosher B, Tariq MF, Zhang R, Koshiya N & Smith JC. (2016). Voltage-Dependent Rhythmogenic Property of Respiratory Pre-Botzinger Complex Glutamatergic, Dbx1- Derived, and Somatostatin-Expressing Neuron Populations Revealed by Graded Optogenetic Inhibition. eNeuro 3.

Kolodny A, Courtwright DT, Hwang CS, Kreiner P, Eadie JL, Clark TW & Alexander GC. (2015). The prescription opioid and heroin crisis: a public health approach to an epidemic of addiction. Annual review of public health 36, 559-574.

Krause KL, Neumueller SE, Marshall BD, Kiner T, Bonis JM, Pan LG, Qian B & Forster HV. (2009). Micro-opioid receptor agonist injections into the presumed pre-Botzinger complex and the surrounding region of awake goats do not alter eupneic breathing. Journal of applied physiology (Bethesda, Md : 1985) 107, 1591-1599.

Kuna ST & Remmers JE. (1999). Premotor input to hypoglossal motoneurons from Kolliker- Fuse neurons in decerebrate cats. Respiration physiology 117, 85-95.

Kuo A, Wyse BD, Meutermans W & Smith MT. (2015). In vivo profiling of seven common opioids for antinociception, constipation and respiratory depression: no two opioids have the same profile. British journal of pharmacology 172, 532-548.

Lalley PM. (2005). D1-dopamine receptor blockade slows respiratory rhythm and enhances opioid-mediated depression. Respiratory physiology & neurobiology 145, 13-22.

Lalley PM. (2008). Opioidergic and dopaminergic modulation of respiration. Respiratory physiology & neurobiology 164, 160-167.

Lan KL, Sarvazyan NA, Taussig R, Mackenzie RG, DiBello PR, Dohlman HG & Neubig RR. (1998). A point mutation in Galphao and Galphai1 blocks interaction with regulator of G protein signaling proteins. The Journal of biological chemistry 273, 12794-12797.

Law PY, Wong YH & Loh HH. (2000). Molecular mechanisms and regulation of opioid receptor signaling. Annual review of pharmacology and toxicology 40, 389-430.

Le Bourdonnec B, Barker WM, Belanger S, Wiant DD, Conway-James NC, Cassel JA, O'Neill TJ, Little PJ, DeHaven RN, DeHaven-Hudkins DL & Dolle RE. (2008). Novel trans-3,4- dimethyl-4-(3-hydroxyphenyl)piperidines as mu opioid receptor antagonists with improved opioid receptor selectivity profiles. Bioorganic & medicinal chemistry letters 18, 2006-2012.

Leal-Pinto E, Gomez-Llorente Y, Sundaram S, Tang QY, Ivanova-Nikolova T, Mahajan R, Baki L, Zhang Z, Chavez J, Ubarretxena-Belandia I & Logothetis DE. (2010). Gating of a G protein-sensitive mammalian Kir3.1 prokaryotic Kir channel chimera in planar lipid bilayers. The Journal of biological chemistry 285, 39790-39800.

Lee LA, Caplan RA, Stephens LS, Posner KL, Terman GW, Voepel-Lewis T & Domino KB. (2015). Postoperative opioid-induced respiratory depression: a closed claims analysis. Anesthesiology 122, 659-665.

Levitt ES, Abdala AP, Paton JF, Bissonnette JM & Williams JT. (2015). mu opioid receptor activation hyperpolarizes respiratory-controlling Kolliker-Fuse neurons and suppresses post-inspiratory drive. The Journal of physiology 593, 4453-4469.

Ling GS, Spiegel K, Lockhart SH & Pasternak GW. (1985). Separation of opioid analgesia from respiratory depression: evidence for different receptor mechanisms. The Journal of pharmacology and experimental therapeutics 232, 149-155.

Lüscher C, Jan LY, Stoffel M, Malenka RC & Nicoll RA. (1997). G protein-coupled inwardly rectifying K+ channels (GIRKs) mediate postsynaptic but not presynaptic transmitter actions in hippocampal neurons. Neuron 19, 687-695.

Luscher C & Slesinger PA. (2010). Emerging roles for G protein-gated inwardly rectifying potassium (GIRK) channels in health and disease. Nature reviews Neuroscience 11, 301- 315.

Machado BH. (2001). Neurotransmission of the cardiovascular reflexes in the nucleus tractus solitarii of awake rats. Annals of the New York Academy of Sciences 940, 179-196.

Macintyre PE, Loadsman JA & Scott DA. (2011). Opioids, ventilation and acute pain management. Anaesthesia and intensive care 39, 545-558.

Madigan LA, Wong GS, Gordon EM, Chen WS, Balenga N, Koziol-White CJ, Panettieri RA, Jr., Levine SJ & Druey KM. (2018). RGS4 Overexpression in Lung Attenuates Airway Hyperresponsiveness in Mice. American journal of respiratory cell and molecular biology 58, 89-98.

Mahmoud S, Yun JK & Ruiz-Velasco V. (2012). Gbeta2 and Gbeta4 participate in the opioid and adrenergic receptor-mediated Ca2+ channel modulation in rat sympathetic neurons. The Journal of physiology 590, 4673-4689.

Majumdar S, Burgman M, Haselton N, Grinnell S, Ocampo J, Pasternak AR & Pasternak GW. (2011). Generation of novel radiolabeled opiates through site-selective iodination. Bioorganic & medicinal chemistry letters 21, 4001-4004.

Manglik A, Lin H, Aryal DK, McCorvy JD, Dengler D, Corder G, Levit A, Kling RC, Bernat V, Hubner H, Huang XP, Sassano MF, Giguere PM, Lober S, Da D, Scherrer G, Kobilka BK, Gmeiner P, Roth BL & Shoichet BK. (2016). Structure-based discovery of opioid analgesics with reduced side effects. Nature 537, 185-190.

Mantyh PW. (1983). Connections of midbrain periaqueductal gray in the monkey. I. Ascending efferent projections. Journal of neurophysiology 49, 567-581.

Mantyh PW & Peschanski M. (1982). Spinal projections from the periaqueductal grey and dorsal raphe in the rat, cat and monkey. Neuroscience 7, 2769-2776.

Mason P. (2001). Contributions of the medullary raphe and ventromedial reticular region to pain modulation and other homeostatic functions. Annual review of neuroscience 24, 737-777.

Mathew SC, Ghosh N, By Y, Berthault A, Virolleaud MA, Carrega L, Chouraqui G, Commeiras L, Condo J, Attolini M, Gaudel-Siri A, Ruf J, Parrain JL, Rodriguez J & Guieu R. (2009). Design, synthesis and biological evaluation of a bivalent micro opiate and adenosine A1 receptor antagonist. Bioorganic & medicinal chemistry letters 19, 6736-6739.

McKay LC, Janczewski WA & Feldman JL. (2005). Sleep-disordered breathing after targeted ablation of preBotzinger complex neurons. Nature neuroscience 8, 1142-1144.

McPherson KB, Leff ER, Li MH, Meurice C, Tai S, Traynor JR & Ingram SL. (2018). Regulators of G-Protein Signaling (RGS) Proteins Promote Receptor Coupling to G- Protein-Coupled Inwardly Rectifying Potassium (GIRK) Channels. The Journal of neuroscience : the official journal of the Society for Neuroscience 38, 8737-8744.

Mellen NM, Janczewski WA, Bocchiaro CM & Feldman JL. (2003). Opioid-induced quantal slowing reveals dual networks for respiratory rhythm generation. Neuron 37, 821-826.

Michaelides M, Miller ML, Egervari G, Primeaux SD, Gomez JL, Ellis RJ, Landry JA, Szutorisz H, Hoffman AF, Lupica CR, Loos RJF, Thanos PK, Bray GA, Neumaier JF, Zachariou V, Wang GJ, Volkow ND & Hurd YL. (2018). Striatal Rgs4 regulates feeding and susceptibility to diet-induced obesity. Molecular psychiatry.

Miller JR, Zuperku EJ, Stuth EAE, Banerjee A, Hopp FA & Stucke AG. (2017). A Subregion of the Parabrachial Nucleus Partially Mediates Respiratory Rate Depression from Intravenous Remifentanil in Young and Adult Rabbits. Anesthesiology 127, 502-514.

Mirshahi T, Robillard L, Zhang H, Hebert TE & Logothetis DE. (2002). Gbeta residues that do not interact with Galpha underlie agonist-independent activity of K+ channels. The Journal of biological chemistry 277, 7348-7355.

Montandon G, Cushing SL, Campbell F, Propst EJ, Horner RL & Narang I. (2016a). Distinct Cortical Signatures Associated with Sedation and Respiratory Rate Depression by Morphine in a Pediatric Population. Anesthesiology 125, 889-903.

Montandon G & Horner R. (2014). CrossTalk proposal: The preBotzinger complex is essential for the respiratory depression following systemic administration of opioid analgesics. The Journal of physiology 592, 1159-1162.

Montandon G & Horner RL. (2013). State-dependent contribution of the hyperpolarization- activated Na+/K+ and persistent Na+ currents to respiratory rhythmogenesis in vivo. J Neurosci 33, 8716-8728.

Montandon G & Horner RL. (2019a). Electrocortical changes associated with sedation and respiratory depression by the analgesic fentanyl. Scientific reports.

Montandon G & Horner RL. (2019b). Electrocortical changes associating sedation and respiratory depression by the opioid analgesic fentanyl. Scientific reports 9, 14122.

Montandon G, Liu H & Horner RL. (2016b). Contribution of the respiratory network to rhythm and motor output revealed by modulation of GIRK channels, somatostatin and neurokinin-1 receptors. Scientific reports 6, 32707.

Montandon G, Qin W, Liu H, Ren J, Greer JJ & Horner RL. (2011). PreBotzinger complex neurokinin-1 receptor-expressing neurons mediate opioid-induced respiratory depression.

The Journal of neuroscience : the official journal of the Society for Neuroscience 31, 1292-1301.

Montandon G, Ren J, Victoria NC, Liu H, Wickman K, Greer JJ & Horner RL. (2016c). G- protein-gated Inwardly Rectifying Potassium Channels Modulate Respiratory Depression by Opioids. Anesthesiology 124, 641-650.

Montandon G & Slutsky AS. (2019). Solving the Opioid Crisis: Respiratory Depression by Opioids as Critical End Point. Chest.

Murakoshi T, Suzue T & Tamai S. (1985). A pharmacological study on respiratory rhythm in the isolated brainstem-spinal cord preparation of the newborn rat. British journal of pharmacology 86, 95-104.

Mustapic S, Radocaj T, Sanchez A, Dogas Z, Stucke AG, Hopp FA, Stuth EA & Zuperku EJ. (2010). Clinically relevant infusion rates of mu-opioid agonist remifentanil cause bradypnea in decerebrate dogs but not via direct effects in the pre-Bötzinger complex region. Journal of neurophysiology 103, 409-418.

Nicoll RA, Alger BE & Jahr CE. (1980). Enkephalin blocks inhibitory pathways in the vertebrate CNS. Nature 287, 22-25.

Oakley RH, Laporte SA, Holt JA, Barak LS & Caron MG. (1999). Association of beta-arrestin with G protein-coupled receptors during clathrin-mediated endocytosis dictates the profile of receptor resensitization. The Journal of biological chemistry 274, 32248-32257.

Onimaru H & Homma I. (2006). Point:Counterpoint: The parafacial respiratory group (pFRG)/pre-Botzinger complex (preBotC) is the primary site of respiratory rhythm generation in the mammal. Point: the PFRG is the primary site of respiratory rhythm generation in the mammal. Journal of applied physiology (Bethesda, Md : 1985) 100, 2094-2095.

Pagliardini S, Janczewski WA, Tan W, Dickson CT, Deisseroth K & Feldman JL. (2011). Active expiration induced by excitation of ventral medulla in adult anesthetized rats. The Journal of neuroscience : the official journal of the Society for Neuroscience 31, 2895- 2905.

Pasternak G & Pan YX. (2011). Mu opioid receptors in pain management. Acta anaesthesiologica Taiwanica : official journal of the Taiwan Society of Anesthesiologists 49, 21-25.

Patil PG, de Leon M, Reed RR, Dubel S, Snutch TP & Yue DT. (1996). Elementary events underlying voltage-dependent G-protein inhibition of N-type calcium channels. Biophysical journal 71, 2509-2521.

Pattinson KT, Governo RJ, MacIntosh BJ, Russell EC, Corfield DR, Tracey I & Wise RG. (2009a). Opioids depress cortical centers responsible for the volitional control of respiration. The Journal of neuroscience : the official journal of the Society for Neuroscience 29, 8177-8186.

Pattinson KT, Mitsis GD, Harvey AK, Jbabdi S, Dirckx S, Mayhew SD, Rogers R, Tracey I & Wise RG. (2009b). Determination of the human brainstem respiratory control network and its cortical connections in vivo using functional and structural imaging. NeuroImage 44, 295-305.

Paxinos G & Watson C. (2014). The rat brain in stereotaxic coordinates: seventh edition. Elsevier.

Perret D & Luo ZD. (2009). Targeting voltage-gated calcium channels for neuropathic pain management. Neurotherapeutics : the journal of the American Society for Experimental NeuroTherapeutics 6, 679-692.

Phillips RS, Cleary DR, Nalwalk JW, Arttamangkul S, Hough LB & Heinricher MM. (2012). Pain-facilitating medullary neurons contribute to opioid-induced respiratory depression. Journal of neurophysiology 108, 2393-2404.

Poulain R, Horvath D, Bonnet B, Eckhoff C, Chapelain B, Bodinier MC & Déprez B. (2001). From hit to lead. Combining two complementary methods for focused library design. Application to mu opiate ligands. Journal of medicinal chemistry 44, 3378-3390.

Prkic I, Mustapic S, Radocaj T, Stucke AG, Stuth EA, Hopp FA, Dean C & Zuperku EJ. (2012). Pontine mu-opioid receptors mediate bradypnea caused by intravenous remifentanil infusions at clinically relevant concentrations in dogs. J Neurophysiol 108, 2430-2441.

Psifogeorgou K, Terzi D, Papachatzaki MM, Varidaki A, Ferguson D, Gold SJ & Zachariou V. (2011). A unique role of RGS9-2 in the striatum as a positive or negative regulator of opiate analgesia. The Journal of neuroscience : the official journal of the Society for Neuroscience 31, 5617-5624.

Raehal KM, Walker JK & Bohn LM. (2005). Morphine side effects in beta-arrestin 2 knockout mice. The Journal of pharmacology and experimental therapeutics 314, 1195-1201.

Raynor K, Kong H, Chen Y, Yasuda K, Yu L, Bell GI & Reisine T. (1994). Pharmacological characterization of the cloned kappa-, delta-, and mu-opioid receptors. Molecular pharmacology 45, 330-334.

Renn CL & Dorsey SG. (2005). The physiology and processing of pain: a review. AACN clinical issues 16, 277-290; quiz 413-275.

Richter DW & Spyer KM. (2001). Studying rhythmogenesis of breathing: comparison of in vivo and in vitro models. Trends in neurosciences 24, 464-472.

Rocco PRM & Zin WA. (2002). Respiratory Effects of Anesthesia, pp. 51-66. Springer Milan, Milano.

Rossi GC, Pasternak GW & Bodnar RJ. (1994). Mu and delta opioid synergy between the periaqueductal gray and the rostro-ventral medulla. Brain research 665, 85-93.

Rudd RA, Seth P, David F & Scholl L. (2016). Increases in Drug and Opioid-Involved Overdose Deaths - United States, 2010-2015. MMWR Morbidity and mortality weekly report 65, 1445-1452.

Sadja R, Alagem N & Reuveny E. (2002). Graded contribution of the Gbeta gamma binding domains to GIRK channel activation. Proceedings of the National Academy of Sciences of the United States of America 99, 10783-10788.

Saitoh O, Masuho I, Terakawa I, Nomoto S, Asano T & Kubo Y. (2001). Regulator of G protein signaling 8 (RGS8) requires its NH2 terminus for subcellular localization and acute desensitization of G protein-gated K+ channels. The Journal of biological chemistry 276, 5052-5058.

Santiago TV & Edelman NH. (1985). Opioids and breathing. Journal of applied physiology (Bethesda, Md : 1985) 59, 1675-1685.

Schmid CL, Kennedy NM, Ross NC, Lovell KM, Yue Z, Morgenweck J, Cameron MD, Bannister TD & Bohn LM. (2017). Bias Factor and Therapeutic Window Correlate to Predict Safer Opioid Analgesics. Cell 171, 1165-1175.e1113.

Schwindinger WF, Borrell BM, Waldman LC & Robishaw JD. (2009). Mice lacking the G protein gamma3-subunit show resistance to opioids and diet induced obesity. American journal of physiology Regulatory, integrative and comparative physiology 297, R1494- 1502.

Severson CA, Wang W, Pieribone VA, Dohle CI & Richerson GB. (2003). Midbrain serotonergic neurons are central pH chemoreceptors. Nature neuroscience 6, 1139-1140.

Shreder K, Zhang L, Dang T, Yaksh TL, Umeno H, DeHaven R, Daubert J & Goodman M. (1998). Synthesis and biological activity of a novel methylamine-bridged enkephalin analogue (MABE): a new route to cyclic peptides and peptidomimetics. Journal of medicinal chemistry 41, 2631-2635.

Sicuteri F, Raino L & Geppetti P. (1983). Substance P and endogenous opioids: how and where they could play a role in cluster headache. Cephalalgia : an international journal of headache 3 Suppl 1, 143-145.

Smith JC, Abdala AP, Rybak IA & Paton JF. (2009). Structural and functional architecture of respiratory networks in the mammalian brainstem. Philosophical transactions of the Royal Society of London Series B, Biological sciences 364, 2577-2587.

Smith JC, Ellenberger HH, Ballanyi K, Richter DW & Feldman JL. (1991). Pre-Botzinger complex: a brainstem region that may generate respiratory rhythm in mammals. Science (New York, NY) 254, 726-729.

Soundararajan M, Willard FS, Kimple AJ, Turnbull AP, Ball LJ, Schoch GA, Gileadi C, Fedorov OY, Dowler EF, Higman VA, Hutsell SQ, Sundstrom M, Doyle DA & Siderovski DP. (2008). Structural diversity in the RGS domain and its interaction with heterotrimeric G protein alpha-subunits. Proceedings of the National Academy of Sciences of the United States of America 105, 6457-6462.

Steeds CE. (2009). The anatomy and physiology of pain. Surgery - Oxford International Edition 27, 507-511.

Stein C. (2018). New concepts in opioid analgesia. Expert opinion on investigational drugs 27, 765-775.

Stewart A, Maity B & Fisher RA. (2015). Two for the Price of One: G Protein-Dependent and - Independent Functions of RGS6 In Vivo. Progress in molecular biology and translational science 133, 123-151.

Stornetta RL, Rosin DL, Wang H, Sevigny CP, Weston MC & Guyenet PG. (2003). A group of glutamatergic interneurons expressing high levels of both neurokinin-1 receptors and somatostatin identifies the region of the pre-Botzinger complex. The Journal of comparative neurology 455, 499-512.

Stucke AG, Miller JR, Prkic I, Zuperku EJ, Hopp FA & Stuth EA. (2015). Opioid-induced Respiratory Depression Is Only Partially Mediated by the preBötzinger Complex in Young and Adult Rabbits In Vivo. Anesthesiology 122, 1288-1298.

Stucke AG, Zuperku EJ, Sanchez A, Tonkovic-Capin M, Tonkovic-Capin V, Mustapic S & Stuth EA. (2008). Opioid receptors on bulbospinal respiratory neurons are not activated during neuronal depression by clinically relevant opioid concentrations. JNeurophysiol 100, 2878-2888.

Subramanian HH, Balnave RJ & Holstege G. (2008). The midbrain periaqueductal gray control of respiration. The Journal of neuroscience : the official journal of the Society for Neuroscience 28, 12274-12283.

Surve CR, Lehmann D & Smrcka AV. (2014). A chemical biology approach demonstrates G protein betagamma subunits are sufficient to mediate directional neutrophil chemotaxis. The Journal of biological chemistry 289, 17791-17801.

Sutton LP, Ostrovskaya O, Dao M, Xie K, Orlandi C, Smith R, Wee S & Martemyanov KA. (2016). Regulator of G-Protein Signaling 7 Regulates Reward Behavior by Controlling Opioid Signaling in the Striatum. Biological psychiatry 80, 235-245.

Sweetnam PM, Caldwell L, Lancaster J, Bauer C, Jr., McMillan B, Kinnier WJ & Price CH. (1993). The role of receptor binding in drug discovery. Journal of natural products 56, 441-455.

Tan W, Janczewski WA, Yang P, Shao XM, Callaway EM & Feldman JL. (2008). Silencing preBotzinger complex somatostatin-expressing neurons induces persistent apnea in awake rat. Nature neuroscience 11, 538-540.

Tan W, Pagliardini S, Yang P, Janczewski WA & Feldman JL. (2010). Projections of preBotzinger complex neurons in adult rats. The Journal of comparative neurology 518, 1862-1878.

Torrecilla M, Marker CL, Cintora SC, Stoffel M, Williams JT & Wickman K. (2002). G-protein- gated potassium channels containing Kir3.2 and Kir3.3 subunits mediate the acute inhibitory effects of opioids on locus ceruleus neurons. The Journal of neuroscience : the official journal of the Society for Neuroscience 22, 4328-4334.

Trescot AM, Datta S, Lee M & Hansen H. (2008). Opioid pharmacology. Pain physician 11, S133-153.

Tupal S, Rieger MA, Ling GY, Park TJ, Dougherty JD, Goodchild AK & Gray PA. (2014). Testing the role of preBotzinger Complex somatostatin neurons in respiratory and vocal behaviors. The European journal of neuroscience 40, 3067-3077. van den Hoogen RH & Colpaert FC. (1986). Respiratory effects of morphine in awake unrestrained rats. The Journal of pharmacology and experimental therapeutics 237, 252- 259. van Gastel J, Hendrickx JO, Leysen H, Santos-Otte P, Luttrell LM, Martin B & Maudsley S. (2018). beta-Arrestin Based Receptor Signaling Paradigms: Potential Therapeutic Targets for Complex Age-Related Disorders. Frontiers in pharmacology 9, 1369.

Vann NC, Pham FD, Hayes JA, Kottick A & Del Negro CA. (2016). Transient Suppression of Dbx1 PreBotzinger Interneurons Disrupts Breathing in Adult Mice. PloS one 11, e0162418.

Varga AG, Reid BT, Kieffer BL & Levitt ES. (2019). Differential impact of two critical respiratory centres in opioid-induced respiratory depression in awake mice. The Journal of physiology.

Varga EV, Hosohata K, Borys D, Navratilova E, Nylen A, Vanderah TW, Porreca F, Roeske WR & Yamamura HI. (2005). Antinociception depends on the presence of G protein gamma2-subunits in brain. European journal of pharmacology 508, 93-98.

Vaughan CW, Ingram SL, Connor MA & Christie MJ. (1997). How opioids inhibit GABA- mediated neurotransmission. Nature 390, 611-614.

Vecchietti V, Clarke GD, Colle R, Dondio G, Giardina G, Petrone G & Sbacchi M. (1992). Substituted 1-(aminomethyl)-2-(arylacetyl)-1,2,3,4-tetrahydroisoquinolines: a novel class of very potent antinociceptive agents with varying degrees of selectivity for kappa and mu opioid receptors. Journal of medicinal chemistry 35, 2970-2978.

Volkow ND & Koroshetz WJ. (2019). The role of neurologists in tackling the opioid epidemic. Nature reviews Neurology 15, 301-305.

Volkow ND & McLellan AT. (2016). Opioid Abuse in Chronic Pain--Misconceptions and Mitigation Strategies. The New England journal of medicine 374, 1253-1263. von Leupoldt A & Dahme B. (2005). Cortical substrates for the perception of dyspnea. Chest 128, 345-354.

Waldhoer M, Bartlett SE & Whistler JL. (2004). Opioid receptors. Annual review of biochemistry 73, 953-990.

Wang D & Teichtahl H. (2007). Opioids, sleep architecture and sleep-disordered breathing. Sleep medicine reviews 11, 35-46.

Wang H, Xu H, Wu LJ, Kim SS, Chen T, Koga K, Descalzi G, Gong B, Vadakkan KI, Zhang X, Kaang BK & Zhuo M. (2011). Identification of an adenylyl cyclase inhibitor for treating neuropathic and inflammatory pain. Science translational medicine 3, 65ra63.

Wang HB, Zhao B, Zhong YQ, Li KC, Li ZY, Wang Q, Lu YJ, Zhang ZN, He SQ, Zheng HC, Wu SX, Hokfelt TG, Bao L & Zhang X. (2010). Coexpression of delta- and mu-opioid receptors in nociceptive sensory neurons. Proceedings of the National Academy of Sciences of the United States of America 107, 13117-13122.

Wang Q, Pronin AN, Levay K, Almaca J, Fornoni A, Caicedo A & Slepak VZ. (2017). Regulator of G-protein signaling Gbeta5-R7 is a crucial activator of muscarinic M3 receptor-stimulated insulin secretion. FASEB journal : official publication of the Federation of American Societies for Experimental Biology.

Wedegaertner PB, Wilson PT & Bourne HR. (1995). Lipid modifications of trimeric G proteins. The Journal of biological chemistry 270, 503-506.

Wei XY, Zhao Y, Wong-Riley MT, Ju G & Liu YY. (2012). Synaptic relationship between somatostatin- and neurokinin-1 receptor-immunoreactive neurons in the pre-Botzinger complex of rats. Journal of neurochemistry 122, 923-933.

Weil JV, McCullough RE, Kline JS & Sodal IE. (1975). Diminished ventilatory response to hypoxia and hypercapnia after morphine in normal man. The New England journal of medicine 292, 1103-1106.

Willis WD, Jr. (1985). The pain system. The neural basis of nociceptive transmission in the mammalian nervous system. Pain and headache 8, 1-346.

Willis WD, Jr. (1988). Anatomy and physiology of descending control of nociceptive responses of dorsal horn neurons: comprehensive review. Progress in brain research 77, 1-29.

Wolak DJ & Thorne RG. (2013). Diffusion of macromolecules in the brain: implications for drug delivery. Molecular pharmaceutics 10, 1492-1504.

Xu X, Zeng W, Popov S, Berman DM, Davignon I, Yu K, Yowe D, Offermanns S, Muallem S & Wilkie TM. (1999). RGS proteins determine signaling specificity of Gq-coupled receptors. The Journal of biological chemistry 274, 3549-3556.

Yang CF & Feldman JL. (2018). Efferent projections of excitatory and inhibitory preBotzinger Complex neurons. The Journal of comparative neurology 526, 1389-1402.

Yang CH, Huang HW, Chen KH, Chen YS, Sheen-Chen SM & Lin CR. (2011). Antinociceptive potentiation and attenuation of tolerance by intrathecal beta-arrestin 2 small interfering RNA in rats. British journal of anaesthesia 107, 774-781.

Yoon SY, Woo J, Park JO, Choi EJ, Shin HS, Roh DH & Kim KS. (2015). Intrathecal RGS4 inhibitor, CCG50014, reduces nociceptive responses and enhances opioid-mediated analgesic effects in the mouse formalin test. Anesthesia and analgesia 120, 671-677.

Yoshikawa DM, Hatwar M & Smrcka AV. (2000). G protein beta 5 subunit interactions with alpha subunits and effectors. Biochemistry 39, 11340-11347.

Young-McCaughan S & Miaskowski C. (2001). Definition of and mechanism for opioid-induced sedation. Pain management nursing : official journal of the American Society of Pain Management Nurses 2, 84-97.

Yu VC, Eiger S, Duan DS, Lameh J & Sadee W. (1990). Regulation of cyclic AMP by the mu- opioid receptor in human neuroblastoma SH-SY5Y cells. Journal of neurochemistry 55, 1390-1396.

Zhang P & Mende U. (2011). Regulators of G-protein signaling in the heart and their potential as therapeutic targets. Circulation research 109, 320-333.

Zhang Z, Xu F, Zhang C & Liang X. (2007). Activation of opioid mu receptors in caudal medullary raphe region inhibits the ventilatory response to hypercapnia in anesthetized rats. Anesthesiology 107, 288-297.

Zhang Z, Zhuang J, Zhang C & Xu F. (2011). Activation of opioid mu-receptors in the commissural subdivision of the nucleus tractus solitarius abolishes the ventilatory response to hypoxia in anesthetized rats. Anesthesiology 115, 353-363.

Zhu W & Pan ZZ. (2005). Mu-opioid-mediated inhibition of glutamate synaptic transmission in rat central amygdala neurons. Neuroscience 133, 97-103.

Zoccal DB, Furuya WI, Bassi M, Colombari DS & Colombari E. (2014). The nucleus of the solitary tract and the coordination of respiratory and sympathetic activities. Frontiers in physiology 5, 238.