Abbreviated Title : Heme Oxygenase and Hypoglossal Motor Neurons 4 5

Obstructive sleep apnea (OSA) is characterized by sporadic collapse of the upper airway leading to periodic disruptions in breathing. Upper airway patency governed by genioglossal nerve activity originates from the hypoglossal motor nucleus. Mice with targeted deletion of the gene Hmox2, encoding the carbon monoxide (CO) producing enzyme, heme oxygenase-2 (HO-2), exhibit severe OSA, yet the contribution of central HO-2 dysregulation to the phenomenon is unknown. Using the rhythmic brainstem slice preparation, which contains the preBötzinger complex (preBötC) and the hypoglossal nucleus, we tested the hypothesis that central HO-2 dysregulation weakens hypoglossal motoneuron output. Disrupting HO-2 activity increased transmission failure as determined by the intermittent inability of the preBötC rhythm to trigger output from the hypoglossal nucleus. Failed transmission was associated with a reduced input-output relationship between the preBötC and the motor nucleus. These network phenomena were related to smaller inspiratory drive currents and reduced intrinsic excitability among hypoglossal neurons. In addition to HO-2, hypoglossal neurons also expressed the CO-regulated H2S producing enzyme cystathionine □-lyase (CSE). H2S abundance was higher in hypoglossal neurons of HO-2 null mice than wild-type controls. Disrupting CSE function normalized transmission in HO-2 null mice and an H2S donor mimicked the effects of HO-2 dysregulation. These findings demonstrate a hitherto uncharacterized modulation of hypoglossal activity through the interaction of HO-2 and CSE-derived H2S, and supports the perspective that centrally derived HO-2 activity plays an important role regulating upper airway control.

neurons 23 and/or the direct modulation of intrinsic excitability of hypoglossal neurons [24][25][26] . It, 78 however, is unknown whether impaired HO-2 signaling in the hypoglossal nucleus influences 79 synaptic and/or intrinsic neuronal properties to alter output from the motor nucleus that may 80 ultimately contribute to upper airway obstruction. 81 We tested this possibility using a combination of electrophysiological, genetic, and 82 pharmacological approaches in the rhythmic medullary brainstem slice preparation. We found 83 that dysregulated HO-2 activity in the hypoglossal nucleus acts through CSE-dependent H 2 S 84 signaling to reduce motor neuron excitability. This in turn, diminishes the input-output 85 relationship between the preBötC and hypoglossal nucleus, and increases the likelihood of 86 transmission failure between the premotor rhythm and motor nucleus output.
These 87 observations indicate that hypoglossal HO-2 / CO and CSE / H 2 S activities interact as important 88 modulators of hypoglossal output that potentially contributes to changed upper airway tone 89 when dysregulated. 90 with glass suction pipettes filled with aCSF 31 . The recorded signal was sampled at 5kHz, 161 amplified 10,000X, with a lowpass filter of 10 kHz using an A-M instruments (A-M Systems, 162 Sequim, WA, USA) extracellular amplifier. The signal was then rectified and integrated using 163 Clampfit electronic filter. Recordings were stored on a computer for posthoc analysis. 164 All intracellular recordings were made using the Multiclamp 700B (Molecular Devices: 165 https://www.moleculardevices.com/systems/conventional-patch-clamp/multiclamp-700b-166 microelectrode-amplifier). Acquisition and post hoc analyses were performed using the Axon 167 pCLAMP10 software suite (Molecular Devices: https: www.moleculardevices.com/system/axon-168 conventional-patch-clamp/pclamp-11-software-suite). from the caudal surface of the slice were recorded. The liquid junction potential was calculated 178 to be -12mV and was subtracted from the membrane potential. The series resistance was 179 compensated and corrected throughout each experiment. In voltage clamp experiments, 180 membrane potential was held at -60mV. Current clamp experiments used a holding potential 181 between 0 and -100pA to establish the baseline resting membrane potential between -55 and -182 70mV. In some cases, we determined rheobases using a ramp protocol in our current clamp 183 recordings. This ramp protocol consisted of a hyperpolarizing step (-100pA) succeeded by the 184 injection of a ramping depolarizing current (122pA/sec; peak current 600pA). 185 Statistical Analyses. Unless otherwise explicitly stated elsewhere, each n value represents an 186 individual animal that served as a biological replicate for a given measurement. included in the experiment. Each row represents sequential cycles from a single slice 198 experiment. As the rhythmic frequency across preparations varied, the number of events (i.e., 199 cycle number) in the 120 s analysis window also varied; therefore, either the total number of 200 cycles or 25 consecutive cycles from in a given analysis window were plotted. 201 Statistics were performed using Origin 8 Pro (OriginLab, RRID:SCR_014212) or Prism 6 202 (GraphPad Software; RRID:SCR_015807). In cases where the distribution of data appeared 203 normal, comparisons between two groups were conducted using either paired or unpaired two-204 tailed t-tests as appropriate. In cases, where the distribution of individual data points did not 205 appear normal, the Wilcoxon match-paired signed rank test was performed.
A one-way 206 ANOVA was performed followed by posthoc Dunnett's test comparing experimental groups to 207 control when a comparison of three or more groups. In plots where the mean ± S.E.M. are 208 presented, the mean and S.E.M. are overlaid on the individual results from the corresponding 209 dataset. Differences were defined when the P-value was less than 0.05. 210

Disrupting HO-2 function impairs hypoglossal inspiratory activity. Extracellular field 216
recordings in the rhythmic brainstem slice were simultaneously recorded from the preBötC and 217 the corresponding motor output from the hypoglossal nucleus. Two approaches were employed 218 to assess the role of HO-2: (1) using Cr(III) Mesoporphyrin IX chloride (ChrMP459, 20µM), an 219 inhibitor of HO 32 ; and (2) using brain slices from HO-2 null mice. 220 Representative extracellular field recordings from the preBötC and hypoglossal nucleus prior 221 and during ChrMP459 exposure were shown in Fig 1B (n=11). While ChrMP459 suppressed 222 hypoglossal burst amplitude (Fig 1C, left; n=11; Baseline: 99.57 ± 0.60%, ChrM459: 81.80 ± 223 10.80%, P=0.007), the HO inhibitor had no effect on preBötC burst amplitude (Fig 1C,  reduced the cycle-to-cycle input-output relationship between preBötC and the motor nucleus as 226 revealed by examining the cycle-to-cycle I/O across preparations (Fig 1D, Baseline: 1.00 ± 0.07 227 vs. ChrMP459: 0.55 ± 0.11; P=.006). In the extreme, altered input-output relationships between 228 preBötC and the hypoglossal nucleus may increase the propensity for transmission failures as 229 determined by the inability of preBötC activity to produce hypoglossal output at the network 230 level 31 . Indeed, the reduced I/O ratio was associated with an increase in failed transmissions of 231 the preBötC activity to output from the hypoglossal nucleus (Fig 1E,  ChrMP459 is a pan HO inhibitor; however, it cannot distinguish activities between heme 236 oxygenase isoforms. To assess the specific contribution of HO-2, we compared rhythmic 237 activities in brain slices from wild type (n=9) and HO-2 null (n=7) mice (Fig 2A). Larger cycle-to-238 cycle I/O ratios were observed in wild type slices as compared to HO-2 null slices (Fig 2B: wild 239 type: 0.99 ± 0.04 vs. HO-2 null: 0.76 ± 0.11, P=0.045). Similarly, transmission of preBötC 240 activity to the hypoglossal nucleus was greater in wild type than in HO-2 null slices (Fig 2 C, 241 wild type: 96.53 ± 1.76% vs. HO-2 null: 62.55 ± 7.93%, P=0.0006). These findings established 242 that genetic elimination of HO-2 produces a similar phenomenon to that of pan HO inhibition 243 indicating that loss of HO-2 activity alone is sufficient for impairing transmission from the 244 preBötC to the hypoglossal nucleus. Given these similarities and the limited availability of HO-2 245 null mice, several of the following studies were performed using the ChrMP459 in rhythmic wild 246 type brainstem slices. 247

HO inhibitor does not affect premotor neuron activity. Intermediary premotor neurons relay 248
drive from the preBötC to the hypoglossal nucleus 29,30 . Therefore, it was possible that HO 249 inhibition impaired transmission of drive from the preBötC by perturbing activity from 250 intermediary premotor neurons. To address this possibility, triple extracellular recordings (n=5) 251 were made from the preBötC, the field of the ipsilateral premotor neurons, and the hypoglossal 252 nucleus. Baseline transmission from the preBötC to the premotor field and to the hypoglossal 253 nucleus was reliable and consistent (Fig. 3A, middle panel). However, ChrMP459 disrupted 254 activity in the hypoglossal nucleus despite unaltered activities in either the preBötC or the 255 intermediate premotor field (Fig 3A, right panel). Indeed, while neither transmission failures nor 256 the cycle-to-cycle I/O ratio from the preBötC to the premotor field was affected by ChrMP459 257 (Fig 3B: left; I/O: Baseline: 1.16 ± 0.09 vs ChrMP 1.16 ± 0.14, P=0.31; right; Transmission: 258 Baseline: 100.0 ± 0.0% vs ChrMP 86.35 ± 11.81%, P=0.312), the HO inhibitor reduced the 259 transmission of activity and the cycle-to-cycle I/O ratio between the premotor field and the 260 hypoglossal nucleus (Fig 3C:

Elevated levels of H 2 S are observed in the hypoglossal nucleus of HO-2 null mice.
We 282 next sought to determine the mechanism(s) by which inhibition of HO-2 affect hypoglossal 283 neuron activity. Earlier studies 33,34 have reported that HO-2 is a negative regulator of CSE-284 dependent H2S production. To test this possibility, we first examined whether the hypoglossal 285 neurons express CSE. Brainstem sections from the wild type hypoglossal nucleus revealed 286 CSE hypoglossal tissue punches from wild type and HO-2 null mice. Relative to the wild type 287 hypoglossal nucleus, H 2 S is expressed in ChAT-positive hypoglossal neurons (Fig 5A, n=3). 288 We then determined H 2 S abundance in hypoglossal tissue punches from wild type and HO-2 289 null mice. Relative to the wild type hypoglossal nucleus ( Fig 5B blue; n=6, 60.58 ± 290 6.37 nmol • mg -1 • h -1 ), H 2 S abundance was greater in the hypoglossal nucleus of HO-2 null 291 mice ( Fig 5B red; n=6, 144.12 ± 8.29 nmol • mg -1 • h -1 ), but not different from the inferior olive 292 brainstem region of HO-2 null mice ( Fig 5B grey; n=4, 56.10 ± 2.88 nmol • mg -1 • h -1 ). These 293 findings suggest that the hypoglossal nucleus expresses CSE and HO-2 negatively regulates 294 H 2 S production in the hypoglossal nucleus. 295

function. 297
If the impaired transmission of inspiratory drive to the hypoglossal nucleus by disrupted HO-2 298 function involves CSE-derived H 2 S then: 1) a H 2 S donor should mimic the effects of disrupted 299 HO-2 activity; 2) CO administration should improve the input-output relationship in respiratory 300 slices from HO-2 null mice and ChrM459 application; and 3) CSE blockade should restore the 301 transmission from the preBötC to the hypoglossal nucleus. The following experiments tested 302 these possibilities. 303 Wild type brainstem slices exhibited a nearly 1:1 ratio of transmission of inspiratory activity from 304 preBötC to hypoglossal neurons (Fig 5C, left). Application of NaHS, a H 2 S donor reduced 305 transmission from preBötC to hypoglossal (Fig. 5C, middle, right) in a dose-dependent manner 306 ( Fig. 5D; 0μM NaHS: n=9, 100 ± 0.73%; 10μM NaHS: n=5, 90.14 ± 6.08%; 50μM NaHS: n=6, 307 84.0 ± 3.91%; 100μM NaHS: n=9, 81.2 ± 5.83%), which coincided with a reduction in I/O ratio 308 by NaHS ( Fig. 5E; 0 μ M NaHS: 1.055 ± 0.028; 10μM NaHS: 0.85 ± 0.09; 50μM NaHS: 0.82 ± 309 0.08; 100μM NaHS: 0.84 ± 0.07). These findings demonstrated increased H 2 S abundance 310 reduces hypoglossal neuronal activity consistent with findings using either ChrM459 or HO-2 311 null mice. Thus, the stability of inspiratory transmission from the preBötC to hypoglossal nucleus 312 appears to be negatively affected either by increasing H 2 S abundance or by disrupting HO-2 313 function. 314 CO produced by HO-2 is known to inhibit CSE-dependent H 2 S production by HO-2 33,34 . 315 Therefore, we sought to assess how CORM-3 (20μM), a pharmacological CO donor, impacted 316 activity in ChrMP459-treated wild type rhythmic slices (n= 3, Fig 6A)  93.39 ± 2.93%, P=0.036). As these findings suggested the absence of HO-2 dependent CO 322 production is a key factor driving transmission failure in the rhythmic slice preparation, we next 323 determined the involvement of CSE. 324 Inspiratory activity in the brainstem slice from HO-2:CSE double null mice appeared to be stable 325 and consistent (Fig 6D). Quantification of simultaneous extracellular field recordings of preBötC 326 activity and hypoglossal nucleus revealed a larger I/O ratio (Fig 6E, HO- Given the observations using HO-2:CSE null mice, we next sought to determine whether acute 330 blockade of CSE could restore transmission relationships between the preBötC and the 331 hypoglossal nucleus in the HO-2 null slice. In vivo L-PAG treatment improved transmission of 332 preBötC activity to the hypoglossal nucleus in the rhythmic slice (Fig. 7A, n=6) as indicated by 333 larger cycle-to-cycle I/O ratios (Fig. 7B, L-PAG = 1.01± 0.03, P=0.008) and greater transmission 334 rates (Fig. 7C, L-PAG 96.31 ± 2.62%, P<0.0001) when compared to the respective metrics from 335 untreated HO-2 null mice. Intermittent transmission failure was also evident in patch clamp 336 recordings from untreated HO-2 null hypoglossal neurons (Fig 7D, left shaded cycles) but not in 337 HO-2 null hypoglossal neurons treated with L-PAG (Fig. 7D, right). These reduced transmission 338 events correlated with smaller individual inspiratory drive currents in HO-2 null hypoglossal 339 neurons when compared to corresponding inspiratory drive currents from L-PAG treated HO-2 340 mice ( Fig. 7D-E FIG 8A1), in others, apamin modestly increased the drive current (<100pA; FIG 8A2). 360 Despite this variability, apamin increased inspiratory drive currents received by ChrMP459 361 treated hypoglossal neurons (FIG 8A3, n= 6 To determine how blockade of K ATP impacted hypoglossal activity during ChrMP459, we used 368 the K ATP channel blocker, tolbutamide (100μM). Tolbutamide did not induce ectopic bursting in 369 the hypoglossal nucleus during ChrMP459 (Supplemental Fig 2B, n=5). In HO-2 null mouse, the incidence of OSA is absent with co-inhibition of CSE 17 , which is 425 consistent with reports that CO generated by HO-2 inhibits CSE-dependent H 2 S production 33,34 . In conclusion, our study provides proof-of-concept for the existence of a central 456 mechanism by which loss of HO-2 leads to reduced upper airway tone by enhancing H 2 S 457 activity in the hypoglossal nucleus. This mechanism appears to involve antagonistic interactions 458 between HO-2 and CSE activities to regulate excitability of hypoglossal neurons and is localized 459 in the neurons themselves. Although OSA in HO-2 null mice has attributed to increased "loop-460 gain" arising from the hypersensitive carotid body reflex 16 , our findings indicate the potential 461 involvement of a disrupted interaction between HO-2 / CO and CSE / H 2 S in hypoglossal motor 462 neurons that contribute could to the sporadic loss of upper airway tone observed in OSA. 463