Ablation of neuromedin B (NMB)‐expressing neurons located within retrotrapezoid nucleus (RTN) reduces the central respiratory chemoreflex (CRC) selectively in conscious rats

Abstract

RTN senses PCO2 and regulates breathing in a CO2‐dependent manner but the full extent of RTN's contribution to the CRC is not known. Lesions produced by microinjecting a substance‐P analog conjugated with saporin (SSP‐SAP) elicit small and/or transient CRC reductions in rats. Interpretative limitations are many: development of compensatory mechanisms, especially of carotid body origin, difficulty to identify the critical neurons and to lesion them sufficiently. As shown recently, neuromedin B (Nmb) seems to be a selective marker of the Phox2b+/Vglut2+ RTN neurons that contribute to the CRC. In the present study, we microinjected saline or SSPSAP (3 injections per side; ng/injection: 0.6, 1.2 or 2.4 ng) to destroy RTN in ~250 g Sprague‐Dawley male rats. Physiological experiments were conducted three weeks later in unanesthetized animals, following which we counted the surviving Nmb+ neurons using fluorescent ISH (FISH) in a 1/6 series of sections. SSP‐lesioned rats were regrouped post hoc into two clusters according to the number of surviving RTN Nmb+ neurons: mild lesion group (ML rats; N= 13; > 55 Nmb+ cells) and large lesion group (LL rats; N=11; 4–55 Nmb+ cells remaining). Control rats (C rats; N= 11) had 324 ± 48 Nmb+ neurons (these and all subsequent values represent mean ± SD). Under normoxia, arterial pH was virtually the same in all groups (LL rats: 7.46 ± 0.03; ML: 7.48 ± 0.01; C: 7.49 ± 0.01). PaO2 was lower in LL rats than in C and ML animals (71 ± 7 vs. 80 ± 6 vs. 78 ± 6 mmHg) but PaCO2 was higher (49 ± 7 vs. 38 ± 3 vs. 39 ± 5 mmHg). BP measured via arterial catheter was the same in all groups. Under normoxia, LL rats had a reduced VT relative to C and ML rats during quiet waking (0.39 ± 0.1 vs. 0.49 ± 0.07 vs. 0.54 ± 0.1 mL/100g), SW sleep (0.32 ± 0.08 vs.0.45 ± 0.06 vs. 0.50 ± 0.1 mL/100g) and REM sleep (0.25 ± 0.06 vs. 0.35±0.05 vs. 0.36 ± 0.1 mL/100g). Short‐term hyperoxia (1 min of 65% FiO2, balance N2) caused a larger decrease in VE in LL and in ML rats compared to C (ΔVE: −11 ± 3 vs. −13 ± 6 vs. −8 ± 3 mL/100g/min). Of note, the hypoventilation still present after 20 min of continuous hyperoxia and was larger in LL than in ML and C rats (ΔVE: −9 ± 6 vs. −5 ± 10 vs. 0.9 ± 4 mL/100g/min). When exposed to 6 % FiCO2 in 65% FiO2 LL rats had a greatly reduced breathing stimulation (hypercapnic ventilatory reflex, HCVR) compared to C and ML rats (ΔVE: 17.8 ± 9 vs. 63.6 ± 12 vs. 49.5 ± 21 mL/100g/min). The HCVR to 9 % FiCO2 in 65% FiO2 was reduced in similar proportion (ΔVE: 25.6 ± 9 vs. 85.9 ± 14 vs. 76.3 ± 27 mL/100g/min). Hypoxia (10% FiO2, balance N2) activated VE similarly in all three groups (ΔVE: 34.1 ± 6 vs. 27.2 ± 8 vs. 28.2 ± 9 in mL/100g/min for LL, C and ML rats respectively) although ΔVT was enhanced and ΔFR reduced in LL rats. In conclusion, >80% RTN Nmb+ neurons need to be destroyed to produce a notable reduction of the HCVR. Large RTN lesions reduce the CRC by ≥ 71% without decreasing the hypoxic ventilatory reflex. Large RTN lesions raise the arterial PCO2 homeostatic set‐point by ≥10 mmHg and lower PaO2 by ~9 mmHg but have little influence on long‐term arterial pH. Finally, the loss of breathing stimulation contributed by RTN is largely compensated by a permanent state of hypoxia and increased carotid body activity.Support or Funding InformationNational Institutes of Health (Grants RO1 HL074011 and RO1 HL 028785 to P.G.G.)This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.