32%, compared with WT before SP, n = 6, ∗p < 0.05) but not Atoh1Phox2bCKO preparations (16.55% ± 5.60%, compared with WT before SP, n = 5, ∗p < 0.05). These data suggest that Atoh1 is important for the RTN neurons to modulate inspiratory frequency, and the RTN neurons are a critical component

of the neonatal rhythmogenic network. The lifelong attenuated ventilatory response check details to hypercapnia is a major contributing factor to fatal apnea in CCHS patients. Such a chemosensory defect might be caused by functional impairments in Phox2b-dependent structures such as the carotid body and RTN neurons (Amiel et al., 2003; Dubreuil et al., 2008). Carotid bodies are the peripheral chemoreceptors that sense the arterial partial pressure of oxygen (pO2) and, to a lesser extent, carbon dioxide (pCO2), along with changes in pH. Although the cellular identities of

central CO2 chemoreceptors remain elusive, it has been shown that the RTN neurons are activated by low pH and can increase ventilation upon sensing high pCO2 (Abbott et al., 2009; Mulkey et al., 2004). The en bloc brainstem preparation responds to lower pH at early embryonic stages, allowing us to test the integrity of embryonic chemosensory network when the RTN is the only affected population. We first recorded inspiratory activities using E16.5 WT Apoptosis inhibitor and Atoh1Phox2bCKO embryos under baseline pH (7.4), and then perfused the brainstems using artificial cerebrospinal fluid (aCSF) with a lower pH (7.2). The Atoh1Phox2bCKO embryos show a slower baseline behavior when compared with WT (Atoh1Phox2bCKO: 58.43% ± 2.24%, n = 11, versus WT: 100% ± 7.14%, n = 7, p < 0.001), consistent with a role for RTN in modulating embryonic inspiratory rhythmogenicity. Interestingly, both WT and Atoh1Phox2bCKO preparations are sensitive to lower pH (Atoh1Phox2bCKO: 227.32% ± 4.99% versus WT: 251.00% ± 5.31%, both compared to baseline WT, p <

0.001) ( Figure 5C). These results indicate that the chemosensory circuits of the Atoh1Phox2bCKO embryos are still capable only to detect pH change at early embryonic stage and is distinct from the effects of RTN deletion in the Egr-2 lineages by expressing Phox2b27Ala ( Ramanantsoa et al., 2011). Normally, the RTN neurons are located at the marginal layer of ventral brainstem where blood vessels deliver CO2 signals (Lazarenko et al., 2009). To determine whether RTN mislocalization affects CO2 detection in free-moving adult animals, we utilized unrestrained whole body plethysmography to monitor the respiration of 3-month-old Atoh1Phox2bCKO survivor mice. The breathing parameters of Atoh1Phox2bCKO mice (n = 9) and their littermates (WT, n = 21) were indistinguishable at rest, but when challenged with hypercapnia (5% CO2), WT mice showed increased respiratory frequency (RF, 340.92 ± 3.79 min−1), tidal volume (VT, 13.73 ± 0.30 μl ⋅ g−1), and minute ventilation (VE, 4.68 ± 0.13 ml ⋅ min−1 ⋅ g−1) ( Figure 6A).