1. Neuroscience

Phox2b mutation mediated by Atoh1 expression impaired respiratory rhythm and ventilatory responses to hypoxia and hypercapnia

  1. Caroline B Ferreira
  2. Talita M Silva
  3. Phelipe E Silva
  4. Claudio L Castro
  5. Catherine Czeisler
  6. José J Otero
  7. Ana C Takakura  Is a corresponding author
  8. Thiago S Moreira  Is a corresponding author
  1. Department of Pharmacology, Instituto de Ciencias Biomedicas, Universidade de Sao Paulo, Brazil
  2. Department of Physiology and Biophysics, Instituto de Ciencias Biomedicas, University of Sao Paulo, Brazil
  3. Department of Pathology, College of Medicine, Ohio State University, United States
https://doi.org/10.7554/eLife.73130
Share this article
Cite this article
  1. Caroline B Ferreira
  2. Talita M Silva
  3. Phelipe E Silva
  4. Claudio L Castro
  5. Catherine Czeisler
  6. José J Otero
  7. Ana C Takakura
  8. Thiago S Moreira
(2022)
Phox2b mutation mediated by Atoh1 expression impaired respiratory rhythm and ventilatory responses to hypoxia and hypercapnia
eLife 11:e73130.
https://doi.org/10.7554/eLife.73130
  2. Download BibTeX
  3. Download .RIS

Abstract

Mutations in the transcription factor Phox2b cause congenital central hypoventilation syndrome (CCHS). The syndrome is characterized by hypoventilation and inability to regulate breathing to maintain adequate O2 and CO2 levels. The mechanism by which CCHS impact respiratory control is incompletely understood, and even less is known about the impact of the non-polyalanine repeat expansion mutations (NPARM) form. Our goal was to investigate the extent by which NPARM Phox2b mutation affect (a) respiratory rhythm; (b) ventilatory responses to hypercapnia (HCVR) and hypoxia (HVR); and (c) number of chemosensitive neurons in mice. We used a transgenic mouse line carrying a conditional Phox2bΔ8 mutation (same found in humans with NPARM CCHS). We crossed them with Atoh1cre mice to introduce mutation in regions involved with respiratory function and central chemoreflex control. Ventilation was measured by plethysmograph during neonatal and adult life. In room air, mutation in neonates and adult did not greatly impact basal ventilation. However, Phox2bΔ8, Atoh1cre increased breath irregularity in adults. The HVR and HCVR were impaired in neonates. The HVR, but not HCVR, was still partially compromised in adults. The mutation reduced the number of Phox2b+/TH--expressing neurons as well as the number of fos-activated cells within the ventral parafacial region (also named retrotrapezoid nucleus [RTN] region) induced by hypercapnia. Our data indicates that Phox2bΔ8 mutation in Atoh1-expressing cells impaired RTN neurons, as well as chemoreflex under hypoxia and hypercapnia specially early in life. This study provided new evidence for mechanisms related to NPARM form of CCHS neuropathology.

Editor's evaluation

The present manuscript by Ferreira and colleagues is of potential interest to researchers working in the field of neural control of breathing and associated respiratory disorders. This study provides some novel insight into some genetic lesions that may underlie some developmental respiratory pathophysiologies.

https://doi.org/10.7554/eLife.73130.sa0

Introduction

Breathing is an essential physiological function soon after birth because it can rapidly regulate O2 and CO2 levels in the blood for the rest of our life. Oxygen is mainly sensed by peripheral chemoreceptors, while CO2 is regulated by central chemoreceptors, and to a lesser extent by peripheral chemoreceptors (Smith et al., 2006; Nattie, 2011; Guyenet, 2014; Guyenet and Bayliss, 2015; Guyenet et al., 2019).

Hypoventilation and inability to increase breathing under low oxygen and high CO2 levels are one of the most impacting symptoms in patients with congenital central hypoventilation syndrome (CCHS). The paired like homeobox 2B (Phox2b) mutations are well known to be involved in the development of CCHS (Weese-Mayer et al., 1993; Amiel et al., 2003). CCHS-related Phox2b mutations occur in two major categories: a trinucleotide, polyalanine repeat expansion mutations (PARM) and a non-polyalanine repeat expansion mutations (NPARM), which includes missense, nonsense, and frameshift mutations (Patwari et al., 2010; Ramanantsoa and Gallego, 2013; Moreira et al., 2016). Phox2b NPARM deletions within exon 3 are correlated with severe CCHS phenotype with complete apnea, profound hypoventilation during sleep, and/or cause of post-neonatal infant mortality (Amiel et al., 2003; Weese-Mayer et al., 2010).

The mechanism by which CCHS impact respiratory control is incompletely understood. Thus, investigating how Phox2b mutation in specific neuronal population could contribute to better understand the clinical respiratory outcomes in CCHS. In a rodent experimental model, Phox2b PARM mutation is specific to retrotrapezoid nucleus (RTN), a well-known region involved with central chemoreflex control, impaired respiratory control, and ventilatory response to hypercapnia in neonates (Ramanantsoa et al., 2011). In contrast, hypoxic ventilatory responses are intact and potentiated (Ramanantsoa et al., 2011). Additionally, genetic deletion of Phox2b from atonal homolog 1 (Atoh1)-expressing cells, that include not only RTN neurons (peri VII region) but also neurons located in the intertrigeminal region (peri V region), also abolished ventilatory response to hypercapnia in neonates (Ruffault et al., 2015). The effect seems to be dependent on neuronal loss of ventral aspect of the parafacial region, also named RTN. Although peri V neurons might also be affected, resection of this region did not impact respiratory response to low pH levels in a brainstem preparation (Ruffault et al., 2015), suggesting that this region is not involved with central chemoreflex. However, the extent to which NPARM in regions that are involved with respiratory control and chemosensitivity remains an open question.

Recently, a human CCHS case postmortem proband was found and the mutation predictably causes a frameshift and a hypomorph protein (Phox2bΔ8) (Di Lascio et al., 2018). The present mutation was used to generate a conditional transgenic mouse line that can be activated by cre recombinase and introduce the humanized NPARM Phox2bΔ8 mutation during different developmental phases and regions (Nobuta et al., 2015). Expression of NPARM Phox2bΔ8 mutation in the ventral visceral motor neuron domain (non-respiratory domain) induced apnea in newborns, loss of visceral motor neurons and Phox2b neurons in the RTN, and pre-Bötzinger complex dysfunction (Alzate-Correa et al., 2021). Thus, in the present study, we proposed to investigate the effect of NPARM Phox2bΔ8 mutation in regions that are directly involved with respiratory control and central chemoreflex. To achieve this we used Atoh1cre line as a promoter. Atoh1 is expressed during development in proliferating cells in the rhombic lip and in postmitotic neurons. In this independent site, postmitotic neurons are the only region that co-express Phox2b and Atoh1 surrounding the paramotor neurons that involves facial motor nucleus (thus peri VII, parafacial/RTN neurons) and trigeminal motor nucleus (peri V). We proposed to investigate the effect of NPARM Phox2bΔ8 mutation in these regions on respiratory function, ventilatory chemoreflex to hypoxia and hypercapnia during neonatal and adulthood. In addition, we proposed to determine the effect of this mutation in the development of Phox2b chemosensitive neurons in the parafacial/RTN region. Our hypothesis is that NPARM Phox2bΔ8 mutation in Atoh1-expressing cells impairs respiratory control, ventilatory responses to hypoxia and hypercapnia, and parafacial/RTN chemosensitive neurons.

We found that NPARM Phox2bΔ8 in Atoh1-expressing cells suppressed breathing activity in response to hypoxia and hypercapnia in neonates. Surprisingly, it did not mainly affect baseline ventilation. We also showed that adult mutant mice increased irregular breathing pattern and the ventilatory response to hypoxia was partially compromised. While ventilatory response to hypercapnia completely recovered. Additionally, anatomical data showed reduced Phox2b+/tyrosine hydroxylase (TH)- immunoreactivity and fos+/TH--activated neurons by hypercapnia in the parafacial/RTN region. Together, our findings imply that NPARM Phox2bΔ8 in Atoh1-expressing cells affects the development of the parafacial/RTN chemosensitive neurons, and consequently impaired breathing under hypoxic and hypercapnic conditions especially in neonates. These data provided new evidence for mechanisms related to CCHS neuropathology.

Results

Functional respiratory changes observed in NPARM Phox2bΔ8 in Atoh1cre-expressing cells

In the first set of experiment, we investigated whether a conditional mutation of Phox2bΔ8 in Atoh1-expressing cells affects ventilation during neonatal and adult phase. Given that all Phox2bΔ8, Atoh1Cre mice survived, respiratory parameters were examined between 1–3 and 30–45 post-natal days.

Body weight during neonatal phase was not different between mutation vs. control littermates (2.2±0.2 g vs. control: 2.3±0.2 g; p=0.731; t=0.348; N=8–10/group). In contrast, mutant mice showed a slightly reduction in body weight compared to controls during adulthood (15±0.7 g vs. control: 17±0.8 g; p=0.031; t=2.393; N=8/group).

The Phox2bΔ8 mutation in Atoh1-expressing cells did not affect respiratory frequency during both neonatal and adult phase (neonate mutant: 179±18 vs. control: 165±11 bpm, p=0.479, t=0.723; adult mutant: 231±7 vs. control: 218±6 bpm, p=0.137, t=1.574; Figure 1A). However, VT was higher in neonate mutant mice vs. control (neonate mutant: 13±0.9 vs. control: 9±0.4 μL/g, p=0.0007, t=4.219, Figure 1B). As a result, VE was higher in neonate mutant compared to control (neonate mutant: 2373±353 vs. control: 1541±123 μL/min/g, p=0.0274, t=2.428; Figure 1C). On the other hand, there was no difference in VT (adult mutants: 15±2 µL/g vs. control: 12±1 μL/g, p=0.1084, t=1.715; Figure 1B) and VE (adult mutants: 3577±370 vs. control: 2705±263 μL/min/g, p=0.0755, t=1.920; Figure 1C) in adults.

Figure 1
Download asset Open asset
Functional respiratory changes observed in the Phox2b∆8 mutation in Atoh1cre-expressing cells.

Changes in (A) respiratory frequency (fR; breaths/min), (B) tidal volume (VT; μL/g), (C) minute ventilation (VE; μL/min/g), (D) inspiratory time (TI; s), (E) expiratory time (TE; s), (F) total cycle duration (TTOT; s), (G) oxygen consumption (VO2, μL/min/g), (H) air convection requirements VE/VO2 (a.u.), and (I) number of apneas in control and mutant (Phox2b∆8, Atoh1cre) mice during neonatal and adult phase. Values are expressed as scatter dot plot with means ± SEM. Neonate (N=8–10/group); adult (N=8/group). *p<0.05 vs. control from Mann-Whitney U test.

Figure 1—source data 1

Raw respiratory parameters of control and Phox2bdelta8/Atoh1-cre mice.

https://cdn.elifesciences.org/articles/73130/elife-73130-fig1-data1-v2.xlsx

Phox2bΔ8, Atoh1cre mutation during neonatal phase did not affect inspiratory time (TI) (neonate mutant: 0.11±0.01 vs. control: 0.120±0.01 s, p=0.734, t=0.345; Figure 1D), expiratory time (TE) (neonate mutant: 0.28±0.04 s vs. control: 0.32±0.03 s, p=0.484, t=0.716; Figure 1E) and total cycle duration (TTOT) (neonate mutant: 0.40±0.05 s vs. control: 0.44±0.04 s; p=0.516; t=0.663; Figure 1F) compared to their control littermates. However, during adult phase, mice carrying Phox2bΔ8 mutation exhibited a reduction in TI (adult mutants: 0.082±0.003 s vs. control: 0.096±0.003 s; p=0.0071, t=3.147; Figure 1D), and an increase in TE (adult mutants: 0.21±0.006 s vs. control: 0.19±0.005 s; p=0.0345; t=2.342; Figure 1E) that did not affect TTOT (adult mutants: 0.29±0.007 s vs. control: 0.29±0.006 s; p=0.557, t=0.600; Figure 1F).

To test whether the increase in VT found in the mutant neonates might be an artifact of the whole-body plethysmograph system, in a subset of neonate, respiratory parameters were analyzed using head-out system (data not shown). Although VT was higher in neonate mutant mice compared to controls (9.4±0.31 vs. control: 8.6±0.4 μL/g, p=0.1143), it did not reach statistic difference due the small number per group (N=4).

To investigate whether changes in body weight and respiratory parameters might be related to changes in metabolic rate, we also measure oxygen consumption (VO2) in neonate and adult mice. VO2 and VE/VO2 did not differ between mutant and control littermates during both neonatal and adult phase (Figure 1G–H). These results suggest that changes in body weight and respiratory parameters cannot be explained by changes in baseline metabolic rate.

NPARM Phox2bΔ8 in Atoh1-expressing cells increased the number of apneas and breath irregularity during adult life

We next analyzed whether Phox2bΔ8 mutation in Atoh1-expressing cells increases the number of apneas and breath irregularity during both neonatal and adult phase. As previously mentioned, the genetic strategy used by the present study is known to affect Phox2b neurons in the parafacial region, and these neurons have been proposed to participate as a generator of respiratory rhythm (Huckstepp et al., 2016; Huckstepp et al., 2018). Interestingly, there was no difference in the number of apneas in neonate mutant compared to control (neonate mutant: 7±0.8 vs. control: 5±0.6 apnea/min; p=0.155; t=1.491; Figure 1I). However, during adulthood the number of apneas in Phox2bΔ8 mutation was higher compared to controls (adult mutants: 7±0.8 vs. control: 3±0.2 apnea/min p=0.0007; t=4.351; Figure 1I; Figure 2).

Figure 2
Download asset Open asset
Breath variability increased in adult mice.

Typical examples of Poincare plot graphs showing SD1 and SD2 from breath duration (TTOT) vs. duration of the subsequent breath (TTOT n+1) in (A) control and (B) mutant mice (Phox2b∆8, Atoh1cre) in neonatal (P3; red circles) and adult (P45; closed circles) phase. (C) Mean ± SEM of SD1 and (D) SD2 during neonatal and adult phases. Neonate (N=8–10/group); adult (N=8/group). *p<0.05 from Mann-Whitney U test.

Figure 2—source data 1

Raw breath variability of control and Phox2bdelta8/Atoh1-cre mice.

https://cdn.elifesciences.org/articles/73130/elife-73130-fig2-data1-v2.xlsx

Breath-to-breath interval was also used as an indicative of breath irregularity. Figures 3A and 4A illustrate breathing recording at rest in controls and mutant mice during both neonate and adult phases, respectively. Phox2bΔ8 mutation did not alter breath-to-breath interval in neonates (neonate mutant: 0.35±0.06 vs. control: 0.33±0.03; p=0.838; t=0.207). In contrast, breath-to-breath interval was significantly higher in mutant adult mice compared to controls (adult mutants: 0.31±0.02 vs. control: 0.18±0.009; p<0.0001; t=5.505).

Figure 3
Download asset Open asset
Phox2bΔ8 in Atoh1cre cells impaired ventilatory responses to hypoxia and hypercapnia in neonates.

(A) Representative plethysmograph breathing traces in a control (top traces) and mutant (Phox2b∆8, Atoh1cre; bottom traces) neonate (P3) mice while ventilated with room air (normoxia; FiO2=0.21); hypoxia (FiO2=0.08); and hypercapnia (FiCO2=0.07). Percentage changes produced by hypoxia or hypercapnia in neonate control and mutant mice. (B) Respiratory frequency (fR; interaction: F(2,32)=0.8, p=0.455; effect of mutation F(1,16)=4.3, p=0.052; effect of hypoxia and hypercapnia: F(2,32)=10.5, p=0.0008). (C) Tidal volume (VT; interaction: F(2,32)=1.92, p=0.162; effect of mutation F(1,16)=2.44, p=0.138); effect of hypoxia and hypercapnia: F(2,32)=4.50, p=0.019. (D) Minute ventilation (VE; interaction: F(2,32)=3.32, p=0.048; effect of mutation F(1,16)=4.48, p=0.0503; effect of hypoxia and hypercapnia: F(2,32)=11.6, p=0.0002). Values are expressed as scatter dot plot with means ± SEM. N=8–10/group. ANOVA two-way Dunnett’s multiple comparisons test.

Figure 3—source data 1

Raw respiratory parameters of control and Phox2bdelta8/Atoh1-cre neonate mice under hypoxia and hypercapnia.

https://cdn.elifesciences.org/articles/73130/elife-73130-fig3-data1-v2.xlsx
Figure 4
Download asset Open asset
Phox2bΔ8 in Atoh1cre cells impaired ventilatory responses to hypoxia in adult.

(A) Representative plethysmograph breathing traces in a control and mutant (Phox2b∆8, Atoh1cre) adult (P45) mice while ventilated with room air (normoxia; FiO2=0.21); hypoxia (FiO2=0.08); and hypercapnia (FiCO2=0.07). Percentage changes produced by hypoxia or hypercapnia in adult control and mutant mice in (B) respiratory frequency (fR; interaction: F(6,84) = 0.97, p=0.448; effect of mutation F(1,14)=0.86, p=0.368; effect of time of hypoxia: F(6,84) = 29.32, p<0.0001); (C) tidal volume (VT, interaction: F(6,84) = 1.26, p=0.285; effect of mutation F(1,14)=20.76, p=0.0004; effect of time of hypoxia: F(6,84) = 17.49, p<0.0001); (D) minute ventilation (VE, interaction: F(6,84) = 1.20, p=0.316; effect of mutation F(1,14)=8.22, p=0.012; effect of time of hypoxia: F(6,84) = 23.92, p<0.0001). N=8/group. *p<0.05 vs. 21% O2 in controls. +p < 0.05 vs. 21% O2 in mutants. ANOVA two-way Dunnett’s multiple comparisons test. (E) Respiratory frequency (fR; interaction: F(6,84) = 0.56, p=0.763; effect of mutation F(1,14)=0.41, p=0.532; effect of time of hypercapnia: F(6,84) = 155.48, p<0.0001); (F) tidal volume (VT, interaction: F(6,84) = 0.22, p=0.968; effect of mutation F(1,14)=0.31, p=0.585; effect of time of hypercapnia: F(6,84) = 69.77, p<0.0001); (G) minute ventilation (VE, interaction: F (6,84)=0.34, p=0.914; effect of mutation F(1,14)=0.38, p=0.547; effect of time of hypercapnia: F(6,84) = 86.85, p<0.0001). N=8/group. *p<0.05 vs. 21% O2 for both control and mutation group. ANOVA two-way Dunnett’s multiple comparisons test.

Figure 4—source data 1

Raw respiratory parameters of control and Phox2bdelta8/Atoh1-cre adult mice under hypoxia and hypercapnia.

https://cdn.elifesciences.org/articles/73130/elife-73130-fig4-data1-v2.xlsx

In addition to the time domain analysis, we also used a nonlinear method to investigate breath variability (Li and Nattie, 2006; Patrone et al., 2018; Fernandes-Junior et al., 2018). We quantified the distribution of the breath duration using the SD1 and SD2 parameters from the Poincare plots (Figure 2A–B). SD1 and SD2 were similar between mutant and control neonates (SD1: 132±31 ms vs. control: 159±36 ms, p=0.571, t=0.577; SD2: 135±28 ms vs. control: 223±54 ms, p=0.193, t=1.357) (Figure 2C–D). However, in agreement with the breath-to-breath interval, Phox2bΔ8 mutation showed higher SD1 (77±6 vs. control: 38±3; p<0.0001; t=5.827) and SD2 (107±9 vs. control: 57±4; p=0.0002; t=4.995) in adult mice (Figure 2C–D). Altogether, these results suggest that breath irregularity is increased in adult mice carrying NPARM Phox2b Δ8 mutation in Atoh1-expressing cells.

NPARM Phox2bΔ8 in Atoh1-expressing cells impaired ventilatory responses to hypoxia and hypercapnia in neonates

A common symptom experienced by patients with the CCHS is an impaired ventilatory response to hypoxia and hypercapnia (Patwari et al., 2010; Moreira et al., 2016). Therefore, we next explored whether a conditional Phox2bΔ8 in Atoh1- expressing cells impairs ventilatory response to hypoxia and hypercapnia during the first days of life. Figure 3A illustrates examples of breathing recording at room air (left traces) and hypoxic challenge (middle traces) in a control (top) and mutant (bottom) mice 3 days after birth. We monitored baseline ventilation while neonates were breathing room air followed by 5 min of hypoxia. We analyzed the first minute of hypoxic exposure because longer than 5 min of low O2 exposure is known to lower body temperature (Kline et al., 1998). Figure 3B–D illustrates percentage change in the respiratory frequency (fR), tidal volume (VT), and ventilation (VE). As expected, neonate control littermates increased fR ≈ 40% (from 100 ± 6% to 139 ± 5%; p=0.0016; Figure 3B) and the VT increased 46% (from: 100±9% to 146 ± 15%; p<0.0001; Figure 3C). That results in a significant increase in VE (from: 100±12% to 205 ± 24%; p<0.0001; Figure 3D). In contrast, neonate mutant failed to significantly increase VE during hypoxia stimulus (from: 100±15% to 128 ± 11%; p=0.341; Figure 3D). Note that we did not observe a significant change in both, fR (from: 100±10% to 123 ± 8%; p=0.269; Figure 3B) and in VT responses (from: 100±10% to 105 ± 9%; p=0.910; Figure 3C).

Our next goal was to investigate whether a conditional Phox2bΔ8 mutation impairs ventilatory response to hypercapnia. Figure 3A illustrates a typical respiratory trace from the same cre-negative and mutant neonate mice but now ventilated with 7% of CO2 (right traces). As expected, neonatal control mice increased fR by approximately 30% during hypercapnia when compared to normoxia (from 100±6% to 129 ± 4%; p=0.0321; Figure 3B). In addition, VT increased significantly from 100±9% to 134 ± 11% (p=0.042; Figure 3C). Therefore, VE increased 71% in the control pups (from 100±12% to 171±11%; p=0.002; Figure 3D). In contrast, Phox2bΔ8 mutation failed to significantly increase VE during hypercapnia (from 100±15% to 144±18%; p=0.096; Figure 3D). The reduction was related to an impairment in fR (from 100±10% to 113±5%; p=0.360; Figure 3B) and VT responses (from 100±6% to 124±13%; p=0.220; Figure 3C).

These data suggest that Phox2bΔ8 in Atoh1-expressing cells affect ventilatory responses to hypoxia and hypercapnia during neonatal phase.

Hypoxia, but not hypercapnic, ventilatory responses still partially compromised in the Phox2bΔ8, Atoh1cre adult mice

Figure 4A illustrates typical breathing traces in a control (top traces) and mutated adult mouse (bottom traces) while ventilated with room air (left traces), hypoxia (middle traces), and hypercapnia (right traces). Figure 4B–D illustrates changes in respiratory frequency, tidal volume, and minute ventilation before and during the hypoxic stimulus (10 min). As expected, breathing activity increased during hypoxia in control adult mice. Respiratory frequency increased at the first minute (from: 100±4% to 145±5%; p=0.0009), then slowly declined until the end of the hypoxia stimuli (Figure 4B). VT significantly increased from min 3 (from: 100±7% to 163±11%; p=0.029) and persistently elevated until min 7 (139±4%; p=0.027) (Figure 4C). Consequently, VE significantly increased from min 1 (from: 100±11% to 200±21%; p=0.024) to min 7 (163±7%; p=0.028) (Figure 4D).

In contrast, mutant mice only had a significant increase in fR at first minute of hypoxia from 100±5% to 154±10% (p=0.032; Figure 4B), but failed to increase VT across the stimulus (Figure 4C). Consequently, the increase in VE was compromised (Figure 4D). These results demonstrate that mutant mice had an impaired ventilatory response to hypoxia in the adult phase.

Interestingly, hypercapnia similarly increased fR, VT, and VE in both, mutant and control adult littermates (Figure 4E–G). These results demonstrate that mutant mice completely recovered the ventilatory response induced by hypercapnia in the adult phase.

NPARM Phox2bΔ8 in Atoh1-expressing cells reduced Phox2b immunoreactivity in the parafacial/RTN region

The CO2-sensitive cells of the ventral aspect of the respiratory parafacial/RTN region belong to a neuronal group with a well-defined phenotype characterized by the presence of Phox2b immunoreactivity and the absence of TH (henceforth called parafacial/RTN neurons) (Stornetta et al., 2006; Shi et al., 2017). According to prior evidence, Phox2b is predominantly expressed by the CO2-activated neurons in the RTN region (Stornetta et al., 2006; Shi et al., 2017). But this marker is also present in a fraction of catecholaminergic neurons (known as C1) located close to the CO2-sensitive neurons (Stornetta et al., 2006; Shi et al., 2017). The C1 neurons are normally bulbospinal blood pressure-regulating neurons (Guyenet, 2006) that can be distinguished from the CO2-sensitive cells by the presence of TH (Takakura et al., 2008; Takakura et al., 2014; Barna et al., 2012; Barna et al., 2014). Thus, to assess the extent to which the mutation affects Phox2b expression in parafacial/RTN and C1 regions, we counted the number of Phox2b-expressing neurons that did not express TH (Phox2b+/TH-) and those that co-express TH (Phox2b+/TH+), respectively.

Figure 5 shows typical photomicrographs and representative diagrams from several Bregma levels in a control (A and B) and mutant (C and D) adult mouse. The total number of Phox2b+ neurons (that include RTN and C1 neurons) was reduced in mutant adult mice (145±36 vs. control: 258±32; p=0.041; t=2.334; Figure 5E). The number of Phox2b+/TH+ (therefore C1 neurons) was similar between mutant and control mice (170±17 vs. control: 156±9; p=0.444; t=0.796; Figure 5E), which strongly suggests that Phox2b mutation in Atoh1-expressing cells did not compromise C1 neurons. On the other hand, the total number of Phox2b+/TH- neurons (RTN neurons) reduced ≈50% compared to controls (124±38 vs. control: 236±31; p=0.047; t=2.257; Figure 5C, D and E). These results indicated that Phox2b mutation in Atoh1-expressing cells compromised chemosensitive neurons (Phox2b+/TH-) in the parafacial/RTN region.

Figure 5
Download asset Open asset
Adult mutant mice reduced Phox2b expression in the parafacial/retrotrapezoid nucleus (RTN) region.

Photomicrographs of ventrolateral medulla from (A) control and (C) mutant (Phox2b∆8, Atoh1cre) adult mice. Schematic drawings represent examples of coronal sections of ventrolateral medulla in (B) control and (D) Phox2b∆8, Atoh1cre mutant mice. Each square represents immunoreactivity for Phox2b and tyrosine hydroxylase (Phox2b+/TH+). The stars represent immunoreactivity for Phox2b and absence of TH (Phox2b+/TH-). The numbers in the middle of each section refer to the location caudal to the Bregma level (in mm) according to the Mouse Brain Atlas of Franklin and Paxinos, 2015. (E) Total number of cells that expressed Phox2b and TH immunoreactivity in the ventrolateral medulla (parafacial/RTN and C1 region) in control and Phox2b∆8, Atoh1cre (N=6/group). (F) Photomicrographs showing locus coeruleus and subcoeruleus region from control and mutant (Phox2b∆8, Atoh1cre) mice. *p<0.05 vs. control, unpaired t-test. Abbreviations: IO, inferior olive; NA, nucleus ambiguous; py, pyramid tract; Sp5, spinal trigeminal tract; VII, facial motor nucleus. Scale bar: C=50 μm applied to A; D=1 mm applied to B; F=100 μm.

Figure 5—source data 1

Raw numbers of neuronal profiles (Phox2b and TH) of control and Phox2bdelta8/Atoh1-cre adult mice.

https://cdn.elifesciences.org/articles/73130/elife-73130-fig5-data1-v2.xlsx

We also examined the effect of the mutation on catecholaminergic cells located in the locus coeruleus (LC) (Figure 5F). Based on TH and Phox2b immunoreactivity, the mutation had no apparent effect on TH+ neurons located in the LC region neither in Phox2b+ cells nor in the sub-LC region (Figure 5F).

NPARM Phox2bΔ8 in Atoh1cre cells reduced the activation of ventral respiratory parafacial/RTN neurons by hypercapnia

As previously shown in Figure 4, Phox2bΔ8, Atoh1cre in adult mice completely recovered ventilatory response induced by hypercapnia. However, these experiments did not rule out whether it involves activation of parafacial/RTN neurons. Thus, the next set of experiments were done to explore the involvement of the remaining parafacial neurons in response to hypercapnia. Mutant and control adult mice were challenged with hypercapnia and fos-immunoreactive was used as a reporter of cell activation. Hypercapnia is well known to induce fos expression in the rodent respiratory parafacial/RTN neurons (Sato et al., 1992; Teppema et al., 1994; Fortuna et al., 2009; Kumar et al., 2015; Shi et al., 2021). To differentiate between parafacial/RTN neurons and adjacent C1 neurons, we also analyzed the expression of TH. Therefore, parafacial/RTN neurons were defined by the presence of fos and absence of TH expression (TH-) (Stornetta et al., 2006; Barna et al., 2012; Barna et al., 2014; Shi et al., 2017).

Excluding the facial motor nucleus, which expresses very low levels of fos-immunoreactive after hypercapnia, the ventrolateral medulla contains two clusters of fos-positive neurons centered predominantly within the rostral aspect. The fos-immunoreactive was expressed in both catecholaminergic (identified by TH+) and non-catecholaminergic neurons (TH-) in control and mutant mice (Figure 6A and C). In control animals, of the total 94±13 fos-immunoreactive neurons within the respiratory parafacial/RTN region, 86±12 (91%) were non-catecholaminergic, that is, presumably chemosensitive neurons (Figure 6A, B and E). On the other hand, in mutant mice, hypercapnia induced fos in only 47±7 neurons and a total of 37±8 were fos+/TH- cells (reduction of 56%) (Figure 6C, D and E). These cells were generally located lateral to the TH+ neurons and under the facial motor nucleus (Figure 6C, D and E). The neurons in this region are well known to belong to a cell group with a well-defined phenotype characterized by the presence of VGlut2 mRNA and the absence of both TH and choline acetyltransferase (Stornetta et al., 2006; Shi et al., 2017). In a subset of animals (N=3), fos expression was found in only six to eight neurons when exposed to room air (data not shown), which strongly suggested the effect of hypercapnia in activated neurons in the parafacial/RTN neurons.

Figure 6
Download asset Open asset
Fos-activated neurons in the parafacial/retrotrapezoid nucleus (RTN) region in response to hypercapnia are reduced in mutant mice.

Photomicrographs of ventrolateral medulla from (A) control and (C) mutant (Phox2b∆8, Atoh1cre) mice exposed to hypercapnia (FiCO2=0.07). Schematic drawings represent coronal sections of ventrolateral medulla in (B) control and (D) Phox2b∆8, Atoh1cre mutant mice. Each square represents tyrosine hydroxylase immunoreactivity (TH+). The stars represent fos and the absence of TH (fos+/TH-). The numbers in the middle of the sections refer to the location caudal to the Bregma level (in mm) according to the Mouse Brain Atlas of Franklin and Paxinos, 2015. (E) Total number of cells that expressed fos and TH immunoreactivity in the ventrolateral medulla (respiratory parafacial/RTN region) in control and Phox2b∆8, Atoh1cre mice (N=6/group). *p<0.05 vs. control; unpaired t-test. Abbreviations: IO, inferior olive; NA, nucleus ambiguous; py, pyramid tract; Sp5, spinal trigeminal tract; VII, facial motor nucleus. Scale bar: C=50 μm applied to A; D=1 mm applied to B.

Figure 6—source data 1

Raw numbers of neuronal profiles (fos and TH) of control and Phox2bdelta8/Atoh1-cre adult mice under hypercapnia.

https://cdn.elifesciences.org/articles/73130/elife-73130-fig6-data1-v2.xlsx

These results indicate that Phox2bΔ8 in Atoh1-expressing cells compromised the number of activated neurons in the parafacial/RTN region induced by hypercapnia.

Discussion

In the present study, we used a conditionally activated NPARM patient-specific transgenic mouse model to investigate the effect of the mutant protein in Atoh1-expressing cells on respiratory function during neonatal and adult life. We found that the mutation resulted in (a) impaired hypoxic and hypercapnic ventilatory responses in neonates; (b) the ventilatory response to hypoxia, but not to hypercapnia, was reduced in adults; (c) the number of irregular breathing pattern increased in adults; (d) Phox2b expression within parafacial/RTN region (Phox2b+/TH-) reduced ≈50%; (e) no significant change in the number of catecholaminergic cells (TH+) located in the ventrolateral medulla (C1 region) or in the dorsolateral pons (LC and subcoeruleus region); (f) the mutation also reduced the number of hypercapnic fos-activated neurons in the parafacial/RTN (fos+/TH-) by 56%. These findings demonstrate for the first time that NPARM Phox2bΔ8 in Atoh1-expressing cells (in the parafacial/RTN and intertrigeminal region) affects regulation of breathing and chemosensory respiratory control for both, hypercapnia and hypoxia, especially in neonates. Furthermore, it showed that despite an impaired RTN region, anatomically and functionally during adulthood, the system adapted and developed appropriate responses to hypercapnia, but not to hypoxia (Figure 7).

Figure 7
Download asset Open asset
Schematic view of the mouse hindbrain control of breathing and the role of transcriptions factors and neuromodulators.

The respiratory parafacial region (pF) contains neurons involved in breathing regulation. Within the ventral aspect of the pF, retrotrapezoid nucleus (RTN) could be defined as a cluster of neurons positive for Phox2b, neuromedin (Nmb), NK1, glutamatergic (VGlut2), pituitary adenylate cyclase-activating peptide (PACAP) and the absence of tyrosine hydroxylase (TH), choline acetyltransferase (ChAT), serotonin, GABA, and glycine. These neurons are activated by CO2 via their intrinsic pH sensitivity and via inputs from the carotid bodies. The RTN of mice has a distinctive developmental lineage that relies on transcription factors Egr2, Phox2b, Lbx1, and Atoh1. Phox2b is the only one that remains expressed in adulthood. RTN progenitors originate from the dB2 domain of rhombomere 5. These progenitors are Phox2b-positive, switch on Lbx1 at the postmitotic stage, migrate ventrally, and activate Atoh-1 expression once they reach the region of the facial motor nucleus. In the respiratory pF also have distinct functional subgroup of neurons, that is, pF ventral neurons (non-RTN) and pF lateral neurons (expiratory oscillators). RTN neurons target various components of the respiratory central pattern generator (rCPG) and are presumed to play a key role in breathing automaticity during anesthesia, sleep, and quiet waking. The carotid body may also influence the activity of the rCPG neurons through connections that bypass the RTN (Stornetta et al., 2006; Takakura et al., 2006). The ventilatory response to CO2 also has a contribution of the catecholaminergic neurons located in the locus coeruleus (LC). Here, we showed that breathing dysfunction of the humanized NPARM Phox2bΔ8 mutation in Atoh1-expressing cells is presumably mediated by loss of cells in the ventral parafacial region. Abbreviations: Atoh1, atonal homolog 1; ChAT, choline acetyltransferase; LC, locus coeruleus; Nmb, neuromedin B; NTS, nucleus of the solitary tract; NK1, tachykinin 1; PACAP, pituitary adenylate cyclase-activating peptide; Phox2b, paired like homeobox 2B; rCPG, respiratory central pattern generator; TH, tyrosine hydroxylase; VGlut2 (Slc17a6), vesicular glutamate transporter 2.

The effect of Phox2bΔ8, Atoh1cre on baseline respiratory function

Our first goal was to investigate the effect of the Phox2bΔ8 mutation on respiratory control at rest. The mutant neonates showed a slightly increase in tidal volume and consequently in total ventilation compared to their control littermates. It was a surprise since hypoventilation is usually found in both humans and experimental model of CCHS (Amiel et al., 2003; Ramanantsoa et al., 2011; Carroll et al., 2014; Hernandez-Miranda et al., 2018). Even when the mutation occurs specific to RTN neurons, it was demonstrated by a previous study that applied PARM Phox2b (Phox2b27Ala) using the Egr2cre (Krox20cre) promoter (Ramanantsoa et al., 2011). However, the reduction in ventilation in their study was due to reduction in total cycle duration with no change in tidal volume.

To further investigate whether the result in our study could be a bias of the plethysmograph method used (whole-body), in a subset of neonates, we used a more accurate method, the head-out plethysmograph. Despite the small number sample (N=4/group) we found similar results, showing a tendency to slightly increase in tidal volume and minute ventilation in the mutant neonate group. It is important to mention that, despite the small increase in ventilation there was no difference in oxygen consumption between mutants and control mice, indicating no changes in metabolic rate induced by the mutation.

Additionally, the small increase in ventilation does not seem to be a consequence of the genetic approach used by us, which target not only ventral parafacial/RTN region (peri VII), but also intertrigeminal neurons (peri V) that express both Phox2b and Atoh1 (Ruffault et al., 2015). The result suggests a direct effect of NPARM Phox2bΔ8 mutation used in our study. Because inactivation of Phox2b from Atoh1-expressing cells, that is, the same population target in our study, found reduction in the ventilation in mutant neonates (Ruffault et al., 2015) and contradict with our study. Unfortunately, we do not know from the former study if the reduction in ventilation was due to change in tidal volume and/or respiratory frequency.

Whether the change is a consequence of NPARM Phox2bΔ8 mutation specific to peri VII and/or peri V region, we do not have a clear picture because we do not use any strategy to target one of those population. Although, there is no data in the literature showing the contribution of peri V Phox2b neurons to regulate tidal volume. When Atoh1 neurons were specifically removed from peri V region, it reduced tidal volume and consequently minute ventilation in mice at 3 weeks of age (van der Heijden and Zoghbi, 2018). It is also important to highlight that all studies cited above measured ventilation in neonates using the whole-body plethysmograph method. Thus, more accurate methods to measure neonatal tidal volume as head-out plethysmograph are required in further studies.

The effect of Phox2bΔ8, Atoh1cre on breath irregularity

In a healthy system, some level of respiratory variability is expected to occur since it can be affected by several factors, for instance chemical drive, excitatory and inhibitory input from many sources (Khoo, 2000). In the present study, adults carrying the Phox2bΔ8 in Atoh1-expressing cells showed higher number of apneas, inter-breath interval (IBI), and breath variability. However, mutant neonates did not differ from controls littermates. Curiously, Phox2bΔ8 in the Nkx2.2-derived progenitor domains (visceral motor non-respiratory neurons) reported apneic phenotype at birth and abnormal respiratory pattern (Alzate-Correa et al., 2021). Additionally, the loss of Phox2b neurons in the RTN region impaired inspiratory rhythmogenesis from pre-BötC. Pre-BötC neurons are known to receive Atoh1-dependent neuronal projections from both peri V and peri VII neurons (Huang et al., 2012). Therefore, Phox2bΔ8 mutation in Atoh1-expressing cells could affect the excitatory tonic drive to pre-BötC neurons and increase irregular respiratory rhythm. Most studies applying different genetic strategies to target Phox2b neurons also reported higher number of apneas as soon as after birth (Ramanantsoa et al., 2011; Ruffault et al., 2015). We still do not know why higher irregular breathing patterns were only identified during adulthood, since breathing is well known to maturate post-natally.

The role of Phox2bΔ8, Atoh1cre on respiratory chemoreception

CCHS is characterized by impaired ventilatory response to hypoxia and hypercapnia. Our physiological data showed that both responses were blunted in neonates carrying NPARM Phox2bΔ8 mutation in Atoh1-expressing cells. However, while hypercapnic ventilatory response completely recovered during adulthood, the hypoxic ventilatory response still partially compromised. The blunted response to hypercapnia in neonates is in line with other findings in the literature that used both PARM Phox2b mutation and genetically removed Phox2b from Atoh1-expressing cells in mice (Ramanantsoa et al., 2011; Ruffault et al., 2015). In addition, as previously found in adults carrying PARM Phox2b mutation restricted to RTN neurons, animals recovered hypercapnic response during adulthood (Ramanantsoa et al., 2011). Although similar findings were reported when deleting Atoh1 from peri V and peri VII region (Huang et al., 2012; Ruffault et al., 2015). It is unknown whether genetic deletion of Phox2b from Atoh1-expressing cells also recovers the ventilatory response to hypercapnia in adults.

The defect in the hypercapnic ventilatory response in neonate seems to be caused by anatomical and functional damage in neurons from peri VII region. In a brainstem spinal cord preparation, the increase in phrenic nerve activity in response to low pH was fully preserved after complete resection of peri V region (Ruffault et al., 2015). These in vitro data indicate that the peri V Phox2b/Atoh1-expressing neurons are not essential to chemosensitivity to CO2/H+. In addition, a recent study showed that loss of Atoh1 specifically from peri V Phox2b/Atoh1 neurons did not compromise in vivo breathing responses to hypercapnia in neonates (van der Heijden and Zoghbi, 2018). Together, these results indicate that the compromised ventilatory response to hypercapnia is due to an impairment in parafacial/RTN neurons.

Phox2b-expressing RTN neurons located in the peri VII region are important CO2 sensors in the brain and receive chemosensory inputs from other cells in the respiratory column in the brainstem (Rosin et al., 2006; Guyenet et al., 2005). Briefly, RTN neurons (a) are sensitive to small changes in CO2/H+ (Mulkey et al., 2004; Onimaru et al., 2008; Wang et al., 2013); (b) are in close opposition to numerous capillaries (Onimaru et al., 2012; Hawkins et al., 2017; Cleary et al., 2020), classifying these neurons to be critical to sense CO2/H+ in the blood; (c) receive afferents from many brainstem sites that contain putative chemosensors (Rosin et al., 2006); (d) respond with depolarization to activation of nearby acid-sensitive astrocytes (Gourine et al., 2010; Wenker et al., 2010; Wenker et al., 2012), and (e) receive excitatory connections from the carotid bodies (Takakura et al., 2006).

In the present study, NPARM Phox2bΔ8 adult mice expressed only 50% of Phox2b+/TH- (likely chemosensitive RTN/parafacial) neurons compared to their control. Neurons were located ventrally and laterally to facial motor nucleus. Similar to it, inactivation of Phox2b from Atoh1-expressing cells only expressed 40% of Phox2b+/Atoh1+ neurons from their controls in the RTN at 18.5 days of embryonic age. In addition, as in our study, cells were located ventral to facial nucleus. A massive loss of Phox2b+/TH- neurons occurred when PARM Phox2b mutation was introduced in the RTN. Thus, loss of RTN chemosensitive neurons might be responsible for the blunted hypercapnic response in neonate. Therefore, the open question is what mechanisms enable the neonate mutant to maintain their ventilation, and presumably normal blood PCO2 in a condition where the chemosensors neurons in the RTN were importantly reduced.

Here, we showed that although adult mutant mice recovered ventilatory response to hypercapnia, there was a reduction in both the Phox2b+/TH- and fos+/TH--activated neurons under hypercapnia in the parafacial/RTN region. Interestingly, former studies applied different strategies to manipulate Phox2b parafacial/RTN neurons, CO2 response was only partially recovered in the adult life (Ruffault et al., 2015; Ramanantsoa et al., 2011; Huang et al., 2012; Hernandez-Miranda et al., 2018). Although, an extensive depletion of parafacial/RTN neurons occurred at embryonic ages. There is no information whether those neurons still depleted during adulthood and whether they are functional. Thus, it complicates further discussion with our finding. The recovery of the CO2 chemoreflex in adults in our study might be due to a late compensation of residual RTN neurons, peripheral chemoreceptor, and/or to some of the multiple chemosensors sites as previously described (Nattie, 2011).

One possibility is that carotid body compensates for the CO2 drive to breathe and then through nucleus of the solitary tract (NTS) activates the respiratory column to maintain breathing activity. The plausible explanation emerges by considering that RTN neurons are strongly activated by carotid body stimulation and provide powerful excitatory input to the respiratory column (Takakura et al., 2006). They may thus be obligatory intermediates for relaying the CO2 response when occur a loss of RTN neurons early in life. The second possibility is that RTN is not an obligatory site for central chemoreceptors in adults when the neurons are damaged at first days of life. Other candidates of chemoreceptor sites could assume the function. Those candidates are serotonergic neurons that have been reported to be pH-sensitive (Wang and Richerson, 1999; Corcoran et al., 2009), the noradrenergic neurons located in the LC (Biancardi et al., 2008) and glial cells (Gourine et al., 2010; Wenker et al., 2010; Wenker et al., 2012; Sobrinho et al., 2014).

The results of previous loss-of-function experiments to assess the role played by RTN neurons in the chemoreflex in adults are not entirely conclusive. In previous work, we evaluated the chemoreflex in which subsets of Phox2b-expressing neurons in the RTN were lesioned using toxin or pharmacological tools (Takakura et al., 2006; Takakura et al., 2008; Takakura et al., 2013; Takakura et al., 2014). Bilateral lesions of the neurokinin1 receptor-expressing neurons in the RTN region by injection of saporin conjugated to a substance P or injection of the GABA-A agonist muscimol reduced hypercapnic ventilatory response in adult rats (Nattie and Li, 2002; Takakura et al., 2008; Takakura et al., 2013; Takakura et al., 2014). However, these experiments lack specificity, and the extension of the lesion or inhibition is difficult to control. Using a more selective approach, Marina et al., 2010, applied a pharmacogenetic tool to silence RTN neurons. Rats that received injection of lentivirus vector expressing the allatostatin receptor from PRSx8 promoter reduced the hypercapnic ventilatory response after administration of allatostatin. However, the PRSx8 promoter used targets Phox2a and Phox2b neurons in the rostral aspect of the ventrolateral medulla, which includes RTN, C1 adrenergic, and A5 noradrenergic neurons (Stornetta et al., 2006; Abbott et al., 2013; Burke et al., 2014; Malheiros-Lima et al., 2018; Malheiros-Lima et al., 2020). Furthermore, it is important to mention that studies that tested loss of function of RTN/parafacial neurons in adult rodents need to be carefully discussed. Since it might exist important differences in neuronal plasticity, when comparing to neurons that were damaged early in the life.

Another important finding in our study was the compromised ventilatory response to hypoxia in both neonates and adult mutants. The hypoxic ventilatory response emerges from a physiological reflex of the already established notion that ventrolateral brainstem respiratory neurons are excited by peripheral chemoreceptors via a direct glutamatergic input from commissural NTS (Guyenet, 2014). Besides the di-synaptic excitatory pathway from commissural NTS to RVLM, we also know that we have a relay via the chemosensitive neurons of the RTN (secondary input) (Takakura et al., 2006). Thus, the compromised ventilatory response to hypoxia could be explained by the fact that this pathway was affected by the conditional Phox2bΔ8 mutation in Atoh1-expressing cells.

The impaired ventilatory response to hypoxia in neonates is in contrast with previous work that used PARM Phox2b mutation specific to RTN neurons. Interestingly, PARM Phox2b mutation in the RTN showed intact and even higher ventilatory response to hypoxia in neonates, despite the abrupt loss of RTN neurons (Ramanantsoa et al., 2011). This difference could be explained by the fact that in their study, peripheral chemoreceptors are potentialized in neonates. When neonate mutants were exposed to hyperoxia (100% O2) they showed higher respiratory depression and apneas compared to their control. Although in the present study we have not tested hyperoxia in neonates, we found that at least in adults it did not cause any effect (data not shown).

The open question that needs to be investigated is by which mechanism NPARM Phox2bΔ8 mutation in Atoh1-expressing cells compromise chemosensory control of breathing in both neonates and adults. Such mechanisms may involve selective loss of neurons, disorganized respiratory circuits, that likely contributes to the irregular breathing pattern and apneic phenotype during adulthood. In addition, further studies could investigate whether the respiratory function and chemoreflex responses in mutants are altered during sleep stages.

Conclusion

Our data established the NPARM Phox2bΔ8 mutation in Atoh1-expressing cells with an impaired ventilatory response to hypercapnia and hypoxia in neonates. Although adult mutant mice recovered the ventilatory response to hypercapnia, the hypoxia ventilatory response still compromised, suggesting a reorganization within the chemoreflex pathways (Figure 7). In other words, the conditional Phox2bΔ8 mutation in Atoh1-expressing cells affects the peripheral chemoreflex pathway and the important cells that serve as relevant chemosensors in the ventral aspect of the parafacial/RTN region. The remaining questions are: (a) How neonates were able to maintain their ventilation even with compromised hypoxic and hypercapnic ventilatory responses? (b) How the hypercapnic ventilatory response was restored in adult NPARM Phox2bΔ8 mutation in Atoh1-expressing cells? Although parafacial/RTN neurons are particularly notable as they are important for respiratory chemoreceptors, substantial evidence has accrued supporting involvement of multiple cell types to maintain stable blood gases parameters, avoiding respiratory acidosis.

We showed that breathing dysfunction of the humanized NPARM Phox2bΔ8 mutation in Atoh1-expressing cells is presumably mediated by loss of cells in the ventral parafacial region. Given that many other physiological processes could be affected by the mutation, our model may help to understand how specific brain areas and neurons generate and control complex behaviors more generally.

Materials and methods

Animals

This study was conducted in accordance with the University of Sao Paulo Institutional Animal Care and Use Committee guidelines (protocol number: 3618221019). Our goal was to introduce the NPARM in regions involved with respiratory function and chemoreflex. We used a transgenic mouse line with a cre-loxP-inducible humanized Phox2b mutation defined as Phox2bΔ8 and crossed them with Atoh-1cre mice (Nobuta et al., 2015; Alzate-Correa et al., 2021). These animals were bred with Atoh1cre mice to allow conditional expression of Phox2b mutant gene in the parafacial and intertrigeminal region. Genotyping was verified by PCR (REDTaq ReadyMix # R4775, Sigma-Aldrich). The primers, genotyping details, and strain number of mice used are delineated in Table 1.

Table 1
Genotyping primers.
Mouse lineStrain nameStrain #Obtained fromPrimersBand sizes
Atoh1CreB6.Cg-Tg(Atoh1-cre)1Bfri/JJax: 011104Jackson LaboratoriesTg FWD 5'-CCG GCA GAG TTT ACA GAA GC-3'Tg = 450 bp
Tg REV 5'-ATG TTT AGC TGG CCC AAA TG-3'CTR = 324 bp
CTR FWD 5'-CTA GGC CAC AGA ATT GAA AGA TCT-3'
CTR REV 5'-GTA GGT GGA AAT TCT AGC ATC ATC C-3'
Phox2bΔ8B6.129(Cg)-Phox2btm1Rth/JJax: 025436David Rowitch, UCSFFWD 5'-GCC CAC AGT GCC TCT TAA CTC-3'Mutant = 450 bp
REV 5'-CGT ACT CTT AAA CGG GCG TCT C-3'Wild type = 334 bp

Ventilation measurements

Request a detailed protocol

Breathing variables of neonatal (P1-3) and adult (P30-45) mice from both sexes were measured noninvasively in unanesthetized and unrestrained using the whole-body plethysmography closed system and the head-out pressure-plethysmography method (Drorbaug and Fenn, 1955; Bartlett and Tenney, 1970; Mortola, 1984; Durand et al., 2004; Mortola and Frappell, 2013; Patrone et al., 2018).

In neonates, part of the respiratory recording was done using the head-out pressure-plethysmography method (N=4/group) and part using whole-body plethysmography closed system (N=10–8/group).

The head-out pressure plethysmograph consists of separate head and body chambers that were 10 and 30 mL for P1-3 mice. The head and body chambers were separated by a pliable neck collar of plastic film that provided an air-tight seal between the two chambers. Three, premixed gas mixtures (room air 21% O2, balance N2; hypercapnia 7% CO2, 21% O2, balance N2; hypoxia 8% O2, balance N2; Oxylumen Gases Industriais Ltda, Sao Paulo, Brazil) were delivered continuously through the head chamber mice a flow rate of 40 mL/min for P1-3. The body chamber was sealed but had two ports, one for the differential pressure transducer (FE 141 Spirometer, ADInstruments, Sydney, Australia) used to monitor pressure oscillations associated with breathing and the other calibration port for injecting and withdrawing known volumes of gas (via a graduated syringe). Calibration of the system via injection of different volumes of air into the body chamber (0.2, 0.4, 0.6, 0.8 mL) established that the pressure signal (mV) was directly proportional to volume and that the relationship was linear (R2=0.999). The pressure signal was amplified (FE 141 Spirometer, ADInstruments), digitized (200 Hz), and stored on computer via acquisition software (PowerLab system, ADInstruments/LabChart Software, version 7.3). The entire plethysmograph system was under a controlled temperature to maintain in the thermoneutral zone for P1-3 age between 32.5°C and 33.5°C (Mortola, 1984).

For whole-body plethysmography closed system, the plethysmograph chamber of neonate had 40 mL and was saturated with water vapor and thermoregulated at 32.5°C and 33.5°C (Mortola, 1984; Durand et al., 2004). The flow rate was set to 40 mL/min to avoid CO2 and water accumulation. Breathing recording in adult mice was all done using whole-body plethysmography closed system in a larger chamber (500 mL) and flow rate was set to 500 mL/min. Experiments occurred at 24–26°C room temperature. The animal chamber was connected to a differential pressure transducer and to a preamplifier (FE 141 Spirometer, ADInstruments) to detect pressure oscillations when chamber was completely closed. Volume calibration was performed for each experiment by injecting 0.2–0.5 mL of air into the neonatal and adult chamber. The signal was digitalized using PowerLab system (ADInstruments). The sample rate was set as 1000 Hz and signal were filtered in 0.5–20 Hz bandwidth.

Breathing variables as breath duration (TTOT; s), inspiratory time (TI; s), expiratory time (TE; s), tidal volume (VT; µL/g), respiratory frequency (fR; breaths/min), and ventilation (VE; µL/min/g) were analyzed offline using Lab Chart software (ADInstruments). Tidal volume in whole-body plethysmography was calculated as previously described (Patrone et al., 2018). Minute ventilation was defined by the product of breathing frequency and tidal volume. Breath variability was analyzed by IBI irregularity and it was defined as IBI irregularity = abs (TTOT (n+1) – TTOT (n))/ TTOT (n) (van der Heijden and Zoghbi, 2018). We also used a nonlinear method of analyses known as Poincare map. This method plots breath duration (TTOT) vs. duration of the subsequent breath (TTOT n+1). We used a total of 100 breaths at rest condition. Next, we calculated SD1 and SD2 that describe the distribution of the points in the ellipse using the Kubios software (version 3.5.0) (Brennan et al., 2002). In summary, it was calculated the width of the variation perpendicular to (SD1) and along the line of identity (SD2) from the ellipse that describes the distribution of the points (Brennan et al., 2002).

To quantify breathing parameters, we first calculated the average of 30 s during a stable condition for each animal during normoxia, hypoxia, and hypercapnia. To quantify changes during hypoxia and hypercapnia, we normalized the data to baseline for each animal and then calculate the relative changes expressed as percentage. Spontaneous apnea‐like events or respiratory pause was defined by the cessation of breathing greater than the average of one respiratory cycle to identify possible breathing pattern abnormalities. The duration of each apnea‐like event was from the end of the first breath to the start of the following breath.

Measurements of O2 consumption

Request a detailed protocol

We used an O2 analyzer (ADInstruments) that was connected to the output port of the animal’s head chamber to pull air through the chamber at 100 mL/min for P1-2 and 500 mL/min for adult mice. A mass flow system (MFS, Sable Systems International, Las Vegas, NV, USA) was coupled to the outlet of the whole-body plethysmograph chamber. The outflow from the chamber was dried through a drierite column before passing through the O2 analyzer where O2 fraction in the outflow gas was continuously sampled (1000 Hz) and digitized via PowerLab (ADInstruments/Chart Software, version 7.3). The fractions of oxygen in the inflow (FiO2) and outflow (FeO2) gas were measured using a gas analyzer (model ML206, ADInstuments) that sampled, alternatively from the input and outflow gas ports. O2 consumption (VO2) was calculated based on the formula (Depocas and Hart, 1957): VO2 = [FLo(FiO2 − FeO2)]/1 − FiO2, where FLo is the outlet flow rate; FiO2 is the inflow O2 fraction; FeO2 is the outflow O2 fraction. VO2 was divided by body mass (in g) and the values reported under standard temperature and pressure, dry (STPD).

Histology

Request a detailed protocol

The mice were deeply anesthetized with isoflurane (5% in 100% O2) and heparin was injected intracardially (500 units) and perfused through the ascending aorta with 20 mL of phosphate-buffered saline (PBS 0.1 M) and with 50 mL of 4% paraformaldehyde (in PBS 0.1 M). The brains were kept overnight immersion in 4% paraformaldehyde and then in a 20% sucrose solution. Brain tissues were sectioned in a coronal plane at 30 μm with a sliding microtome and stored in cryoprotectant solution (20% glycerol plus 30% ethylene glycol in 50 mM phosphate buffer, pH 7.4) at –20°C until histological processing. All histochemical procedures were completed using free-floating sections.

For immunofluorescence, the following primary antibodies were used: (a) anti-Phox2b (rabbit anti-Phox2b 1:1000; a gift from JF Brunet, Ecole Normale Supèrieure, Paris, France); (b) anti-TH (mouse anti-TH, 1:1000; Millipore, Burlington, MA, USA); (c) anti-fos (rabbit anti-fos, 1:1000; Santa Cruz Biotechnology, Santa Cruz, CA, USA). All primary antibodies were diluted in PBS containing 2% normal donkey serum (Jackson ImmunoResearch Laboratories) and 0.3% Triton X-100 and were incubated overnight. Sections were subsequently rinsed in PBS and incubated for 2 hr in an appropriate secondary antibody (1:500). The sections were mounted in slides and covered with DPX (Sigma-Aldrich, Milwaukee, WI, USA).

Mapping

Request a detailed protocol

A series of three 30 µm transverse sections through the brainstem were examined for each experiment using a Zeiss AxioImager A1 microscope (Carl Zeiss Microimaging, Thornwood, NY, USA). Images were taken with a Zeiss MRC camera (resolution 1388×1040 pixels). Only cell profiles that included a nucleus were counted and/or mapped bilaterally. Balance and contrast were adjusted to reflect true rendering as much as possible. No other ‘photo retouching’ was performed.

The total number of Phox2b+, Phox2b+/TH+, and Phox2b+/TH- cells in the parafacial/RTN region (between 5.99 and 6.75 mm caudal to Bregma level) was plotted as the mean ± SEM (8 sections/animal). We also analyzed fos+ and TH- cells in the parafacial/RTN region. The neuroanatomical nomenclature employed during experimentation and in this manuscript was defined by the Mouse Brain Atlas from Franklin and Paxinos, 2015.

Experimental protocols

Experiment 1: Effect of Phox2bΔ8 mutation in Atoh1cre-expressing cells on breathing and chemoreflex activation during neonatal phase

Request a detailed protocol

Pups were placed in the plethysmography chambers (head-out or whole-body system) and acclimated 5 min prior to the experiment. To record breath parameters in the whole-body system, the flow was interrupted, and the chamber was closed for 1 min. We recorded a total of 3–5 min of ventilation in room air to determine the baseline. To induce chemoreflex challenge, pups were ventilated during 5 min in hypercapnia (7% CO2, 21% O2, balance N2) or hypoxia (8% O2, balance N2) separated by a 10 min of recovery period (room air). In a separate experiment, we also measure VO2 in neonates to investigate whether any change in body weight and baseline respiratory parameters might be related to changes in metabolic rate.

Experiment 2: Effect of Phox2bΔ8 mutation in Atoh1cre-expressing cells on breathing and chemoreflex activation during adult phase

Request a detailed protocol

Adult mice were familiarized during 30 min in 3 consecutive days in the plethysmography chambers (whole-body system). At the day of the breathing recording, animals were acclimated 30–45 min prior to the experiment. After this acclimation, we recorded 10 min in room air breathing to determine the baseline. Animals were then exposed to hypercapnia or hypoxia during 10 min separated by a 20 min of recovery period in room air. In a separate experiment, we also measure VO2 in adults to investigate whether any change in body weight and baseline respiratory parameters could be related to changes in metabolic rate.

Experiment 3: Anatomical changes induced by Phox2bΔ8 mutation in the parafacial/ RTN region

Request a detailed protocol

To investigate whether Phox2bΔ8 mutation compromised Phox2b expression in the parafacial/RTN neurons, adult mice were anesthetized and perfused transcardially. Next, tissues were processed by immunohistochemistry to identify Phox2b expression and absence of TH (see details in Histology section).

Experiment 4: Effect of hypercapnia on fos expression in the parafacial/RTN neurons induced by Phox2bΔ8 mutation

Request a detailed protocol

To investigate whether Phox2bΔ8 mutation compromised the activation of parafacial/RTN neurons by hypercapnia, we analyze fos expression in adult mice. Animals were habituated in the plethysmography chambers and ventilated in room air (0.5 L/min) during 3 consecutive days. At the day of experiment, mice were acclimated 1 hr prior to the hypercapnic challenge. Then, animals were exposed to hypercapnia (7% CO2, 21% O2, balance N2) for 45 min. After exposure, mice were ventilated for additional 45 min in room air. Finally, animals were anesthetized and perfused transcardially as described above in Histology section. All experiments were conducted between 9:00 a.m. and 3:00 p.m.

Statistical analysis

Request a detailed protocol

Results are presented as mean ± SEM. All statistics were performed using GraphPad Prism (version 9, GraphPad Software), with parametric tests used for normally distributed datasets. Details of specific tests are provided in the legend of each figure. The significance level was set as p<0.05.

Data availability

All data generated or analyzed during this study are included in the manuscript and supporting file; Source Data files have been provided for Figures 1-6.

References

    1. Drorbaug JE
    2. Fenn WO
    (1955)
    A barometric method for measuring ventilation in newborn infants
    Pediatrics 16:81–87.
  19. Book
    1. Franklin K
    2. Paxinos G
    (2015)
    Paxinos and Franklin’s Mouse Brain Stereotaxic Atals Coordinates
    Massachusetts, USA: Academic Press.

Decision letter

  1. Muriel Thoby-Brisson
    Reviewing Editor; CNRS Université de Bordeaux, France
  2. Martin R Pollak
    Senior Editor; Harvard Medical School, United States
  3. Hiroshi Onimaru
    Reviewer; Showa University, Japan
  4. Luis Rodrigo Hernandez-Miranda
    Reviewer; Charite Universitätsmedizin Berlin, Germany

Our editorial process produces two outputs: (i) public reviews designed to be posted alongside the preprint for the benefit of readers; (ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "Phox2b mutation mediated by Atoh1 expression impaired respiratory rhythm and ventilatory responses to hypoxia and hypercapnia" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Martin Pollak as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Hiroshi Onimaru (Reviewer #1); Luis Rodrigo Hernandez-Miranda (Reviewer #3).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

The present manuscript by Ferreira and colleagues is of potential interest to researchers working in the field of neural control of breathing and associated respiratory disorders. This study provides some novel insight into some genetic lesions that may underlie some developmental respiratory pathophysiologies. However, limitations in novelty, methodological approaches and data interpretations prevent acceptance of the paper in its present form.

To sum,

1 – Despite the fact that the authors propose a new genetic strategy to manipulate the development of the RTN in mice, there is a limited level of novelty when compared to previous already published works that have already clearly established the importance of Phox2b and Atoh1 regarding their role in specifying RTN constitutive neurons with their functions in central chemoception and breathing rhythm in general. You should strongly and deeply argue and discuss the differences between your present results and already known data addressing the role of Phox2b and Atoh1 in the RTN function and respiratory disorders.

2 – In addition, there are several methodological aspects and interpretation that deserve attention: the most important one being the practices for plethysmographic recordings in newborn and adult mice that must be performed in correct conditions in order to be able to draw strong and unambiguous conclusions; the identity of cFos-positive cells should be better described (are they indeed Phox2b+?). A fate map of GFP-expressing cells is required, including at different developmental stages. This is crucial as a precedent study has shown mislocation of RTN cells after genetic manipulations, this possibility is by the way not documented, examined or discussed. In addition, the status of other structures co-expressing Phox2b and Atoh1 must be described, as they might also affect breathing behavior.

3 – Illustrations must be ameliorated (Figure 1 is useless, Figure 4 requires traces before gas challenges and back to control, Fig8: the VII nucleus must be Phox2b positive) and additional data should be presented (map of GFP+ cells as stated above, activity of RTN cells during challenges).

Reviewer #1 (Recommendations for the authors):

Authors introduced new strategy of genetic manipulation in mice to reveal functional development of the RTN neurons that may relate to CCHS neuropathology. The methods and results are fairly clear. However, it might be rather difficult and complicated work to discuss the present findings because the results should be compared with those of various previous works including genetic manipulations of the RTN. I think that authors struggled to these issues, but I appreciate if they did more effort for this.

General comments

To understand strategy of this study, it would be required to know precise relationship between Phox2b and Atoh1 expression in the RTN during development: e.g. whether Phox2b expression requires preceding Atoh1 expression or not, whether Atoh1 is not required for Phox2b expression but does for correct localization of Phox2b expressing cells, etc. Although some information was mentioned, please add a brief further explanation for this issue in the Introduction to facilitate understanding of their strategy for general readers (see also comments for Discussion section).

In addition, it would be important to correctly understand differences between genetic removal and induction of (conditional) mutation of Phox2b (or Atoh1). It seems to be rather complicated for general readers to understand these issues and what are new findings of the present study, because various genetic manipulations to investigate the RTN function have been performed.

Specific comments

Abstract

The last sentence, "…of the RTN neurons and are essential for the activation of breathing under hypoxic and hypercapnia condition…": What are essential for the activation? (Same as the last sentence in the Introduction).

Methods

Authors described "All experiments, were conducted between 9:00 A.M. and 3:00 P.M."

How do you think about the sleep-wake cycle condition of mice during measurements? This might be important in considering relation to CCHS.

Results

Although the number of specimens for statistical analysis is described in Figure 2 legend, it is not clear whether the same number could be applied in other figures. For instance, the number in Figure 3 seems to be different from those in Figure 2 (e.g. in neonate control and adult control). Moreover, the specimen number for body weight analysis was not described (Page 10, para 3). Please check the number of all experiments. It would be better to describe them in each figure or in the text.

Section 6: The fos expression was counted in RTN neurons lacking TH immunoreactivity. Is it correct that these cells are Phox2b-positive?

Section 7: The results showed that number of Phox2b expressing cells in mutant mice was about half of that in control mice. Authors should discuss why these results were obtained, because it was previously reported that almost 100% of Atoh1 expressing cells were also Phox2-positive (Dubreuil et al. 2009). Did the present results indicate that 50 % of Phox2b expressing cells in the mutant RTN were Atoh1 independent? Or some of Phox2b△8 mutated cells are remaining?

Section 7: Did authors confirmed that Phox2b-positive and TH-positive cells in the medulla and pons were not affected by this mutation?

Discussion

If authors could summarize possible relationships between Atoh1 and Phox2b (and maybe other related transcription factors) involved in the RTN development, considering previous and present results, it would be very helpful for understanding the present situation and future problem of this field. I strongly recommend that authors give a figure for this purpose.

Reviewer #2 (Recommendations for the authors):

The manner in which the plethysmography for both the neonates and adults is not technically sound. There are several problems with their approaches. For neonates, whole body plethysmography is not an appropriate method to measure tidal volume. As the authors themselves note in their methods section, whole body plethysmography is used "to detected pressure oscillations as a result of changes in temperature promoted by ventilation when chamber was completely closed" [sic]. In neonate plethysmography, P1-3 pups typically equilibrate with their environmental chamber temp. Thus, the waveform is not a derivative of heating and cooling that can be related to tidal volume by applying Bartlett and Tenney corrections but rather a function of airway resistance as the compression in the chamber equilibrates with the rarefaction in the lungs upon inspiration (PMID: 25017785). This phenomenon is also a component of adult respiratory measurements, but less so with a large enough temperature differential (30-33 chamber temp) making the barometric component the predominant feature. Facemask pneumotachography or headout plethysmography can give a more accurate and consistent estimate of tidal volume in neonates.

In adult plethysmography, several necessary measurements were not taken or presented. Body temperatures is not reported, nor are VO2 or VCO2. In plethysmography, especially in chemosensory studies, these are critical measurements that should be taken concurrently with breathing measurements and reported singularly and as VE/VO2. If any metabolic or temperature differences exist compared to control groups, this will have significant impacts on breathing outcomes. Changes in metabolism can drive persistent states of alkalosis and acidosis that would impact responses to chemosensory challenges (though such states may be mitigated by renal compensation). Metabolism is state dependent and plethysmography, even with habituation, is still stressful (PMID: 31178741). As noted by authors and further elaborated by Frappel and Mortola (PMID: 1621857, VO2 and temperature will drop in response to hypoxia. Additionally, poikilocapnic hypoxic hyperventilation (vs isocapnic)) causes a drop in pCO2 that reduces drive to breath (PMID: 23690557). Lastly, the chamber temperature for the adult studies should be held at thermo-neutrality rather than room temperature (30-33C). The cold challenge to mice that is room temperature has confounding effects on drive to breath and metabolism. By measuring adults in thermo-neutral conditions, potential metabolic effects are minimized. Isocapnic hypoxia may also be considered.

Given that there is a reported body weight difference and the authors are using a cold chamber for adults, there very well may be an unappreciated difference in metabolism or in metabolic changes due to changes in state, response to hypoxia, or the cold challenge that impacts the reported outcomes. Metabolic differences may also arise (leading to weight loss) through other Phox2B – Atoh1 overlapping populations perturbed in this model not considered in the manuscript.

The observed phenotypes cannot be exclusively assigned to the RTN. A full assessment of Atoh1 and Phox2b overlap using cumulative fatemapping afforded by the Atoh1_Cre; Phox2bΔ8 model should be reported as other areas of overlap could either impact breathing directly or indirectly through metabolism and stress responses (PMID 8184995). The role of the previously identified Atoh1; Phox2B para and intra – trigeminal neurons should be accounted for in the phenotypes and/or the Atoh1; Phox2B RTN neurons tested in isolation.

Reviewer #3 (Recommendations for the authors):

Ferreira and colleagues provide a novel mouse model (Atoh1Cre,Phox2bdelta8) for the study of the central respiratory chemoreceptor circuit and, therefore, of interest for the respiratory physiology community. Nonetheless, in its present form, this work still lacks more physiological, developmental, and anatomical characterizations to place this study in a broader context and gain new insights into the physiology of respiratory chemoreflexes.

I hope the authors find my below comments of use to enrich their work.

1. The major caveat of this work is that it does not significantly differ from previously published reports using very similar approaches (including a Atoh1Cre,Phox2bflox/flox strategy in Ruffault et al., 2015, eLife DOI: 10.7554/eLife.07051).

2. For today's standards, the display of the alleles (genetic strategy) used in this study cannot occupy a full main figure (Figure 1).

3. The plethysmograph traces presented in Figures 2, 4, 5 and 6 should be accompanied by a period of pre-Gas exposure and post-Gas exposure.

4. Figure 7 and 8 could be combined, and more representative photographs should be presented. The assignment of the facial motor nucleus seems to be arbitrary, as it lacks Phox2b immunoreactive cells. Facial motor neurons do express Phox2b in addition to ChAT in the adult life of mice.

5. The general outline of the manuscript is nothing I have seen before in eLife, that is numbered points for Materials and methods and the result section, although I admit that it helped a lot with the reading.

6. The authors show a reduction (about half) in the number of Phox2b+/TH- cells in adult Atoh1Cre,Phox2bdelta8 mice, and assume that this is indicative of a reduction in the number of retrotrapezoid neurons. This is not necessarily true. I would recommend that the authors present first an anatomical/developmental characterization of retrotrapezoid neurons in Atoh1Cre,Phox2bdelta8 at embryonic and neonatal states (E12.5, E16.5 and P0). At this stages, retrotrapezoid neurons have a well-established molecular signature: Phox2b, Atoh1 and Lbx1 expression. Whereas there are not great commercial Lbx1 and Atoh1 antibodies, the authors could consider combining in situ hybridization for Lbx1 and/or Atoh1 with Phox2b immunoreactivity, the Phox2b antibody that the authors used in this study is great, and compatible with in situ hybridization (my own experience). Other studies by the groups of Huda Zoghbi and Jean-Francois Brunet have shown that interfering with Phox2b and Atoh1 expression in retrotrapezoid neurons results in the incorrect location of this cells dorsally to the facial motor nucleus, is this phenotype also present in Atoh1Cre,Phox2bdelta8 mice?

7. The authors show a reduction of Fos+/TH- cells in adult Atoh1Cre,Phox2bdelta8 mice that were exposed to high levels of CO2 in air. From this result the authors conclude that the number of Fos-activated retrotrapezoid neurons is decreased. Again, this is not necessarily true. To better define this, it is necessary to demonstrate that Fos is not express in Phox2b+/TH- retrotrapezoid neurons. Whereas the Fos and Phox2b antibodies used in this study are both generated in rabbits, the authors could make use of the eGFP expression present in the Phox2bdelta8 allele (Figure 1), therefore, it is necessary to combine the immunoreactivity for eGFP (Phox2b), Fos and TH. This is central for this study, as if indeed less retrotrapezoid neurons are activated by CO2 in adult Atoh1Cre,Phox2bdelta8 mice, it is astonishing that these mice can have a full response to hypercapnia. This is intriguing because other mouse models, in which the number of retrotrapezoid neurons are reduced in greater numbers, do not show a full response to CO2 in the adult life, for instance in: P2b::CreBAC1;Atoh1lox/lox (Ruffault et al., 2015), Egr2cre;P2b27Alacki (Ramanantsoa et al., 2011, DOI: 10.1523/JNEUROSCI.1721-11.2011), Atoh1Phox2bCKO mice (Huang et al., 2017, DOI: 10.1016/j.neuron.2012.06.027) and Egr2cre;Lbx1FS (Hernandez-Miranda et al., 2018, DOI: 10.1073/pnas.1813520115).

8. The authors do not address if retrotrapezoid neurons/parafacial cells are rhythmically active and responsive for pH changes in the embryonic/neonatal life of Atoh1Cre,Phox2bdelta8 animals. The fact that these mice can fully respond to hypercapnia in the adult life but not in the neonatal stages might imply that retrotrapezoid neurons are present but somehow silenced in Atoh1Cre,Phox2bdelta8 neonates. Therefore the anatomical/developmental characterization of retrotrapezoid neurons (as suggested above) should be complemented with in vitro calcium imaging in Atoh1Cre,Phox2bdelta8 embryos or neonates. This could explain why mice that completely lack retrotrapezoid neurons (Egr2cre;Lbx1FS) do not fully display the hypercapnic reflex, whereas Atoh1Cre,Phox2bdelta8 mice do.

https://doi.org/10.7554/eLife.73130.sa1

Author response

Essential revisions:

The present manuscript by Ferreira and colleagues is of potential interest to researchers working in the field of neural control of breathing and associated respiratory disorders. This study provides some novel insight into some genetic lesions that may underlie some developmental respiratory pathophysiologies. However, limitations in novelty, methodological approaches and data interpretations prevent acceptance of the paper in its present form.

We thank the reviewers and the editors for the constructive comments on our manuscript. We have carefully addressed most of the major comments. We modified the manuscript accordingly. We believe that the revised version of our manuscript has been greatly improved and we hope that it will be suitable for publication on the eLife.

To sum,

1 – Despite the fact that the authors propose a new genetic strategy to manipulate the development of the RTN in mice, there is a limited level of novelty when compared to previous already published works that have already clearly established the importance of Phox2b and Atoh1 regarding their role in specifying RTN constitutive neurons with their functions in central chemoception and breathing rhythm in general. You should strongly and deeply argue and discuss the differences between your present results and already known data addressing the role of Phox2b and Atoh1 in the RTN function and respiratory disorders.

We would like to thank the reviewers and editor for the important concern and the opportunity to better discuss the differences between our study and former studies that have already described the role of Phox2b mutation or genetic deletion on chemoreception and breathing regulation. We implemented several modifications in the manuscript to address this question. Briefly, the main novelty of our study was to introduce a humanized NPARM Phox2bΔ8 in Atoh1cre-expressing cells of rodents. Atoh1 is expressed during development in proliferating cells in the rhombic lip and in postmitotic neurons. In this independent site, postmitotic neurons are the only region that co-express Phox2b and Atoh1 surround the paramotor neurons that involves facial motor nucleus (peri VII thus, RTN/parafacial neurons) and trigeminal motor nucleus (periV also known as intertrigeminal region). The present study highlighted some differences compared to deletion of a transcription factor (Phox2b or Atoh1) or introduction of PARM CCHS mutation revealed by previous studies. Surprisingly, (1) Phox2bΔ8 in Atoh1expressing cells did not induce important functional change of baseline respiration in neonates. In contrast, it increased the number of apneas and respiratory irregularity in adult mice. (2) Ventilatory responses to hypoxia and hypercapnia are compromised in neonates. Interestingly, same lack of hypercapnic response have been demonstrated after introducing PARM mutation specific to RTN (DOI:10.1523/JNEUROSCI.1721-11.2011) or by deleting Phox2b from same Phox2b/Atoh1 expressing neurons (periV and peri VII) (DOI: 10.7554/eLife.07051.001). However, the compromised ventilatory response to hypoxia showed by us contrast with the study that introduced PARM mutation in the RTN (DOI:10.1523/JNEUROSCI.1721-11.2011). Furthermore, (3) in adults, the ventilatory response to hypoxia still partially impaired. But they recovered hypercapnic ventilatory response. Despite hypercapnic responses is recovered, the number of Phox2b+/TH- neurons in the RTN/parafacial region reduced by approximately 50%. Additionally, the number of fos+/TH- expressing neurons in the RTN/parafacial region induced by hypercapnia drastically reduced in adults. Interestingly, formers studies that applied different strategies to manipulate Phox2b RTN/parafacial neurons, CO2 response was only partially recovered in the adult life (DOI: 10.7554/eLife.07051; DOI:10.1523/JNEUROSCI.1721-11.2011; http://dx.doi.org/10.1016/j.neuron.2012.06.027; https://doi.org/10.1073/pnas.1813520115). Although, an extensive depletion of RTN/parafacial neurons occurred at embryonic ages. There is no information whether those neurons still depleted during adulthood and whether they are functional. Thus, it complicates further comparison with our finding. Together these results highlighted important differences that certainly imply different mechanisms between our strategy and previous studies. The mechanism by which NPARM Phox2bΔ8 mutation impact ventilatory responses induced by hypoxia and hypercapnia will be focus of future studies in the laboratory.

2 – In addition, there are several methodological aspects and interpretation that deserve attention: the most important one being the practices for plethysmographic recordings in newborn and adult mice that must be performed in correct conditions in order to be able to draw strong and unambiguous conclusions; the identity of cFos-positive cells should be better described (are they indeed Phox2b+?). A fate map of GFP-expressing cells is required, including at different developmental stages. This is crucial as a precedent study has shown mislocation of RTN cells after genetic manipulations, this possibility is by the way not documented, examined or discussed. In addition, the status of other structures co-expressing Phox2b and Atoh1 must be described, as they might also affect breathing behavior.

Thank you for several comments on methodology. We acknowledge the limitations of the barometric plethysmography for the precise measurement of tidal volume. For that reason, breath volumes were normalized to calibrations made during each recording. The amplitude ratio of breaths compared to calibrations was used to determine a value in microliter for each breath. In addition, modifications induced by hypoxia/hypercapnia stimulus was showed as percentage change and not absolute values. To account for differences in body size, breath volumes were also normalized to body weight (μl/g). As a result, we do not expect these limitations to have a significant impact on the interpretation of our data. In addition, to further investigate whether the plethysmography method used would change the former tidal volume results, specialty during baseline conditions where the absolute values were described, we performed head-out plethysmograph system in a subset of neonate (control and mutant). We found that neonate mutants mice presented a slightly increase in tidal volume. However, it did not reach statistic differences, presumably due to the small number of the sample (N = 4/group). In any case, we added the results in the result section and mentioned in the text the importance to use more accurate methods since it has not been used by majority of the studies, including those that we referenced in the present study (Mortola, 1984, Durand et al., 2004; Mortola and Frappell, 2013; Patrone et al., 2018).

In relation to the fos expression, we better described our protocol and clarify the methodology used in the text (methods, results and Discussion section). We used the absence of coexpression with tyrosine hydroxylase, as it has been extensively described to identify RTN activated neurons located ventrally to facial motor nucleus (Stornetta et al., 2006. Barna et al., 2012; 2014; Kumar et al., 2015). Unfortunately, the fos and Phox2b antibodies that we have in the laboratory are made in the same species (i.e, rabbit) and could not be ran together. Furthermore, we investigated RTN/parafacial neurons by analyzing the presence of Phox2b and absence of TH expression. Phox2b+/TH- neurons are well known to be chemosensitive RTN neurons (Stornetta et al., 2006; Onimaru et al., 2008). We found that the number of Phox2b+/TH- neurons greatly reduced in adult mutant mice. Importantly, the neurons were located ventral and laterally to facial motor nucleus. In agreement with previous study that deleted Phox2b from same neuronal region (DOI: 10.7554/eLife.07051.001). More specifically, from Phox2b/Atoh1 post mitotic neurons in the peri V and periVII region. Interestingly, the misslocation only occurred when manipulating Atoh1 neurons (DOI: 10.7554/eLife.07051.001; DOI:https://doi.org/10.1016/j.neuron.2012.06.027 ).

Regarding the fate map of GFP-expressing cells, in concordance with our prior data (DOI 10.1007/s00401-015-1441-0; doi:10.1111/bpa.12877), GFP expression is only detected in embryonic Phox2b expressing cells. The goal of the present study was to analyze respiratory and anatomical effect after birth and not during embryonic phase. For that reason, a fate map showing GFP-expressing cells during different developmental stages was not performed.

3 – Illustrations must be ameliorated (Figure 1 is useless, Figure 4 requires traces before gas challenges and back to control, Fig8: the VII nucleus must be Phox2b positive) and additional data should be presented (map of GFP+ cells as stated above, activity of RTN cells during challenges).

Figure 1 was removed. We edited and added traces showing pre and pos stimulus. Regarding the Phox2b positive cells in the VII nucleus, we do not have a final answer why there is no Phox2b-immunoreactive in the facial motor nucleus. However, previous experiments from our lab (PMID: 22015927, 24286756, 24363384, 27633663) and from different labs (PMID: 17021186, 17255166, 18440993, 26068853, 29066557) did not find Phox2b-expressing neurons in the facial motor nucleus of postnatal or adult rodents.

Reviewer #1 (Recommendations for the authors):

Authors introduced new strategy of genetic manipulation in mice to reveal functional development of the RTN neurons that may relate to CCHS neuropathology. The methods and results are fairly clear. However, it might be rather difficult and complicated work to discuss the present findings because the results should be compared with those of various previous works including genetic manipulations of the RTN. I think that authors struggled to these issues, but I appreciate if they did more effort for this.

We agree with the reviewer. In the present version of the manuscript, we better discussed our data in comparison with former studies. We rewrote the manuscript and most of the information was added in the introduction and Discussion section. Briefly, the main novelty of our study was to introduce a humanized NPARM Phox2bΔ8 in Atoh1cre-expressing cells of rodents. Atoh1 is expressed during development in proliferating cells in the rhombic lip and in postmitotic neurons. In this independent site, postmitotic neurons are the only region that co-express Phox2b and Atoh1 surround the paramotor neurons that involves facial motor nucleus (peri VII thus, RTN/parafacial neurons) and trigeminal motor nucleus (periV also known as intertrigeminal region). The present study highlighted some differences compared to deletion of a transcription factor (Phox2b or Atoh1) or introduction of PARM CCHS mutation revealed by previous studies.

Surprisingly, (1) Phox2bΔ8 in Atoh1-expressing cells did not induce important functional change of baseline respiration in neonates. In contrast, it increased the number of apneas and respiratory irregularity in adult mice. (2) Ventilatory responses to hypoxia and hypercapnia are compromised in neonates. Interestingly, same lack of hypercapnic response have been demonstrated after introducing PARM mutation specific to RTN (DOI:10.1523/JNEUROSCI.1721-11.2011) or by deleting Phox2b from same Phox2b/Atoh1 expressing neurons (periV and peri VII) (DOI: 10.7554/eLife.07051.001). However, the compromised ventilatory response to hypoxia showed by us contrast with the study that introduced PARM mutation in the RTN (DOI:10.1523/JNEUROSCI.1721-11.2011). Furthermore, (3) in adults, the ventilatory response to hypoxia still partially impaired. But they recovered hypercapnic ventilatory response. Despite hypercapnic responses is recovered, the number of Phox2b+/TH- neurons in the RTN/parafacial region reduced by approximately 50%. Additionally, the number of fos+/TH- expressing neurons in the RTN/parafacial region induced by hypercapnia drastically reduced in adults. Interestingly, formers studies that applied different strategies to manipulate Phox2b RTN/parafacial neurons, CO2 response was only partially recovered in the adult life (DOI: 10.7554/eLife.07051; DOI:10.1523/JNEUROSCI.1721-11.2011; http://dx.doi.org/10.1016/j.neuron.2012.06.027; https://doi.org/10.1073/pnas.1813520115). Although, an extensive depletion of RTN/parafacial neurons occurred at embryonic ages. There is no information whether those neurons still depleted during adulthood and whether they are functional. Thus, it complicates further comparison with our finding. Together these results highlighted important differences that certainly imply different mechanisms between our strategy and previous studies. The mechanism by which NPARM Phox2bΔ8 mutation impact ventilatory responses induced by hypoxia and hypercapnia will be focus of future studies in the laboratory.

General comments

To understand strategy of this study, it would be required to know precise relationship between Phox2b and Atoh1 expression in the RTN during development: e.g. whether Phox2b expression requires preceding Atoh1 expression or not, whether Atoh1 is not required for Phox2b expression but does for correct localization of Phox2b expressing cells, etc. Although some information was mentioned, please add a brief further explanation for this issue in the Introduction to facilitate understanding of their strategy for general readers (see also comments for Discussion section).

In the present study, we introduced the humanized NPARM Phox2bΔ8 in Atoh1expressing cells of rodents. Atoh1 in the brainstem is expressed during development in proliferating cells in the rhombic lip (at early stages) and in postmitotic neurons (during late stage) independently. The postmitotic neurons are the only region that co-express Phox2b and Atoh1 surround the paramotor neurons that involves facial motor nucleus (peri VII thus, RTN/parafacial neurons) and trigeminal motor nucleus (periV also known as intertrigeminal region) (DOI: 10.7554/eLife.07051.001). Thus, NPARM Phox2bΔ8 was introduced in the periVII and periV region. Our goal with this strategy was to investigate the effect of the NPARM Phox2bΔ8 in regions that are well known to be involved with respiratory control and central chemosensitivity. The idea was to induce late activation of NPARM Phox2bΔ8 and, then analyze the effect on respiratory function and neuroanatomy of RTN after birth (neonatal and adulthood). The results in the present study showing immunoreactivity to Phox2b, and absence of tyrosine hydroxylase strongly indicate the phenotype of RTN/parafacial chemosensitive neurons. Furthermore, we showed that when the mutation is activated in the postmitotic Phox2b/Atoh1 neurons, it greatly impacts RTN/parafacial neurons (reduction in approximately 50% compared to controls). Deletion of Phox2b using similar strategy (from same periV and periVII region) reduced Phox2b+/Atoh1+ expression by 60% during embryonic age 18.5.

We also added more information in the Introduction and Discussion to be clear for general readers.

In addition, it would be important to correctly understand differences between genetic removal and induction of (conditional) mutation of Phox2b (or Atoh1). It seems to be rather complicated for general readers to understand these issues and what are new findings of the present study, because various genetic manipulations to investigate the RTN function have been performed.

We apologize for the confusion and have better described the genetic approach used. The present study used a conditional “humanized” Phox2b NPARM mutation inspired by a proband previously reported by Dr. Otero´s group. This locus includes a human exon 3 mutation with an 8-nucleotide deletion. It is not induced by Atoh1. Rather, cells that express Atoh1 undergo the cre-mediated recombination at the engineered Phox2b locus and will express the NPARM phox2b only when the gene is normally expressed. Phox2b shows expression early in mammalian development, and therefore this construct provides for the study of NPARM Phox2b effects in respiratory nuclei after initial embryonic Phox2b specification.

Specific comments

Abstract

The last sentence, "…of the RTN neurons and are essential for the activation of breathing under hypoxic and hypercapnia condition…": What are essential for the activation? (Same as the last sentence in the Introduction).

We modified the entire Abstract section.

Methods

Authors described "All experiments, were conducted between 9:00 A.M. and 3:00 P.M."

How do you think about the sleep-wake cycle condition of mice during measurements? This might be important in considering relation to CCHS.

Despite we have done all recordings during light cycle where the percentage of sleep is higher, unfortunately, we did not record it concomitant with respiration. So, we cannot guarantee if animals were sleeping or not during the breathing recordings. Some preliminary data from our lab (not shown in this manuscript) found that percentage of sleep-awake cycle was similar between control and mutated mice. We are planning to better evaluate in a future study a comparison of respiratory control between wake and sleep state in both control and mutant mice (Phox2bΔ8 mutation in Atoh1cre).

Results

Although the number of specimens for statistical analysis is described in Figure 2 legend, it is not clear whether the same number could be applied in other figures. For instance, the number in Figure 3 seems to be different from those in Figure 2 (e.g. in neonate control and adult control). Moreover, the specimen number for body weight analysis was not described (Page 10, para 3). Please check the number of all experiments. It would be better to describe them in each figure or in the text.

We acknowledged the observation. In the present version, we checked all the numbers and statistics. We apologized for the mistake. In addition, we found that the mean of inter-breath intervals was done incorrectly, i.e., not converting to absolute values. So, we also correct it in the present version of the manuscript, figure and excel sheet. To clarify, none of the results or statistics had been changed after correcting those mistakes.

Section 6: The fos expression was counted in RTN neurons lacking TH immunoreactivity. Is it correct that these cells are Phox2b-positive?

We counted the number of fos+ neurons in the RTN/parafacial region that lack TH immunoreactivity. It is likely that these cells also express Phox2b because in this region we have the bulk of Phox2b-expressing neurons. We did not perform Phox2b and fos because our antibodies are made in rabbits. We avoid the TH cells, which also express Phox2b but it is not the chemoreceptors RTN neurons.

Section 7: The results showed that number of Phox2b expressing cells in mutant mice was about half of that in control mice. Authors should discuss why these results were obtained, because it was previously reported that almost 100% of Atoh1 expressing cells were also Phox2-positive (Dubreuil et al. 2009). Did the present results indicate that 50 % of Phox2b expressing cells in the mutant RTN were Atoh1 independent? Or some of Phox2b△8 mutated cells are remaining?

In the present version, we did not analyze the percentage of Atoh1 cells that colocalize with Phox2b in the RTN. Previous studies that used in situ hybridization showed that approximately all Phox2b neurons of the RTN also express Atoh1 (PMID: 22958821, 25866925). We used the cre line as a tool to introduce the conditional Phox2b mutation during Atoh1 expression. Thus, cells that express Atoh1 undergo the cre-mediated recombination at the engineered Phox2b locus and will express the NPARM phox2b only when the gene is normally expressed. Phox2b shows expression early in mammalian development, and therefore this construct provides for the study of NPARM Phox2b effects in respiratory nuclei after initial embryonic Phox2b specification.

Section 7: Did authors confirmed that Phox2b-positive and TH-positive cells in the medulla and pons were not affected by this mutation?

Yes, we ran a new series of experiments and noticed that Phox2b+ neurons in the pons as well as the number of TH cells in the A1, A2, A6, and C1 were not affected by the mutation.

Discussion

If authors could summarize possible relationships between Atoh1 and Phox2b (and maybe other related transcription factors) involved in the RTN development, considering previous and present results, it would be very helpful for understanding the present situation and future problem of this field. I strongly recommend that authors give a figure for this purpose.

Thank you for the suggestion. In the present version of the manuscript, we made a summary figure illustrating what we have in the literature and what our work added to the field of breathing control.

Reviewer #2 (Recommendations for the authors):

The manner in which the plethysmography for both the neonates and adults is not technically sound. There are several problems with their approaches. For neonates, whole body plethysmography is not an appropriate method to measure tidal volume. As the authors themselves note in their methods section, whole body plethysmography is used "to detected pressure oscillations as a result of changes in temperature promoted by ventilation when chamber was completely closed" [sic]. In neonate plethysmography, P1-3 pups typically equilibrate with their environmental chamber temp. Thus, the waveform is not a derivative of heating and cooling that can be related to tidal volume by applying Bartlett and Tenney corrections but rather a function of airway resistance as the compression in the chamber equilibrates with the rarefaction in the lungs upon inspiration (PMID: 25017785). This phenomenon is also a component of adult respiratory measurements, but less so with a large enough temperature differential (30-33 chamber temp) making the barometric component the predominant feature. Facemask pneumotachography or headout plethysmography can give a more accurate and consistent estimate of tidal volume in neonates.

We acknowledge the limitations of the barometric plethysmography for precise measurement of tidal volume in neonates used in the present study. For that reason, in a subset of control and mutant neonate mice we analyze respiration using head-out plethysmograph. Despite the small sample size, we found that tidal volume was slightly higher in the mutant compared to controls littermate. We added more information in the material and methods ((page 5; 2) Ventilation measurements), results (page 11, para 1st), and Discussion section (page 16, para 2nd). Additionally, in the experiments using whole body plethysmograph, breath volumes were normalized to calibrations made during each recording. The amplitude ratio of breaths compared to calibrations was used to determine a value in microliter for each breath. More importantly, changes induced by hypoxia/hypercapnia stimulus were showed as percentage change and not absolute values. To account for differences in body size, breath volumes were also normalized to body weight (μl/g). As a result, we do not expect that these limitations significantly impact the interpretation of our data.

In adult plethysmography, several necessary measurements were not taken or presented. Body temperatures is not reported, nor are VO2 or VCO2. In plethysmography, especially in chemosensory studies, these are critical measurements that should be taken concurrently with breathing measurements and reported singularly and as VE/VO2. If any metabolic or temperature differences exist compared to control groups, this will have significant impacts on breathing outcomes. Changes in metabolism can drive persistent states of alkalosis and acidosis that would impact responses to chemosensory challenges (though such states may be mitigated by renal compensation). Metabolism is state dependent and plethysmography, even with habituation, is still stressful (PMID: 31178741). As noted by authors and further elaborated by Frappel and Mortola (PMID: 1621857, VO2) and temperature will drop in response to hypoxia. Additionally, poikilocapnic hypoxic hyperventilation (vs isocapnic) causes a drop in pCO2 that reduces drive to breath (PMID: 23690557). Lastly, the chamber temperature for the adult studies should be held at thermo-neutrality rather than room temperature (30-33C). The cold challenge to mice that is room temperature has confounding effects on drive to breath and metabolism. By measuring adults in thermo-neutral conditions, potential metabolic effects are minimized. Isocapnic hypoxia may also be considered.

Thank you for pointing out important issues in our previous manuscript. We agree with the reviewer. In fact, some information and misconception were in the previous version. Now, we added the correct way in which the respiratory parameters were measured in both neonate and adult mice (Material and methods section; page 5 – ventilation measurements). We also measured VO2 and VE/VO2 in control and mutant mice during neonatal and adult phase. Despite the slightly reduction in body weight in the mutants during adulthood, oxygen consumption was not different between groups. Those results indicate that NPARM Phox2bΔ8 in Atoh1cre-expressing cells did not impact metabolism.

Given that there is a reported body weight difference and the authors are using a cold chamber for adults, there very well may be an unappreciated difference in metabolism or in metabolic changes due to changes in state, response to hypoxia, or the cold challenge that impacts the reported outcomes. Metabolic differences may also arise (leading to weight loss) through other Phox2B – Atoh1 overlapping populations perturbed in this model not considered in the manuscript.

As mentioned above some information and misconception were in the previous version. We performed additional experiments to analyze oxygen consumption. We found that mutation did not change VO2 and VE/VO2 in neonates nor in adults compare to controls. Thus, the reduction in body weight in adult mutant cannot be assigned to changes in the metabolism.

Unfortunately, most of the studies that genetically deleted Phox2b/Atoh1 (PMID: 25866925; PMID: 22958821; PMID: 29972353) did not report body weight. However, when PARM Phox2b mutation was induced specific to RTN neurons (PMID: 21900566), i.e., without directly manipulating periV region, mutated neonate (P9) had lower body weight when compared to controls. Nevertheless, oxygen consumption was not evaluated by the former study.

The observed phenotypes cannot be exclusively assigned to the RTN. A full assessment of Atoh1 and Phox2b overlap using cumulative fatemapping afforded by the Atoh1_Cre; Phox2bΔ8 model should be reported as other areas of overlap could either impact breathing directly or indirectly through metabolism and stress responses (PMID 8184995). The role of the previously identified Atoh1; Phox2B para and intra – trigeminal neurons should be accounted for in the phenotypes and/or the Atoh1; Phox2B RTN neurons tested in isolation.

We would like to thank the reviewer for the excellent point raised. Regarding the stress response, the study cited (PMID 8184995) demonstrated that children with CCHS had lower levels of anxiety compared to all other children, including those with asthma. However, there is no information about CCHS form (PARM or NPARM) included in the group. It will be an interesting topic to investigate in the future.

In relation to the cumulative fate mapping, we had several problems with our Phox2bΔ8; Atoh1cre colony which prevented us to continue the experiments to show differences in the expression of Phox2b within the para or intraregional regions. We are trying to recover the colony and future experiments will be carried out to account for the phenotypes asked by the reviewer.

Reviewer #3 (Recommendations for the authors):

Ferreira and colleagues provide a novel mouse model (Atoh1Cre,Phox2bdelta8) for the study of the central respiratory chemoreceptor circuit and, therefore, of interest for the respiratory physiology community. Nonetheless, in its present form, this work still lacks more physiological, developmental, and anatomical characterizations to place this study in a broader context and gain new insights into the physiology of respiratory chemoreflexes.

I hope the authors find my below comments of use to enrich their work.

In the new version of the manuscript, we have made significant changes as suggested by the reviewer. For example, we reorganize the anatomical data including new analysis and performed new functional experiments to strengthen our work. We are very enthusiastic about our reviewed version, and we believe it will open new questions that need to be addressed in future studies

1. The major caveat of this work is that it does not significantly differ from previously published reports using very similar approaches (including a Atoh1Cre,Phox2bflox/flox strategy in Ruffault et al., 2015, eLife DOI: 10.7554/eLife.07051).

We would like to thank the reviewer for the important concerns and the opportunity to better discuss the differences between our study and former studies that have already described the role of Phox2b mutation or genetic deletion on chemoreception and breathing regulation. We implemented several modifications in the manuscript to address this question. We rewrote the manuscript and most of the information was added in the introduction and Discussion section. Briefly, the main novelty of our study was to introduce a humanized NPARM Phox2bΔ8 in Atoh1creexpressing cells of rodents. Atoh1 is expressed during development in proliferating cells in the rhombic lip and in postmitotic neurons. In this independent site, postmitotic neurons are the only region that co-express Phox2b and Atoh1 surround the paramotor neurons that involves facial motor nucleus (peri VII thus, RTN/parafacial neurons) and trigeminal motor nucleus (periV also known as intertrigeminal region). The present study highlighted some differences compared to deletion of a transcription factor (Phox2b or Atoh1) or introduction of PARM CCHS mutation revealed by previous studies. Surprisingly, (1) Phox2bΔ8 in Atoh1-expressing cells did not induce important functional change of baseline respiration in neonates. In contrast, it increased the number of apneas and respiratory irregularity in adult mice. (2) Ventilatory responses to hypoxia and hypercapnia are compromised in neonates. Interestingly, similar lack of hypercapnic response have been demonstrated after introducing PARM mutation specific to RTN (DOI:10.1523/JNEUROSCI.1721-11.2011) or by deleting Phox2b from same Phox2b/Atoh1 expressing neurons (periV and peri VII) (DOI: 10.7554/eLife.07051.001). However, the compromised ventilatory response to hypoxia showed by us contrast with the study that introduced PARM mutation in the RTN (DOI:10.1523/JNEUROSCI.1721-11.2011). Furthermore, (3) in adults, the ventilatory response to hypoxia still partially impaired. But they recovered hypercapnic ventilatory response. Despite hypercapnic responses is recovered, the number of Phox2b+/TH- neurons in the RTN/parafacial region reduced by approximately 50%. Additionally, the number of fos+/TH- expressing neurons in the RTN/parafacial region induced by hypercapnia drastically reduced in adults. Interestingly, formers studies that applied different strategies to manipulate Phox2b RTN/parafacial neurons, CO2 response was only partially recovered in the adult life (DOI: 10.7554/eLife.07051; DOI:10.1523/JNEUROSCI.1721-11.2011; http://dx.doi.org/10.1016/j.neuron.2012.06.027; https://doi.org/10.1073/pnas.1813520115). Although, an extensive depletion of RTN/parafacial neurons occurred at embryonic ages. There is no information whether those neurons still depleted during adulthood and whether they are functional. Thus, it complicates further comparison with our finding. Together these results highlighted important differences that certainly imply different mechanisms between our strategy and previous studies. The mechanism by which NPARM Phox2bΔ8 mutation impact ventilatory responses induced by hypoxia and hypercapnia will be focus of future studies in the laboratory.

2. For today's standards, the display of the alleles (genetic strategy) used in this study cannot occupy a full main figure (Figure 1).

Figure 1 was removed.

3. The plethysmograph traces presented in Figures 2, 4, 5 and 6 should be accompanied by a period of pre-Gas exposure and post-Gas exposure.

Pre-gas exposure and post-gas exposure was added in Figures: 3 and 4. Additionally, all figures were restructured.

4. Figure 7 and 8 could be combined, and more representative photographs should be presented. The assignment of the facial motor nucleus seems to be arbitrary, as it lacks Phox2b immunoreactive cells. Facial motor neurons do express Phox2b in addition to ChAT in the adult life of mice.

We added more representative figures, as well as control experiments. For that reason, we keep two figures for the anatomical representations. We do not have a final answer why we did not have an Phox2b-immunoreactive in the facial motor nucleus. However, previous experiments from our lab (PMID: 22015927, 24286756, 24363384, 27633663) and from different labs (PMID: 17021186, 17255166, 18440993, 26068853, 29066557) performed in postnatal/adult mice did not found Phox2b-expressing neurons in the facial motor nucleus. That are in contrast to what have been showed during embryonic stages.

5. The general outline of the manuscript is nothing I have seen before in eLife, that is numbered points for Materials and methods and the result section, although I admit that it helped a lot with the reading.

Thank you for the positive comments.

6. The authors show a reduction (about half) in the number of Phox2b+/TH- cells in adult Atoh1Cre,Phox2bdelta8 mice, and assume that this is indicative of a reduction in the number of retrotrapezoid neurons. This is not necessarily true. I would recommend that the authors present first an anatomical/developmental characterization of retrotrapezoid neurons in Atoh1Cre,Phox2bdelta8 at embryonic and neonatal states (E12.5, E16.5 and P0). At this stages, retrotrapezoid neurons have a well-established molecular signature: Phox2b, Atoh1 and Lbx1 expression. Whereas there are not great commercial Lbx1 and Atoh1 antibodies, the authors could consider combining in situ hybridization for Lbx1 and/or Atoh1 with Phox2b immunoreactivity, the Phox2b antibody that the authors used in this study is great, and compatible with in situ hybridization (my own experience). Other studies by the groups of Huda Zoghbi and Jean-Francois Brunet have shown that interfering with Phox2b and Atoh1 expression in retrotrapezoid neurons results in the incorrect location of this cells dorsally to the facial motor nucleus, is this phenotype also present in Atoh1Cre,Phox2bdelta8 mice?

The Phox2b+/TH- staining to characterize chemosensitive neurons in the RTN/parafacial region has been extensively demonstrated (PMID: 26068853; 29066557). Additionally, in vitro, and anatomical data showed that Phox2b+/TH- neurons in the RTN responded to CO2/H+ and majority express TASK-2 and GPR4 receptor. Respectively, a H+inhibited background K+ channel, and a H+-activated G-protein-coupled receptor, that strongly suggest mediating the chemosensitivity of RTN/parafacial neurons. Thus, it is unlikely that the population showed in our study did not involve chemosensitive neurons in the RTN/parafacial region. In addition, we used a second strategy, Fos+/TH- immunostaining, to investigate the RTN/parafacial neuronal activation induced by hypercapnia and to avoid the C1 adjacent neurons. Fos+/TH- neurons expression reduced by 50% in adult mutated mice compared to control. Together, these results indicate that the reduction in Phox2b+/TH- neurons in the mutant mice greatly reduced the number of functional chemosensitive neurons from RTN/parafacial region.

We understand the importance to describe the anatomical characterization of RTN/parafacial neurons at embryonic stages, but it was not the main goal of the present study. We did not analyze any functional respiratory parameters during embryonic stages, as most studies have already done. We focus in characterize functional and anatomical changes induced by the NPARM Phox2bΔ8 mutation in a mature system. As it corresponds with the timeline when CCHS is detected, to investigate how the anatomical changes could correlate with functional changes when the mutation is activated at late embryonic stages in a restricted neuronal population (periVII and periV). From now, to better understand the mechanisms we will consider the suggestions in the future studies.

The conditional NPARM Phox2bΔ8 in Atoh1-expressing cells did not result in miss location of Phox2b neurons dorsally to facial motor nucleus. Our study used a similar strategy that Ruffault et al. (2015) used in their study. The difference is that instead activated the NPARM Phox2bΔ8, they deleted Phox2b from Atoh1-expressing cells (Atoh1Cre;Phox2blox/lox). Interestingly, improper location seems to depend directly on Atoh1. Since, deletion of Atoh1 from same population (P2b::CreBAC1;Atoh1lox/lox) shifted Phox2b neurons dorsally. In addition, similar results occurred when Atoh1 was deleted within the HoxA4 domain (PMID: 22958821).

7. The authors show a reduction of Fos+/TH- cells in adult Atoh1Cre,Phox2bdelta8 mice that were exposed to high levels of CO2 in air. From this result the authors conclude that the number of Fos-activated retrotrapezoid neurons is decreased. Again, this is not necessarily true. To better define this, it is necessary to demonstrate that Fos is not express in Phox2b+/TH- retrotrapezoid neurons. Whereas the Fos and Phox2b antibodies used in this study are both generated in rabbits, the authors could make use of the eGFP expression present in the Phox2bdelta8 allele (Figure 1), therefore, it is necessary to combine the immunoreactivity for eGFP (Phox2b), Fos and TH. This is central for this study, as if indeed less retrotrapezoid neurons are activated by CO2 in adult Atoh1Cre,Phox2bdelta8 mice, it is astonishing that these mice can have a full response to hypercapnia. This is intriguing because other mouse models, in which the number of retrotrapezoid neurons are reduced in greater numbers, do not show a full response to CO2 in the adult life, for instance in: P2b::CreBAC1;Atoh1lox/lox (Ruffault et al., 2015), Egr2cre;P2b27Alacki (Ramanantsoa et al., 2011, DOI: 10.1523/JNEUROSCI.1721-11.2011), Atoh1Phox2bCKO mice (Huang et al., 2017, DOI: 10.1016/j.neuron.2012.06.027) and Egr2cre;Lbx1FS (Hernandez-Miranda et al., 2018, DOI: 10.1073/pnas.1813520115).

The Phox2b and fos antibody that we currently used in the laboratory are both made in the same species (rabbit). Thus, to further avoid any misinterpretation by counting C1 adjacent neurons that also could be activated by hypercapnia. We applied the absence of TH- immunostaining together with anatomical well-defined region (ventral and laterally to facial nucleus) as a criterion to define RTN/parafacial neurons. We also added more information to the manuscript to be clear to the reader. Regarding the use of GFP, in concordance with our prior data (Nobuta et al., Alzate et al.), GFP expression is only detected in embryonic Phox2b expressing cells. Although the most likely reason for this is NPARM Phox2b induced lethality, we acknowledge that other possibilities such as incomplete cre-mediated recombination or Phox2b mutation self-repression may mediate these effects (PMID: 2310352; 17765533).

As far as the above studies mentioned by the reviewer, the RTN/parafacial neurons investigated were done only during embryonic stages. Therefore, we did not know if in adult, the number of neurons would be comparable to the embryonic age showed to further discusses with the data presented by our study. In addition, the former study, they did not perform any experiment to investigate whether the remaining parafacial/RTN neurons were functional or not.

8. The authors do not address if retrotrapezoid neurons/parafacial cells are rhythmically active and responsive for pH changes in the embryonic/neonatal life of Atoh1Cre,Phox2bdelta8 animals. The fact that these mice can fully respond to hypercapnia in the adult life but not in the neonatal stages might imply that retrotrapezoid neurons are present but somehow silenced in Atoh1Cre,Phox2bdelta8 neonates. Therefore the anatomical/developmental characterization of retrotrapezoid neurons (as suggested above) should be complemented with in vitro calcium imaging in Atoh1Cre,Phox2bdelta8 embryos or neonates. This could explain why mice that completely lack retrotrapezoid neurons (Egr2cre;Lbx1FS) do not fully display the hypercapnic reflex, whereas Atoh1Cre,Phox2bdelta8 mice do.

Thank you for your comments, which are essential to complement our study. However, in our laboratory, we do not have the option of carrying out calcium imaging experiments. Future projects and experiments and maybe some collaborations are being proposed, so that we can answer this question, as well as others.

https://doi.org/10.7554/eLife.73130.sa2

Article and author information

Author details

  1. Caroline B Ferreira

    Department of Pharmacology, Instituto de Ciencias Biomedicas, Universidade de Sao Paulo, Sao Paulo, Brazil
    Contribution
    Conceptualization, Data curation, Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared

  2. Talita M Silva

    Department of Physiology and Biophysics, Instituto de Ciencias Biomedicas, University of Sao Paulo, Sao Paulo, Brazil
    Contribution
    Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared

  3. Phelipe E Silva

    Department of Physiology and Biophysics, Instituto de Ciencias Biomedicas, University of Sao Paulo, Sao Paulo, Brazil
    Contribution
    Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared

  4. Claudio L Castro

    Department of Physiology and Biophysics, Instituto de Ciencias Biomedicas, University of Sao Paulo, Sao Paulo, Brazil
    Contribution
    Data curation, Formal analysis, Investigation, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared

  5. Catherine Czeisler

    Department of Pathology, College of Medicine, Ohio State University, Columbus, United States
    Contribution
    Conceptualization, Resources, Funding acquisition, Project administration, Writing - review and editing
    Competing interests
    No competing interests declared

  6. José J Otero

    Department of Pathology, College of Medicine, Ohio State University, Columbus, United States
    Contribution
    Conceptualization, Resources, Data curation, Software, Formal analysis, Funding acquisition, Project administration, Writing - review and editing
    Competing interests
    No competing interests declared

  7. Ana C Takakura

    Department of Pharmacology, Instituto de Ciencias Biomedicas, Universidade de Sao Paulo, Sao Paulo, Brazil
    Contribution
    Conceptualization, Resources, Data curation, Software, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    takakura@icb.usp.br
    Competing interests
    No competing interests declared

  8. Thiago S Moreira

    Department of Physiology and Biophysics, Instituto de Ciencias Biomedicas, University of Sao Paulo, Sao Paulo, Brazil
    Contribution
    Conceptualization, Resources, Data curation, Software, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    tmoreira@icb.usp.br
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-9789-8296

Funding

Fundação de Amparo à Pesquisa do Estado de São Paulo (2009/01236-4)

  • Ana C Takakura

Fundação de Amparo à Pesquisa do Estado de São Paulo (2015/23376-1)

  • Thiago S Moreira

NHLBI Division of Intramural Research (RO1HL132355)

  • José J Otero

Conselho Nacional de Desenvolvimento Científico e Tecnológico (302334/2019-0)

  • Thiago S Moreira

Conselho Nacional de Desenvolvimento Científico e Tecnológico (302288/2019-8)

  • Ana C Takakura

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

This research work was supported by public funding from São Paulo Research Foundation (FAPESP) (Grants: 2019/01236-4 to ACT and 2015/23376-1 to TSM), and by funds from FAPESP fellowship (2017/12678-2 to TMS and 2019/20990-1 to PES), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) grant (408647/2018-3 to ACT) and fellowships (302334/2019-0 to TSM and 302288/2019-8 to ACT) and NHLBI/NIH (Grant: R01HL132355 to CMC and JJO). This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior-Brasil (CAPES) – Finance Code 001.

Ethics

This study was conducted in accordance with the University of Sao Paulo Institutional Animal Care and Use Committee guidelines (protocol number: 3618221019).

Senior Editor

  1. Martin R Pollak, Harvard Medical School, United States

Reviewing Editor

  1. Muriel Thoby-Brisson, CNRS Université de Bordeaux, France

Reviewers

  1. Hiroshi Onimaru, Showa University, Japan
  2. Luis Rodrigo Hernandez-Miranda, Charite Universitätsmedizin Berlin, Germany

Version history

  1. Preprint posted: August 10, 2021 (view preprint)
  2. Received: August 17, 2021
  3. Accepted: November 14, 2022
  4. Accepted Manuscript published: November 17, 2022 (version 1)
  5. Version of Record published: November 23, 2022 (version 2)

Copyright

© 2022, Ferreira et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 503
    Page views
  • 92
    Downloads
  • 2
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Caroline B Ferreira
  2. Talita M Silva
  3. Phelipe E Silva
  4. Claudio L Castro
  5. Catherine Czeisler
  6. José J Otero
  7. Ana C Takakura
  8. Thiago S Moreira
(2022)
Phox2b mutation mediated by Atoh1 expression impaired respiratory rhythm and ventilatory responses to hypoxia and hypercapnia
eLife 11:e73130.
https://doi.org/10.7554/eLife.73130

Categories and tags

Research organism

Further reading

    1. Genetics and Genomics
    2. Neuroscience

    Translation of dipeptide repeat proteins in C9ORF72 ALS/FTD through unique and redundant AUG initiation codons

    Yoshifumi Sonobe, Soojin Lee ... Paschalis Kratsios
    Research Article Updated

    A hexanucleotide repeat expansion in C9ORF72 is the most common genetic cause of amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). A hallmark of ALS/FTD pathology is the presence of dipeptide repeat (DPR) proteins, produced from both sense GGGGCC (poly-GA, poly-GP, poly-GR) and antisense CCCCGG (poly-PR, poly-PG, poly-PA) transcripts. Translation of sense DPRs, such as poly-GA and poly-GR, depends on non-canonical (non-AUG) initiation codons. Here, we provide evidence for canonical AUG-dependent translation of two antisense DPRs, poly-PR and poly-PG. A single AUG is required for synthesis of poly-PR, one of the most toxic DPRs. Unexpectedly, we found redundancy between three AUG codons necessary for poly-PG translation. Further, the eukaryotic translation initiation factor 2D (EIF2D), which was previously implicated in sense DPR synthesis, is not required for AUG-dependent poly-PR or poly-PG translation, suggesting that distinct translation initiation factors control DPR synthesis from sense and antisense transcripts. Our findings on DPR synthesis from the C9ORF72 locus may be broadly applicable to many other nucleotide repeat expansion disorders.

    1. Cell Biology
    2. Neuroscience

    Removal of extracellular human amyloid beta aggregates by extracellular proteases in C. elegans

    Elisabeth Jongsma, Anita Goyala ... Collin Yvès Ewald
    Research Article Updated

    The amyloid beta (Aβ) plaques found in Alzheimer’s disease (AD) patients’ brains contain collagens and are embedded extracellularly. Several collagens have been proposed to influence Aβ aggregate formation, yet their role in clearance is unknown. To investigate the potential role of collagens in forming and clearance of extracellular aggregates in vivo, we created a transgenic Caenorhabditis elegans strain that expresses and secretes human Aβ1-42. This secreted Aβ forms aggregates in two distinct places within the extracellular matrix. In a screen for extracellular human Aβ aggregation regulators, we identified different collagens to ameliorate or potentiate Aβ aggregation. We show that a disintegrin and metalloprotease a disintegrin and metalloprotease 2 (ADM-2), an ortholog of ADAM9, reduces the load of extracellular Aβ aggregates. ADM-2 is required and sufficient to remove the extracellular Aβ aggregates. Thus, we provide in vivo evidence of collagens essential for aggregate formation and metalloprotease participating in extracellular Aβ aggregate removal.

    1. Computational and Systems Biology
    2. Neuroscience

    Task-dependent optimal representations for cerebellar learning

    Marjorie Xie, Samuel P Muscinelli ... Ashok Litwin-Kumar
    Research Article Updated

    The cerebellar granule cell layer has inspired numerous theoretical models of neural representations that support learned behaviors, beginning with the work of Marr and Albus. In these models, granule cells form a sparse, combinatorial encoding of diverse sensorimotor inputs. Such sparse representations are optimal for learning to discriminate random stimuli. However, recent observations of dense, low-dimensional activity across granule cells have called into question the role of sparse coding in these neurons. Here, we generalize theories of cerebellar learning to determine the optimal granule cell representation for tasks beyond random stimulus discrimination, including continuous input-output transformations as required for smooth motor control. We show that for such tasks, the optimal granule cell representation is substantially denser than predicted by classical theories. Our results provide a general theory of learning in cerebellum-like systems and suggest that optimal cerebellar representations are task-dependent.