Distinct roles of Hoxa2 and Krox20 in the development of rhythmic neural networks controlling inspiratory depth, respiratory frequency, and jaw opening

Abstract

Background: Little is known about the involvement of molecular determinants of segmental patterning of rhombomeres (r) in the development of rhythmic neural networks in the mouse hindbrain. Here, we compare the phenotypes of mice carrying targeted inactivations of Hoxa2, the only Hox gene expressed up to r2, and of Krox20, expressed in r3 and r5. We investigated the impact of such mutations on the neural circuits controlling jaw opening and breathing in newborn mice, compatible with Hoxa2-dependent trigeminal defects and direct regulation of Hoxa2 by Krox20 in r3.

Results: We found that Hoxa2 mutants displayed an impaired oro-buccal reflex, similarly to Krox20 mutants. In contrast, while Krox20 is required for the development of the rhythm-promoting parafacial respiratory group (pFRG) modulating respiratory frequency, Hoxa2 inactivation did not affect neonatal breathing frequency. Instead, we found that Hoxa2-/- but not Krox20-/- mutation leads to the elimination of a transient control of the inspiratory amplitude normally occurring during the first hours following birth. Tracing of r2-specific progenies of Hoxa2 expressing cells indicated that the control of inspiratory activity resides in rostral pontine areas and required an intact r2-derived territory.

Conclusion: Thus, inspiratory shaping and respiratory frequency are under the control of distinct Hox-dependent segmental cues in the mammalian brain. Moreover, these data point to the importance of rhombomere-specific genetic control in the development of modular neural networks in the mammalian hindbrain.

Figure 1

Phenotypic traits of Hoxa2^-/- mutants at birth: impaired oro-buccal behavior and increased tidal volume. (a) Plethsymographic recordings of wild-type (top), and heterozygous (middle) and homozygous (bottom) Hoxa2 mutant mice at P0. Inspiration is upward. Note that in Hoxa2^-/- mice, there is a two-fold increase in tidal volume compared with Hoxa2^+/- and wild-type littermates, whereas the frequency is the same (about 110 breaths/minute). (b, c) Individual data relating tidal volume (V_T, abscissa) and number (nb) of jaw openings (ordinates) at P0.1. Each symbol corresponds to one animal. Black triangles are for Hoxa2^-/- mutants (b, c), open circles represent Hoxa2^+/- mutants (c) and open squares correspond to wild-type animals (b). Note that Hoxa2^-/- mutants can be separated from other genotypes at P0.1, due to their two-fold increased tidal volume and their reduced number of jaw openings. Broken lines indicate the values used to calculate penetrance of the phenotype (V_T, all data inferior to mean – 1 standard deviation; jaw openings, all data superior to mean + 1 standard deviation).

Figure 2

Tidal volume of Hoxa2^-/- and Krox20^-/- mutants during the first postnatal days. Inset on the left: plethysmographic recordings of wild-type mice at different times after birth (P0) during the first postnatal day. Inspiration is upward and expiration is downward. Note the evolution of tidal volume (respiratory amplitude) and respiratory frequency during the first day. Calibration bars: abscissa, 1 s; ordinates, 10 μl. Graphs present the evolution of mean ± standard error of the mean. Tidal volume (V_T) in wild-type mice (open squares, dotted line) and (a) Hoxa2^-/- and (b) Krox20^-/- mutants (black triangles, continuous line) during the first two days after birth. All mutants were dead shortly after P0.75, therefore explaining the lack of further data. Note that in wild-type and Krox20^-/- animals, tidal volume rapidly increased during the first 12 hours of life whereas in Hoxa2^-/- mice, tidal volume was already two-fold greater at P0.1. ***P < 0.001.

Figure 3

Respiratory frequency of Hoxa2^-/- and Krox20^-/- mutants during the first postnatal days. Graphs present the evolution of mean ± standard error of the mean respiratory frequency in wild-type mice (open squares, dotted line) and (a) Hoxa2^-/- and (b) Krox20^-/- mutants (black triangles, continuous line) during the first two days after birth. All mutants were dead shortly after P0.75, therefore explaining the lack of further data. Deficiency of the respiratory frequency is lethal in Krox20^-/- [17]; frequency is normal in Hoxa2^-/- mutants. **P < 0.01, ***P < 0.001.

Figure 4

r3 and r5 are reduced in size in Krox20^Cre/flox embryos. (a, b) The size of r3 was estimated at E9.5 on flat-mounted hindbrains by labeling adjacent rhombomeres. r4 was delimited by in situ hybridization with a Hoxb1 probe and r2 by detection of the alkaline phosphatase activity from an r2-specific transgene [47, 48]. The negative territory located in between corresponds to r3 and is reduced in Krox20^Cre/flox (b) compared to control Krox20^Cre/+ (a) embryos. A few Hoxb1-positive cells are also observed within r3 in embryos (arrows). (c, d) Flat mounts of Krox20^Cre/+ (c) and Krox20^Cre/flox(d) hindbrains immunolabeled with an antibody directed against the 155 kDa component of neurofilaments (2H3). r3 and r5 can be distinguished from even-numbered rhombomeres by their less advanced differentiation of reticular neurons, revealed by lower neurofilament immunoreactivity. The r5/r6 boundary is clearly visible since it is followed by axons (arrow in (c, d)). Both r3 and r5 are reduced in Krox20^Cre/flox embryos, the effect being more dramatic in r3 (arrowheads). (e) Breathing frequency at birth in heterozygous Krox20^lacZ/+ (left) and in Krox20^Cre/flox (middle) mice is the same as in wild-type mice (WT, white columns); it is lower than normal in homozygous Krox20^lacZ/lacZ mice (right). *** : p<0.001

Figure 5

Naloxone treatment was effective on ventilation in Hoxa2^-/- mice but did not increase survival. (a) Plethysmographic recordings before (control) and after naloxone (NLX) treatment in Hoxa2^-/- animals. Note the frequency increase and the reduction of apneas. (b, c) Effects of the subcutanueous injection of NLX (1 mg/kg) at P0.5 upon mean ± standard error of the mean respiratory frequency calculated without apneic episodes in wild-type, Hoxa2^+/-, Hoxa2^-/- and Krox20^-/- animals (b) or survival in the same genotypes (c). In (b, c) white bars indicate control values and black bars labeled NLX indicate values in the same animals one hour (b) (respiratory frequency) or 1.5 days (c) (survival) after NLX injection. Although NLX eliminates apneas (a) and increases respiratory frequency (by 31 + 11% in Hoxa2^-/- and 51 ± 37 % in Krox20^-/-) (b), it does not allow survival of Hoxa2^-/- mutants (c).

Figure 6

Hyperplasia of the r1-derived dorsal pontine tegmentum in Hoxa2^-/- mutants. Para-sagittal sections of the brainstem of (a, b, d, e, g) Hoxa2^{EGFP(lox-neo-lox)/+};r2::Cre (labeled Hoxa2^EGFP/+;r2::Cre) or (c, f, h) Hoxa2^{EGFP(lox-neo-lox)/-};r2::Cre (labeled Hoxa2^EGFP/-;r2::Cre) E16 mice at different latero-medial levels, from more lateral (KF level) (a-c), to more medial (LC level) (g, h) through an intermediate level (trigemino-facial level) (d, e). (a, d) Violet cresyl stainings from which the delimitations of the different brainstem nuclei appearing as lines in other panels were drawn. (b, c, e, f) Immunodetection of EGFP, showing r2-derived cells. Note the reduction of the r2-derived domain and the ventral expansion of r1-derived nuclei such as the estimated KF (arrowheads in (c)), PB and RtTg (arrowheads in (f)) in Hoxa2^{EGFP(lox-neo-lox)/-};r2::Cre mice. (g, h) Immunodetection of EGFP at the LC level showing reduction of the r2-derived territory in Hoxa2^{EGFP(lox-neo-lox)/-};r2::Cre; insets show immunodetection of PHOX2A, a marker for the LC noradrenergic neurons – note the ventral expansion in Hoxa2^{EGFP(lox-neo-lox)/-};r2::Cre mice (arrows in (h)). 5Mo, trigeminal motor nucleus; 7Mo, facial motor nucleus; Cb, cerebellum; KF, estimated position of the Kölliker-Fuse nucleus; LC, locus coeruleus; MVe, medial vestibular nucleus; PB, estimated position of the parabrachial nucleus; Pn, pontine nuclei; Pr5, principal sensory trigeminal nucleus; RtTg, reticulo-tegmental nucleus of the pons; Sp5, spinal trigeminal tract; Su5, supratrigeminal nucleus; vsc, ventral spinocerebellar tract.