5247
J Physiol 589.21 (2011) pp 5247–5256
Postnatal loss of brainstem serotonin neurones compromises the ability of neonatal rats to survive episodic severe hypoxia Kevin J. Cummings1 , Julie C. Hewitt2 , Aihua Li2 , John A. Daubenspeck2 and Eugene E. Nattie2 1
The Journal of Physiology
2
Department of Biomedical Sciences, University of Missouri-Columbia, Columbia, MO 65211, USA Department of Physiology and Neurobiology, Dartmouth Medical School, Lebanon, NH 03756-0001, USA
Non-technical summary Failure to withstand severe hypoxia (failed autoresuscitation) has been documented in sudden infant death syndrome (SIDS) cases, and SIDS is associated with a constellation of defects within the brainstem serotonin system, including reduced serotonin content. Neonatal mice with reduced numbers of serotonergic neurones beginning in utero also have major defects in autoresuscitation at the beginning of the second postnatal week. Here we injected a neurotoxin into the brainstems of 2- to 3-day-old rat pups, reducing serotonin content by ∼80%, and at P7–10 tested their ability to autoresuscitate when challenged with repeated episodes of environmental anoxia. Pups with reduced serotonin content have delayed gasping in response to each challenge, as well as a significantly higher mortality by the last episode. These new data imply that a postnatal loss of brainstem serotonin in infants could increase the risk of SIDS during severely hypoxic conditions. Abstract Pet-1−/− mice with a prenatal, genetically induced loss of 5-hydroxytryptamine (5-HT, serotonin) neurones are compromised in their ability to withstand episodic environmental anoxia via autoresuscitation. Given the prenatal role of 5-HT neurones in the development of neural networks, here we ask if a postnatal loss of 5-HT neurones also compromises autoresuscitation. We treated neonatal rat pups at postnatal day (P)2–3 with an intra-cisternal injection of 5,7-dihydroxytryptamine (5,7-DHT; ∼40 μg; n = 8) to pharmacologically lesion the 5-HT system, or vehicle (control; n = 14). At P7–10 we exposed unanaesthetized treated and control pups to 15 episodes of environmental anoxia (97% N2 , 3% CO2 ). Medullary 5-HT content was reduced 80% by 5,7-DHT treatment (P < 0.001). Baseline ventilation (V˙ E ), metabolic rate (V˙ O2 ), ventilatory equivalent (V˙ E /V˙ O2 ), heart rate (HR), heart rate variability (HRV) and arterial haemoglobin saturation (S aO2 ) were no different in 5-HT-deficient pups compared to controls. However, only 25% of 5-HT-deficient pups survived all 15 episodes of environmental anoxia, compared to 79% of control littermates (P = 0.007). High mortality of 5,7-DHT-treated pups was associated with delayed onset of gasping (P < 0.001), delayed recovery of HR from hypoxic-induced bradycardia (P < 0.001), and delayed recovery of eupnoea from hypoxic-induced apnoea (P < 0.001). Treatment with 5,7-DHT affected neither the gasping pattern once initiated, nor HR, V˙ E /V˙ O2 or S aO2 during the intervening episodes of room air. A significant increase in HRV occurred in all animals with repeated exposure, and in 5-HT-deficient pups this increase occurred immediately prior to death. We conclude that a postnatal loss of brainstem 5-HT content compromises autoresuscitation in response to environmental anoxia. This report provides new evidence in rat pups that 5-HT neurones serve a physiological role in autoresuscitation. Our data may be relevant to understanding the aetiology of the sudden infant death
C 2011 The Authors. Journal compilation C 2011 The Physiological Society
DOI: 10.1113/jphysiol.2011.214445
5248
K. J. Cummings and others
J Physiol 589.21
syndrome (SIDS), in which there is medullary 5-HT deficiency and in some cases evidence of severe hypoxia and failed autoresuscitation. (Received 20 June 2011; accepted after revision 11 September 2011; first published online 12 September 2011) Corresponding author K. J. Cummings: Department of Biomedical Sciences, University of Missouri-Columbia, E109 Veterinary Medicine Building, 1600 E. Rollins St, Columbia, MO 65211, USA. Email:
[email protected] Abbreviations 5,7-DHT, 5,7-dihydroxytryptamine; HF, high frequency; HR, heart rate; HRV, heart rate variability; 5-HT, 5-hydroxytryptamine; LF, low frequency; SIDS, sudden infant death syndrome.
Introduction In adult animals, 5-HT neurones within the medullary raphe nuclei contribute to respiratory, thermoregulatory and autonomic control (Hodges et al. 2008; Ramage & Villalon, 2008; Corcoran et al. 2009). A growing body of evidence supports the idea that medullary 5-HT neurones also participate in cardio-respiratory homeostasis during the early postnatal period, including the stabilization of both breathing and heart rate. Neonatal mice and rats having lost most of their medullary 5-HT neurones experience augmented spontaneous bradycardias and apnoeas (Cummings et al. 2009, 2010; Hodges et al. 2009), probably exposing them to episodic hypoxia with potentially deleterious consequences for growth and survival. In addition to respiratory and cardiac dysfunction in normoxic conditions, neonatal mice lacking 60–70% of brainstem 5-HT neurones from early embryogenesis onwards (Pet-1−/− ; Hendricks et al. 2003; Kiyasova et al. 2011) have major defects in autoresuscitation, a life-preserving process utilized by neonatal mammals in severely hypoxic conditions (Fewell, 2005; Erickson et al. 2007; Cummings et al. 2011a). Not only do Pet-1−/− neonates have an unusually prolonged apnoea in response to a single episode of environmental anoxia (Erickson & Sposato, 2009), but these animals are also more likely to succumb to repeated episodes of environmental anoxia (Cummings et al. 2011a). Notwithstanding a normal gasping pattern, Pet-1−/− neonates show a marked delay in restoring HR from the hypoxic-induced bradycardia, and an unusually large increase in HRV after a single episode of environmental anoxia (Cummings et al. 2011a). After two or three episodes, gasping is completely ineffectual at restoring HR. An especially interesting aspect of this phenotype is the critical period in development in which it presents (at ∼P8; Cummings et al. 2011a), and until ∼P10 (KJC, unpublished observations). In addition to acute effects on physiological control systems, brainstem 5-HT neurones have potent developmental effects during early embryogenesis. Inputs from 5-HT neurones influence the proliferation, differentiation and migration of neurones throughout the CNS, including those residing within brainstem cardio-respiratory control networks (Lauder &
Krebs, 1978; Bou-Flores et al. 2000). Thus, the compromised ability of Pet-1−/− animals to survive episodic environmental anoxia may result from either a developmental defect or the loss of acute, 5-HT-mediated physiological responses to hypoxic stress. Understanding how a postnatal loss of 5-HT neurones affects cardio-respiratory recovery from conditions of severe hypoxia (i.e. autoresuscitation) could help in our understanding of the aetiology of SIDS, the leading cause of death in the post-neonatal period. First, SIDS is associated with multiple 5-HT system defects including 5-HT deficiency (Kinney et al. 2003; Paterson et al. 2006; Duncan et al. 2010). We do not know whether this deficiency and associated defects occur prenatally or postnatally. Second, some eventual SIDS cases experience episodes of apnoea (Kahn et al. 1992) and bradycardia (Poets et al. 1999; Sridhar et al. 2003) prior to death, which suggests exposure to episodic hypoxia/ischaemia. And some SIDS cases exhibit multiple histopathological markers of episodic hypoxia/ischaemia (Kinney, 2009). Third, and perhaps most important, some SIDS cases ultimately fail to autoresuscitate (Poets et al. 1999; Sridhar et al. 2003). Here we hypothesize that a postnatal loss of 5-HT neurones is sufficient to compromise autoresuscitation during repeated episodes of environmental anoxia. To test this hypothesis we induced in rat pups a pharmacological loss of 5-HT neurones in the first few days of life. We then tested the cardio-respiratory responses and survival of these unanaesthetized animals within the critical period when Pet-1−/− neonates are susceptible to failed autoresuscitation. We find that rat pups with a pharmacologically induced loss of 5-HT neurones also have marked defects in autoresuscitation, supporting a physiological and not a developmental role for these neurones in the recovery of cardio-respiratory function after severely hypoxic conditions.
Methods Animals
Pregnant dams were provided food and water ad libitum and were housed with a 12 h light–dark cycle at an ambient C 2011 The Authors. Journal compilation C 2011 The Physiological Society
J Physiol 589.21
5-HT and autoresuscitation in neonatal rats
temperature of 21–23◦ C. Pups were obtained from three litters from three breeding pairs. From each litter, pups were assigned to either vehicle (saline; n = 14; 8 males, 6 females) or drug (5,7-dihydroxytryptamine; 5,7-DHT; n = 8; 5 males, 3 females) groups. Drug or vehicle injection was performed via the cisterna magna on P2–3, and autoresuscitation was tested at P7–10. In a previous publication (Cummings et al. 2009) we demonstrated that an identical, intra-cisternal injection of 5,7-DHT significantly reduced (by ∼80%) the number of 5-HT-positive neurones in the raphe magnus and raphe obscurus, but not in the raphe pallidus. In addition, 5-HT-positive axonal projections were dysmorphic and sparse throughout the medulla. We chose to study the animals at P7–10 because defective autoresuscitation has been documented in Pet-1−/− animals within this age range (Cummings et al. 2011a and KJC, unpublished observations). All pups assigned to drug treatment were injected I.P. with 20 mg kg−1 desipramine prior to 5,7-DHT injection to protect catecholaminergic neurones. Five pups injected with vehicle were also pre-treated with desipramine as a control. All experimental protocols were approved by the Institutional Animal Care and Use Committee of Dartmouth College, and conform to principles of UK regulation. Surgery
At ∼30 min to 1 h prior to surgery, pups assigned to either vehicle or drug groups were injected with desipramine. Pups were cold-anaesthetized for ∼10–15 min within a glass cylinder immersed in ice water until they were motionless. Pups were then removed from the bath and placed on a bed of crushed ice covered in aluminum foil to maintain deep hypothermia. While in the prone position, the head of the animal was placed over a small bar within a stereotactic setup. A ∼1 cm incision was made to expose the section of thin membrane covering the entrance to the cisterna magna. The tip of a 30-gauge needle was inserted into the cisterna magna, and 1 μl of either 100 mM 5,7-DHT (∼40 μg total) or vehicle (0.1% ascorbic acid in saline) was injected. The animal was maintained in the prone position with the needle held in place for 10 min to allow for drug diffusion. The needle was then removed, and the muscle and epidermal layers sutured. Animals were then allowed to warm gradually at room temperature.
5249
temperature, held at 36 ± 0.5◦ C in all animals throughout the experiment, was controlled by changing the temperature of the water perfusing the glass chamber, thereby changing the ambient temperature within the chamber. Ambient temperature at the beginning of the experiment was ∼33◦ C, but owing to the suppressive effects of hypoxia on metabolic rate (V˙ O2 ) and body temperature we increased the temperature of the water bath ∼1–2◦ C over the duration of the experiment to keep body temperature constant. Ventilation (V˙ E ) was measured with a head-out system. The head chamber was made by fitting a section of vinyl over the end of the syringe tube (volume ∼3 ml), held in place with another rubber gasket (Terumo Medical Corp., Japan) that fitted into the anterior end of the chamber. The snout of the animal (devoid of fur) was placed into a small hole in the vinyl and sealed with polyether material (Impregum F Polyether Impression material, 3M, St Paul, MN, USA). A downstream pump (AEI Technologies, Naperville, IL, USA) connected to the outlet port of the mask pulled air through the pneumotach and mask at a flow of 150 ml min−1 . The expired gas was drawn through a small column of Drierite (W. A. Hammond Drierite Co. Ltd, Xenia, OH, USA), and then an O2 analyser (AEI Technologies, Pittsburgh, PA, USA), allowing the determination of V˙ O2 . Arterial haemoglobin saturation (S aO2 ) data were obtained using a pulse oximeter fixed to the base of the tail (Starr Life Sciences, Pittsburgh, PA, USA). Anoxia (97% N2 /3% CO2 ) was delivered directly from a tank to the surrounds of the pneumotach through the open end of a 50 ml syringe placed over the end of the pneumotach. In this way, the downstream pump pulled the hypoxic gas through the mask with a fast wash-in time (∼2–3 s) but with no change in mask pressure. Body temperature was continually monitored with fine thermocouples (Omega Engineering, Stamford, CT, USA). Surface ECG was used to measure heart rate (HR) using electrodes embedded within a vest made from tensor bandage. Thermocouples, ECG leads and pulse oximeter leads were exteriorized through a hole in a rubber gasket at the posterior end of the chamber. Inspiratory and expiratory airflows were detected by connecting both side-arms of the pneumotach to a differential pressure transducer (Validyne Engineering, Northridge, CA, USA). Integration of the flow trace provided respiratory volume, calibrated by injecting and withdrawing known volumes of air (0.05, 0.1, 0.2 ml) at the end of each experiment. The pneumotach responded in a linear fashion to these volumes.
Experimental setup
The setup was based on one previously described (Cummings et al. 2009; Cummings & Frappell, 2009). Briefly, the animal chamber (volume ∼40 ml) was constructed from a water-jacketed glass cylinder. Body C 2011 The Authors. Journal compilation C 2011 The Physiological Society
Experimental protocol
Animals were removed from the litter and fur on the snout was removed with commercially available product (Nair;
5250
K. J. Cummings and others
Church and Dwight, Princeton, NJ, USA). The animal was then instrumented for ECG, body temperature and S aO2 , and placed within the pre-heated glass chamber, at an ambient temperature of ∼33◦ C, within the thermoneutral zone of young rat pups (Mortola & Dotta, 1992). The snout of the animal was sealed within the head chamber, and the head chamber was fixed by way of the rubber gasket into the chamber’s anterior end. Body temperature was allowed to warm to ∼36◦ C (∼15 min). We then recorded 10 min of baseline parameters, followed by 15 episodes of environmental anoxia, each separated by 5 min of normoxia. It was previously shown that normal rat pups of comparable age fail to autoresuscitate at or around 15 anoxic exposures (Serdarevich & Fewell, 1999). Environmental anoxia was only given until pups were apnoeic and motionless (∼40–50 s for each group). Our setup is designed such that when the anoxic gas is removed, the head chamber is immediately flushed with room air. In this way, the inspirate in the head chamber was normoxic when the first gasp occurred. At the end of the hypoxic trials, 5,7-DHT-treated (n = 8) and control littermates (n = 6) were killed with an overdose of ketamine–xylazine. For determination of 5-HT and noradrenaline content by HPLC, whole brains were extracted and medullas separated from cerebellar and pontine tissue. Tissue was then frozen and stored at –80◦ C. Medullas were homogenized in a solution of 0.1 M trichloroacetic acid, 10 mM sodium acetate, 0.1 mM EDTA, 1 mM isoproterenol, followed by centrifugation. Supernatants were used for HPLC analysis.
*
[Monoamine] (ng/mg protein)
16 14 12 10 8 6 4 2 0 5-HT
NA
Figure 1. Medullary serotonin (5-HT) content is significantly reduced by 5,7-DHT treatment HPLC determination of 5-HT and noradrenaline (NA) content in the medullas of vehicle-treated rat pups (filled bars, n = 6), and those given an intracerebroventricular injection of 5,7-DHT after treatment with desipramine (n = 8, open bars) to protect NA neurones. ∗ Significant difference from vehicle (P < 0.001). All data are means ± SEM.
J Physiol 589.21
Data analysis
In addition to assessing survival across the 15 hypoxic episodes, we measured respiratory frequency (f B , breaths min−1 ), tidal volume (V T , ml kg−1 ), V˙ E (f × V T ; ml min−1 kg−1 ), V˙ O2 (ml min−1 kg−1 ), V˙ E /V˙ O2 , HR (beats min−1 ), and S aO2 prior to episode 1, 5, 10 and/or last episode survived. Mass-specific V˙ O2 was determined using the formula: V˙ O2 = (0.21 – fractional O2 exhausted from mask) × flow (ml min−1 )/mass (kg). Autoresuscitation variables were measured including gasp latency (duration of hypoxic apnoea), the time constant for the restoration of HR (63% recovery), time required for eupnoea recovery, and the frequency and amplitude of gasps from the first to the last survived episode. To assess whether a postnatal loss of 5-HT neurones altered the autonomic response to anoxia, we measured low- (LF) and high- (HF) frequency components of HRV in treated and control pups during the 5 min baseline in normoxia, after the first episode of anoxia and prior to the 5th and last survived episode of anoxia (∼2000 beats for each period). HR was sampled at 1 kHz and the inter-beat intervals (RRI) were computed using the peak detection function on LabChart 6.0. Text files were exported for analysis using an instrument scripted in MATLAB (Mathworks, Natick, MA, USA). The RRI spectral estimation used the Lomb analysis technique to avoid the spectral distortion caused by constant-interval re-sampling (Moody, 1993). Spectral power was estimated by summation within LF (0.15–1.5 Hz) and HF (1.5–6 Hz) bands as suggested by Just and colleagues (Just et al. 2000). The LF/HF ratio was also obtained for each segment. Differences between treated and control groups at baseline were assessed with two-tailed t tests. Some animals did not survive all 15 bouts, so we assessed differences in pre-episode or autoresuscitation variables between groups over the 1st, 5th and last survived episodes using a 2-factor, repeated-measures ANOVA (factor 1: vehicle or drug treatment; factor 2: hypoxic episode). Tukey’s post hoc tests were performed when overall significant differences were found. Given that the HRV data were not normally distributed, differences between control and drug-treated groups across hypoxic episodes were assessed with a 2-factor, repeated-measures ANOVA on ranked data, and confirmed with a Kruskal–Wallis non-parametric test. Survival curves for vehicle- and drug-treated animals were compared across the 15 episodes using Kaplan–Meier log rank analysis. Significant differences between vehicleand drug-treated groups were evaluated at an α level of P < 0.05. Results Medullary 5-HT content was reduced by ∼80% in 5,7-DHT-treated pups (P < 0.001; Fig. 1). Mass and C 2011 The Authors. Journal compilation C 2011 The Physiological Society
5-HT and autoresuscitation in neonatal rats
J Physiol 589.21
5251
Table 1. Baseline parameters for vehicle- (n = 14) and 5,7-DHT-treated rat pups (n = 8). All data are means ± SEM
Vehicle 5,7-DHT
Mass (g)
HR (beats min−1 )
V˙ E (ml min−1 kg−1 )
VT (ml kg−1 )
fB (breaths min−1 )
V˙ E /V˙ O2
V˙ O2 (ml min−1 kg−1 )
S aO2 (%)
19.9 ± 0.7 15.0 ± 1.0∗
404 ± 6 406 ± 12
955 ± 42 1031 ± 126
6.3 ± 0.2 7.2 ± 0.8
152 ± 5 146 ± 10
22.0 ± 1.0 24.8 ± 2.5
44.5 ± 2.2 43.4 ± 6.0
93.1 ± 1.1 93.8 ± 2.1
∗ Significantly
different from vehicle (P < 0.05).
baseline cardio-respiratory variables prior to challenge are shown in Table 1. Drug-treated animals weighed ∼25% less than vehicle-treated animals (P = 0.02), but no other significant differences were observed between the two groups during baseline, room-air conditions. A loss of brainstem 5-HT significantly affects the ability of pups to withstand episodic, severely hypoxic conditions. 79% of control pups survive all 15 episodes, compared to just 25% of 5-HT-deficient pups (Fig. 2A, P = 0.007). The median number of survived episodes is 15 and 12 for controls and treated animals, respectively (P = 0.003). The death of 5-HT-deficient animals was associated with several key defects in autoresuscitation. Prolonged gasp latency (i.e. prolonged hypoxic apnoea) is evident from the first episode of severe hypoxia, and a delay in the recovery of HR and breathing emerges after subsequent episodes (Fig. 2B). Across all episodes, gasp latencies are on average twice as long in 5-HT-deficient animals compared to controls (P = 0.001; Fig. 3A), and are not further prolonged with subsequent episodes. Beyond the delay in its initiation, gasping in 5-HT-deficient animals is normal with respect to frequency (Fig. 3B) and volume (Fig. 3C). In both groups, the recovery of HR and breathing is prolonged from the first to the last episode (episode effect: P < 0.001 for both), but 5,7-DHT-treated pups require more time to recover HR across all episodes (treatment effect: P = 0.002; Fig. 4A). The delay in 5-HT-deficient animals becomes more striking with subsequent episodes of severe hypoxia, increasing 5-fold from the first to the last survived episode (Fig. 4A). In contrast, HR recovery in control littermates increases only ∼25% from the first to the last episode (treatment × episode interaction: P < 0.001; Fig. 4A). Along with delayed HR recovery, pups with reduced brainstem 5-HT are also delayed in their recovery of eupnoea; this becomes further prolonged over the course of subsequent episodes of severe hypoxia. From the first to the last episode, 5-HT-deficient animals require 5-fold more time to recover eupnoea, while the recovery time for control littermates is only increased 2-fold (treatment effect: P = 0.002; treatment × episode effect: P < 0.001; Fig. 4B). Note the strong effect of 5-HT deficiency on autoresuscitation parameters from the first to last episode despite the fact that, on average, treated animals receive significantly fewer challenges of environmental anoxia compared to controls. C 2011 The Authors. Journal compilation C 2011 The Physiological Society
We examined whether the increased mortality of 5-HT-deficient pups was reflected in cardio-respiratory parameters during the intervening period of room air prior to the 1st, 5th and last challenge. HR falls over the course of the challenges, but the magnitude of the effect is no different between treated animals (–16 ± 3%) and littermate controls (–14 ± 2%) (episode effect: P < 0.001). Despite the delay in the recovery of eupnoea, the progressive hyperventilation that we observed during the intervening periods of room air is not significantly different in 5-HT-deficient rat pups (+79 ± 15%) compared to control littermates (+96 ± 13%) (episode effect: P < 0.001). The hyperventilation results from a progressive increase in V T over the course of the episodes (not shown). While S aO2 dropped precipitously in response to each hypoxic challenge, there was no effect of 5,7-DHT treatment on S aO2 across normoxic recovery periods (not shown). To examine whether a postnatal loss of 5-HT neurones alters the autonomic response to severe hypoxia, we examined HRV in treated and control pups at baseline, after the first hypoxic episode, and prior to the fifth and last survived episode (Fig. 5). For both groups, there is a significant increase in HRV from the first to the last episode. This effect is of the same magnitude in control and treated pups (episode effect: P < 0.001 for Total, LF and HF power, Fig. 5A–C, respectively), but occurs sooner, just prior to death, in treated animals. Because both the LF and HF components of HRV increased by the same magnitude, in both groups there was no significant change in LF/HF from the first to the last survived episode (Fig. 5D).
Discussion Here we demonstrate that neonatal rats having sustained a loss of 5-HT neurones induced postnatally are severely compromised in their ability to withstand episodic environmental anoxia. The high mortality of 5-HTdeficient rat pups is associated with key defects in autoresuscitation, including a significant delay in the initiation of gasping that is apparent upon the first hypoxic episode. And once gasping is initiated, 5-HT-deficient pups take progressively longer to restore HR and breathing over subsequent episodes. These data extend our previous work in animals with a prenatal loss of 5-HT neurones (Pet-1−/− )
5252
K. J. Cummings and others
These data may be relevant to the pathogenesis of SIDS, a syndrome associated with multiple defects within regions of the medullary 5-HT system, including a 26% reduction in 5-HT content and a 22% reduction in the expression of tryptophan hydroxylase-2, the enzyme catalysing the rate-limiting step in 5-HT biosynthesis (Duncan et al.
and suggest that 5-HT neurones are necessary in postnatal life for an appropriate physiological response to severe hypoxia. Insults to the medullary 5-HT system occurring in early postnatal life could compromise the physiological response necessary for human infants to survive severely hypoxic conditions.
A
J Physiol 589.21
100
% Survival
80
60
40
*
20
0 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Episode no.
B
Figure 2. Postnatal loss of medullary 5-HT neurones decreases survival over the course of episodic, environmental anoxia A, probability of survival for vehicle-treated pups (filled circles, n = 12) and 5,7-DHT-treated pups (open circles, n = 8) across the 15 episodes of anoxia. ∗ Significantly different from vehicle-treated pups (P < 0.001). B, raw traces of heart rate (HR) and respiration (Resp Volume) over the 1st, 10th and last survived hypoxic episode for a control and 5,7-DHT-treated littermate. Period of anoxia is denoted by an open bar. Hypoxic apnoea (i.e. gasp latency) is shown with an arrow. The first gasp is indicated with an asterisk (∗ ). Note the prolonged gasp latency across all episodes and delay in HR recovery (time to 63% recovery; shaded region) that emerges in the 5,7-DHT-treated animal by episode 12 (last survived episode). C 2011 The Authors. Journal compilation C 2011 The Physiological Society
5-HT and autoresuscitation in neonatal rats
J Physiol 589.21
2010). That episodic hypoxia precedes death in SIDS cases is suggested by subtle gliosis, apoptosis, reactive microglia, an up-regulation of antioxidant enzymes and a down-regulation of mitochondrial-associated protein-2 (Kinney, 2009). Most striking is the failure of autoresuscitation observed in some SIDS cases during what appears to be severely hypoxic conditions (Poets et al.
A
50
1
Gasp Latency (s)
40
1999; Sridhar et al. 2003). Taken together with previous findings (Hodges et al. 2008; Cummings et al. 2009, 2010), our data suggest that reduced 5-HT content may not only increase the number and intensity of hypoxic episodes during infancy through increased apnoea and bradycardia, but could also increase the risk of death during more severely hypoxic conditions that might occur during airway obstruction or re-breathing while in prone or face-covered conditions (Poets et al. 1999; Pasquale-Styles et al. 2007; Kinney, 2009). 5-HT-deficient rat pups require twice as long to initiate gasping in response to severe hypoxia compared to control littermates. Unlike the delay in the recovery
30
A
20
160
1
5
10
LAST
30
Gasp fB
25
HR recovery (s)
0
100 80 60 40 20
15
0
Veh: 14.4 5,7-DHT: 11.8
1
5
10
LAST
10
1,2,3
B 160
0 5
10
LAST
30 25 20
Eupnoea recovery (s)
140 1
Gasp VT
120
20
5
C
1,2,3
140
10
B
5253
120 100 80 60 40
15
20
10
0
Veh: 14.4 5,7-DHT: 11.8
1 5
5
10
LA S T
Episode no.
0 1
5
10
LAST
Episode no.
Figure 3. Delay in the initiation of gasping in 5-HT-deficient rat pups A, across all episodes of severe hypoxia, rat pups with 5-HT deficiency (open circles, n = 8) take twice as long as control littermates (filled circles, n = 12) to initiate gasping. Data for the last survived bout (LAST) is shown for each group. 1 indicates treatment effect: P < 0.001. Neither the frequency (f B ) (B) nor amplitude (C) of gasps is affected by a loss of 5-HT neurones. All data are means ± SEM. C 2011 The Authors. Journal compilation C 2011 The Physiological Society
Figure 4. Delayed recovery of heart rate and eupnoea in rat pups with a loss of 5-HT with subsequent episodes of severe hypoxia A, time required for the recovery of heart rate (HR) across episodes 1, 5, 10 and the last survived episode (LAST) in vehicle-treated (filled circles, n = 14) and 5,7-DHT-treated (open circles, n = 8) rat pups. B, time required for recovery of eupnoea in each group across hypoxic episodes. Note on x-axis that the last episode occurs sooner in treated animals (5,7-DHT; 11.8 episodes) compared to vehicle-treated controls (Veh; 14.4 episodes).1 : significant effect of 5,7-DHT treatment relative to control littermates.2 : significant effect of hypoxic episodes. 3 : significant interaction between treatment and hypoxic episode. All data are means ± SEM.
5254
K. J. Cummings and others
of HR and eupnoea which only emerge in treated pups after several hypoxic episodes, prolonged hypoxic apnoea is apparent during the first episode and may therefore be a key factor determining survival. These data are supported by data obtained from neonatal Pet-1−/− mice (Erickson & Sposato, 2009; Cummings et al. 2011a), and from a reduced preparation demonstrating an important contribution from medullary 5-HT2A -receptor activation for the initiation of gasping (Tryba et al. 2006). It may be that the initiation of gasping by medullary respiratory neurones (e.g. neurones within the pre-B¨otzinger complex) relies in part on excitatory inputs derived from medullary 5-HT neurones. Alternatively, a loss of 5-HT neurones may lead to enhanced tonic inhibition (Solomon, 2000) or to a lowering of the P O2 threshold at which gasping emerges (Fewell, 2005). Once initiated, gasping appears to be normal in Pet-1−/− mice in vivo (Cummings et al. 2011a) and in situ (St-John et al. 2009), and with pharmacological blockade of 5-HT receptors in situ (St-John & Leiter, 2008) suggesting that
A
J Physiol 589.21
medullary 5-HT neurones have little role in shaping the gasping pattern once it is generated. The mechanism(s) responsible for the delay in HR and respiratory recovery observed in 5-HT-deficient rat pups is unknown. The delay in recovery may be secondary to the prolonged apnoea and associated factors produced or released during severe hypoxia. One possibility is an increase in adenosinergic activity occurring in a background of 5-HT deficiency. Adenosine, a ubiquitous molecule released by hypoxia, inhibits respiratory neurones directly through its binding to A1 receptors or indirectly through binding to stimulatory A2A receptors on GABAergic interneurones (Zaidi et al. 2006; Abu-Shaweesh & Martin, 2008), and modulates the respiratory and HR responses to anoxia (Fewell et al. 2007). Despite prolonging respiratory recovery, a loss of 5-HT does not influence the progressive increase in V˙ E /V˙ O2 occurring during the intervening normoxic periods. Although it is not influenced by a loss of 5-HT, this phenomenon may be a form of respiratory plasticity akin
B 1×10–1
1×10–1
*
1×10–2 HF (a.u.)
LF (a.u.)
1×10–2
1×10–3
1×10–4
1×10–3
1×10–4 Veh:14.4 5,7-DTH: 11.8
1×10–5 B
C
1
5
1×10–5
L
D
1×10–1
Veh:14.4 5,7-DTH: 11.8 B
1
5
L
3
*
LF/HF
1×10–2 Total (a.u.)
*
1×10–3
2
1 1×10–4 Veh:14.4 5,7-DTH: 11.8
1×10–5 B
1
5
L
Veh:14.4 5,7-DTH: 11.8
0 B
1
5
L
Figure 5. Effect of episodic severe hypoxia and 5-HT deficiency on heart rate variability in rat pups Low-frequency (LF) (A), high-frequency (HF) (B), Total (C) power and LF/HF (D) at baseline (B), after the first episode (1) and prior to the fifth (5) and last survived episode (L) of severe hypoxia in control (filled circles, n = 14) and 5,7-DHT-treated (open circles, n = 8) littermates. Note on x-axis that the last episode occurs sooner in treated animals (11.8 episodes) compared to controls (14.4 episodes). ∗ : significant effect of episode on LF, HF and Total power. C 2011 The Authors. Journal compilation C 2011 The Physiological Society
J Physiol 589.21
5-HT and autoresuscitation in neonatal rats
to the long-term facilitation of ventilation that has been described in other preparations (Mahamed & Mitchell, 2007; Julien et al. 2008). In both treated and control pups there is an eventual increase in autonomic drive that occurs with exposure to severe hypoxia that manifests in both the HF and LF domains of HRV (Fig. 5). This increase occurs immediately prior to autoresuscitation failure in 5-HT-deficient pups. An increase in HRV after a single episode of also occurs in Pet-1−/− mice that eventually fail to autoresuscitate (Cummings et al. 2011a), and while its significance is unknown, increased HRV may be a marker indicating susceptibility to severe hypoxia. The lesion imparted to the medullary 5-HT system in our rat pups is similar to Pet-1−/− neonates with respect to 5-HT content (Cummings et al. 2011b), and the overall extent of 5-HT neurone depletion (Cummings et al. 2009). However, the phenotype in the rat pups is less severe. Rat pups with reduced brainstem 5-HT can withstand at least nine anoxic episodes, while Pet-1−/− neonates succumb after only two or three episodes (Cummings et al. 2011a). Likewise, the initiation of gasping and recovery of HR and eupnoea is more profoundly delayed in Pet-1−/− mice. This may be related to the more prolonged hypoxic apnoea in Pet-1−/− mice compared to 5,7-DHT-treated rat pups. One might expect the phenotype of 5,7-DHT-treated animals to be more, not less, dramatic than the Pet-1−/− phenotype; Pet-1−/− animals retain a population of 5-HT neurones projecting to specific brainstem regions, including the nucleus of the solitary tract, nucleus ambiguus and ventrolateral medulla (Kiyasova et al. 2011). In contrast, 5,7-DHT is applied to the entire brainstem and its cytotoxicity is indiscriminate with respect to these 5-HT neurone populations. It may be that the critical period during which 5-HT exerts its influence on autoresuscitation is shifted in rats relative to mice. We focused solely on the critical period (∼P7–10) in mice when a loss of brainstem 5-HT neurones compromises autoresuscitation (Cummings et al. 2011a). As development is slightly (∼1–1.5 days) delayed in rats relative to mice (Schneider & Norton, 1979), it may be that at slightly older ages the sensitivity of 5-HT-deficient rat pups to episodic, severe hypoxia would more closely resemble that of Pet-1−/− mice. Regardless, future investigations into the specific mode of action of medullary 5-HT neurones within cardio-respiratory or autonomic nuclei participating in autoresuscitation should be aided by the development of a rat model in which 5-HT can be pharmacologically reduced in postnatal life. While our data may have relevance to SIDS, we note that the lesion imparted to the 5-HT system of our rat pups (∼80% reduction in 5-HT) is more severe than that observed in SIDS (∼25%) (Duncan et al. 2010). Beyond the medullary 5-HT system, defects in other brainstem neurotransmitter systems have been observed in C 2011 The Authors. Journal compilation C 2011 The Physiological Society
5255
SIDS cases, and SIDS may involve environmental insults other than hypoxia (Kinney et al. 2009). Nevertheless, our data support the concept that a loss of medullary 5-HT occurring postnatally can compromise autoresuscitation, a life-preserving process initiated by severely hypoxic conditions and which has been shown to be defective in SIDS (Poets et al. 1999; Sridhar et al. 2003). References Abu-Shaweesh JM & Martin RJ (2008). Neonatal apnea: what’s new? Pediatr Pulmonol 43, 937–944. Bou-Flores C, Lajard AM, Monteau R, De Maeyer E, Seif I, Lanoir J & Hilaire G (2000). Abnormal phrenic motoneuron activity and morphology in neonatal monoamine oxidase A-deficient transgenic mice: possible role of a serotonin excess. J Neurosci 20, 4646–4656. Corcoran AE, Hodges MR, Wu Y, Wang W, Wylie CJ, Deneris ES & Richerson GB (2009). Medullary serotonin neurons and central CO2 chemoreception. Respir Physiol Neurobiol 168, 49–58. Cummings KJ, Commons KG, Fan KC, Li A & Nattie EE (2009). Severe spontaneous bradycardia associated with respiratory disruptions in rat pups with fewer brain stem 5-HT neurons. Am J Physiol Regul Integr Comp Physiol 296, R1783–R1796. Cummings KJ, Commons KG, Hewitt JC, Daubenspeck JA, Kinney HC & Nattie EE (2011a). Failed heart rate recovery at a critical age in 5-HT-deficient mice exposed to episodic anoxia: Implications for SIDS. J Appl Physiol 111, 825–833. Cummings KJ & Frappell PB (2009). Breath-to-breath hypercapnic response in neonatal rats: temperature dependency of the chemoreflexes and potential implications for breathing stability. Am J Physiol Regul Integr Comp Physiol 297, R124–R134. Cummings KJ, Li A, Deneris ES & Nattie EE (2010). Bradycardia in serotonin-deficient Pet-1−/− mice: influence of respiratory dysfunction and hyperthermia over the first 2 postnatal weeks. Am J Physiol Regul Integr Comp Physiol 298, R1333–R1342. Cummings KJ, Li A & Nattie EE (2011b). Brainstem serotonin-deficiency in the neonatal period: autonomic dysregulation during mild cold stress. J Physiol 589, 2055–2064. Duncan JR, Paterson DS, Hoffman JM, Mokler DJ, Borenstein NS, Belliveau RA, Krous HF, Haas EA, Stanley C, Nattie EE, Trachtenberg FL & Kinney HC (2010). Brainstem serotonergic deficiency in sudden infant death syndrome. JAMA 303, 430–437. Erickson JT, Shafer G, Rossetti MD, Wilson CG & Deneris ES (2007). Arrest of 5HT neuron differentiation delays respiratory maturation and impairs neonatal homeostatic responses to environmental challenges. Respir Physiol Neurobiol 159, 85–101. Erickson JT & Sposato BC (2009). Autoresuscitation responses to hypoxia-induced apnea are delayed in newborn 5-HT-deficient Pet-1 homozygous mice. J Appl Physiol 106, 1785–1792. Fewell JE (2005). Protective responses of the newborn to hypoxia. Respir Physiol Neurobiol 149, 243–255.
5256
K. J. Cummings and others
Fewell JE, Zhang C & Gillis AM (2007). Influence of adenosine A1 -receptor blockade and vagotomy on the gasping and heart rate response to hypoxia in rats during early postnatal maturation. J Appl Physiol 103, 1234–1241. Hendricks TJ, Fyodorov DV, Wegman LJ, Lelutiu NB, Pehek EA, Yamamoto B, Silver J, Weeber EJ, Sweatt JD & Deneris ES (2003). Pet-1 ETS gene plays a critical role in 5-HT neuron development and is required for normal anxiety-like and aggressive behavior. Neuron 37, 233–247. Hodges MR, Tattersall GJ, Harris MB, McEvoy SD, Richerson DN, Deneris ES, Johnson RL, Chen ZF & Richerson GB (2008). Defects in breathing and thermoregulation in mice with near-complete absence of central serotonin neurons. J Neurosci 28, 2495–2505. Hodges MR, Wehner M, Aungst J, Smith JC & Richerson GB (2009). Transgenic mice lacking serotonin neurons have severe apnea and high mortality during development. J Neurosci 29, 10341–10349. Julien C, Bairam A & Joseph V (2008). Chronic intermittent hypoxia reduces ventilatory long-term facilitation and enhances apnea frequency in newborn rats. Am J Physiol Regul Integr Comp Physiol 294, R1356–R1366. Just A, Faulhaber J & Ehmke H (2000). Autonomic cardiovascular control in conscious mice. Am J Physiol Regul Integr Comp Physiol 279, R2214–R2221. Kahn A, Groswasser J, Rebuffat E, Sottiaux M, Blum D, Foerster M, Franco P, Bochner A, Alexander M, Bachy A et al. (1992). Sleep and cardiorespiratory characteristics of infant victims of sudden death: a prospective case-control study. Sleep 15, 287–292. Kinney HC (2009). Neuropathology provides new insight in the pathogenesis of the sudden infant death syndrome. Acta Neuropathol 117, 247–255. Kinney HC, Randall LL, Sleeper LA, Willinger M, Belliveau RA, Zec N, Rava LA, Dominici L, Iyasu S, Randall B, Habbe D, Wilson H, Mandell F, McClain M & Welty TK (2003). Serotonergic brainstem abnormalities in Northern Plains Indians with the sudden infant death syndrome. J Neuropathol Exp Neurol 62, 1178–1191. Kinney HC, Richerson GB, Dymecki SM, Darnall RA & Nattie EE (2009). The brainstem and serotonin in the sudden infant death syndrome. Annu Rev Pathol 4, 517–550. Kiyasova V, Fernandez SP, Laine J, Stankovski L, Muzerelle A, Doly S & Gaspar P (2011). A genetically defined morphologically and functionally unique subset of 5-HT neurons in the mouse raphe nuclei. J Neurosci 31, 2756–2768. Lauder JM & Krebs H (1978). Serotonin as a differentiation signal in early neurogenesis. Dev Neurosci 1, 15–30. Mahamed S & Mitchell GS (2007). Is there a link between intermittent hypoxia-induced respiratory plasticity and obstructive sleep apnoea? Exp Physiol 92, 27–37. Moody GB (1993). Spectral analysis of heart rate without resampling. Comput Cardiol 20, 715–718. Mortola JP & Dotta A (1992). Effects of hypoxia and ambient temperature on gaseous metabolism of newborn rats. Am J Physiol Regul Integr Comp Physiol 263, R267–R272. Pasquale-Styles MA, Tackitt PL & Schmidt CJ (2007). Infant death scene investigation and the assessment of potential risk factors for asphyxia: a review of 209 sudden unexpected infant deaths. J Forensic Sci 52, 924–929.
J Physiol 589.21
Paterson DS, Trachtenberg FL, Thompson EG, Belliveau RA, Beggs AH, Darnall R, Chadwick AE, Krous HF & Kinney HC (2006). Multiple serotonergic brainstem abnormalities in sudden infant death syndrome. JAMA 296, 2124–2132. Poets CF, Meny RG, Chobanian MR & Bonofiglo RE (1999). Gasping and other cardiorespiratory patterns during sudden infant deaths. Pediatr Res 45, 350–354. Ramage AG & Villalon CM (2008). 5-Hydroxytryptamine and cardiovascular regulation. Trends Pharmacol Sci 29, 472–481. Schneider BF & Norton S (1979). Equivalent ages in rat, mouse and chick embryos. Teratology 19, 273–278. Serdarevich C & Fewell JE (1999). Influence of core temperature on autoresuscitation during repeated exposure to hypoxia in normal rat pups. J Appl Physiol 87, 1346–1353. Solomon IC (2000). Excitation of phrenic and sympathetic output during acute hypoxia: contribution of medullary oxygen detectors. Respir Physiol 121, 101–117. Sridhar R, Thach BT, Kelly DH & Henslee JA (2003). Characterization of successful and failed autoresuscitation in human infants, including those dying of SIDS. Pediatr Pulmonol 36, 113–122. St-John WM & Leiter JC (2008). Maintenance of gasping and restoration of eupnea after hypoxia is impaired following blockers of α1 -adrenergic receptors and serotonin 5-HT2 receptors. J Appl Physiol 104, 665–673. St-John WM, Li A & Leiter JC (2009). Genesis of gasping is independent of levels of serotonin in the Pet-1 knockout mouse. J Appl Physiol 107, 679–685. Tryba AK, Pena F & Ramirez JM (2006). Gasping activity in vitro: a rhythm dependent on 5-HT2A receptors. J Neurosci 26, 2623–2634. Zaidi SI, Jafri A, Martin RJ & Haxhiu MA (2006). Adenosine A2A receptors are expressed by GABAergic neurons of medulla oblongata in developing rat. Brain Res 1071, 42–53. Author contributions All authors contributed to either the concepts or experimental design, the analysis and presentation of data or the drafting and revision of the manuscript. All authors have also approved the final version of this manuscript for publication. All experiments were performed at Dartmouth Medical School. Acknowledgements Funding for this study was provided by an NIH Program Project grant HD36379 (NICHD, PI, H. C. Kinney, and PROJECT 2 PI, E. E. Nattie) and NIH grant HL28066, (PI, E. E. Nattie). HPLC determinations were performed by the CMN/KC Neurochemistry Core Lab at Vanderbilt University. The CMN/KC Neurochemistry Core Lab is supported by Vanderbilt Kennedy Centre for Research on Human Development, Vanderbilt Conte Centre for Neuroscience Research and The Vanderbilt Centre for Molecular Neuroscience. The authors also acknowledge the generous support of the Parker B. Francis Family and its Foundation (Fellowship to K. J. Cummings). C 2011 The Authors. Journal compilation C 2011 The Physiological Society
