Dba degree program

In obstructive sleep apnea, hypoxic ventilatory sensitivity may affect the degree of hypoxic stress and sleep disruption that occurs in response to upper airway obstruction. We induced (1) sleep-induced hypoxia (SIH) or (2) sleep fragmentation (SF) without hypoxia for 5 days (12-hour light/dark cycle) in two inbred mouse strains with low (A/J) and high (DBA/2J) hypoxic ventilatory sensitivities. During SIH, the time to arousal (26.4 ?¡À 1.1 vs. 21.3 ?¡À 1.5 seconds, p

Keywords: carotid body; genetics; hypoxic ventilatory response; obstructive sleep apnea; sleep fragmentation

Sleep-disordered breathing is a prevalent condition affecting 2-4% of the adult population in the United States (1). Obstructive sleep apnea (OSA) is the most common form of sleep-disordered breathing and is characterized by repetitive periods of intermittent hypoxia and arousal from sleep (2). The intermittent hypoxia and sleep fragmentation (SF) of OSA have been implicated in numerous pathophysiologic sequelae, including neurologic, cardiovascular, and metabolic disturbances, resulting in significant morbidity and mortality (3-7). Although the magnitude of the hypoxic stimulus has been linked to the degree of pathophysiologic dysfunction in OSA (7), the factors that determine the rate and severity of intermittent hypoxia and SF are unclear.

The severity of the intermittent hypoxia and the degree of SF can vary considerably between OSA patients. Even in normal individuals, in whom airway obstruction is induced experimentally during sleep, the duration of hypoxic periods and the frequency of events exhibit marked variability (8). One potential source of the variability in apnea characteristics is due to differences in the hypoxic ventilatory response during sleep (9, 10). More specifically, the body's primary hypoxic sensing organ, the carotid body, can play a crucial role in eliciting an arousal response and terminating an apnea (11-13). Indeed, Bowes and colleagues (12) demonstrated in a dog model of OSA that carotid body denervation resulted in an escalating hypoxic stimulus that was unable to elicit an arousal response. Thus, the carotid body and hypoxic ventilatory responsiveness can play a key role in respiratory responses to airway obstruction during sleep.

We have previously demonstrated that different inbred mouse strains exhibit a wide variation in hypoxic ventilatory responsiveness (14). Within the strains studied, the DBA/2J strain exhibited the largest ventilatory response to hypoxia, and the A/J strain exhibited the smallest ventilatory response to hypoxia. Furthermore, we have recently shown that the carotid body of the DBA/2J strain is four times larger, contains many more typical glomus cells, and is markedly more responsive to the neurotransmitter acetylcholine than the A/J strain (15). The purpose of this study was to determine whether genetic differences in hypoxic sensitivity and carotid body structure and function between the DBA/2J and A/J strains affect the phenotypic expression of sleep-disordered breathing. To examine genetic influences on sleep-disordered breathing severity, we have developed a murine model of sleep-induced hypoxia (SIH), in which the rate and severity of hypoxic events occur independent of upper airway properties (16). In the model, polysomnography is measured continuously, and the sleep/wake state is scored in real time, enabling the delivery of an increasing hypoxic stimulus at sleep onset, followed by reoxygenation at arousal. Responses were compared between the two inbred mouse strains, and we hypothesized that the severity of SIH would be greater in the low hypoxic-responding A/J strain than the high hypoxic-responding DBA/2J strain. In contrast, we further hypothesized that that the severity of nonhypoxic SF would be comparable between the two strains. Some of the results of these studies have been previously reported in the form of an abstract (17).

METHODS

Adult male A/J (24.3 ?¡À 0.5 g, 109 ?¡À 7 days of age) and DBA/2J (25.6 ?¡À 0.9 g, 96 ?¡À 6 days of age) mice from Jackson Laboratory (Bar Harbor, ME) were used in the study, which was approved by the Johns Hopkins University Animal Use and Care Committee and complied with the American Physiological Society Guidelines. Animals were euthanized with penlobarbital (60 mg intraperitoneally).

Procedures

In isoflurane (1-2%)-anesthetized animals, three electroencephalographic electrodes and two nuchal clectromyographic electrodes were implanted as previously described (16). At 5-7 days after surgery, a connector cable from the animal was fixed above to a low friction mercury swivel, allowing 360-degree unrestricted movement of the tethered mouse. The mouse was housed in a customized chamber with a 12-hour light/dark cycle in which gas entered through inlet ports in the base and exhausted through an open hole at the apex (see online supplement for further details).

The techniques and procedures related to our online sleep/wake detection system have been previously described (16). In brief, the program used a series of investigator-determined thresholds for frequency distribution of the electroencephalogram and amplitude of the nuchal electromyogram to determine sleep/wake state over 5-second epochs. In this study using DBA/2J and A/J strains, the computer algorithm detected wakefulness and NREM sleep with an accuracy consistent with our previously published values in C57BE/6J mice (16), but during REM sleep fell below our previously published level of 85.6 ?¡À 5.0% accuracy. However, misclassified REM periods were always scored as NREM sleep and did not affect the triggering of SIH or SF (discussed later here). Overall sleep architecture data are reported as wakefulness and sleep (NREM and REM sleep combined).

SIH and SF

SIH. During continuous periods of wakefulness, room air was delivered through the cage at a rate of 4 L/minute. When sleep was detected for a continuous 15-second period, the output signal from the computer caused the airflow to cease and 100% N^sub 2^ to be delivered at 4 L/minute until arousal occurred, at which time the delivery of room air was restored to produce reoxygenation.

SF without hypoxia. A system was developed to cause arousal using a tactile and auditory stimulus after a comparable period of time from sleep onset as that produced with hypoxia. The sampling ports for measurement of FI^sub O^sub 2^^, during SIH were adapted and used to deliver a sequentially increasing airflow stimulus until arousal occurred (10 L/ minute from 0-5 seconds, 20 L/minute from 5-10 seconds, and 50 L/ minute from 10-15 seconds).

Protocol

At the completion of a 3- to 5-day acclimation period in the housing chamber, 5 consecutive days of either SIH or SF were bracketed by 24-hour baseline and recovery periods during which sleep was recorded.

Analyses

We determined the time and duration of each event and periods of wakefulness and sleep (we used previously described combining rules to produce contiguous epochs of sleep/wake) (16). The relative propensity of strains (DBA/2J versus A/J) to fall asleep was evaluated using the Andersen-Gill model of multiple failure-time survival analysis (18). Statistical significance between factors of SIH or SF, light cycle (light, dark), or changes over time (Days 1-5) was determined by analysis of variance. If the analysis of variance was significant for a factor, a Dunnett test was used to determine which means were significantly different. All the results are presented as mean ?¡À SEM.

RESULTS

SIH

Event characteristics. A recording example of consecutive hypoxic events during NREM sleep is shown in Figure 1 for an A/J (upper tracing) mouse and a DBA/2J (lower tracing) mouse. The figure demonstrates that the A/J mouse has longer hypoxic events when compared with the DBA/2J mouse. Group data show that over the 5 days of SIH, A/J mice had hypoxic events that lasted on average 26.4 ?¡À 1.1 seconds and were significantly longer (p

In addition to having longer more severe hypoxic events, the A/J mice also had significantly more events (p

Sleep architecture. The total sleep time during the 24-hour baseline period, the 5 days of SIH, and the 24-hour recovery period is shown in Figure 3. The DBA/2J strain slept on average for 39.3 ?¡À 3.8% of the 24-hour control period (Figure 3B, lower panel). There was a significant decrease (p