INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

MANY PHYSIOLOGICAL PARAMETERS vary regularly over time. Among the most studied are the daily oscillations in body temperature (T_b), metabolic rate [oxygen consumption (O₂)], and activity. These rhythms have an endogenous origin controlled by a biological clock located at the level of the suprachiasmatic nucleus (16). Their natural period in free-running conditions, i.e., in the absence of external synchronizers, usually differs from 24 h, and the daily light-dark cycle is believed to be the most powerful synchronizer (9).

The circadian oscillations in O₂ and T_b, which occur even in the absence of daily fluctuations in activity (3, 13), are thought to be contributed to by changes in thermoregulatory mechanisms (13). Many mammals respond to hypoxia with a drop in O₂ and T_b (10), largely because of an inhibition of heat production in all of its forms, shivering, nonshivering, and behavioral thermogenesis, coupled to a lowering in the set point of thermoregulation (11). We tested the hypothesis that hypoxia blunts the circadian oscillations of T_b. Experiments were performed on rats during synchronized conditions (12:12-h light-dark cycle) and then in constant light (free running) to further examine whether the effects of hypoxia on the circadian pattern may involve the biological clock.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All experiments were performed on Sprague-Dawley male rats of 200-250 g, after approval from the Animal Ethics Committee of this institution. Rats were kept either in 12:12-h dark-light conditions (light off at 7:00 PM) or in constant-light conditions. In either case, light intensity was 300-600 lx. T_b and activity were measured by standard telemetric techniques. A few days before the first measurements, animals were instrumented with an intra-abdominal transmitter, powered by an energizer receiver unit (4000E, Mini Mitter, Sunriver, OR). T_b was obtained from the frequency of the transmitter, and activity was the total score of counts registered by the radiating coils of the energizer receiver platform. Data were sampled at 1 Hz, averaged in 2-min bins, and stored on computer. O₂ was measured by an open-flow system, with continuous recording of the inflowing and outflowing O₂ and CO₂ concentrations of the gases delivered through the metabolic chamber at the constant flow rate of 1,800 ml/min (5). Hypoxia was obtained in normobaric or hypobaric conditions and was maintained for 1 wk. Normobaric hypoxia was used for the 12:12-h dark-light experiments (T_b and activity, n = 8; O₂, n = 4), by delivering an hypoxic gas mixture from pressurized tanks under continuous monitoring of the gas concentration in the chamber. Hypobaric hypoxia was adopted for the experiments in free-running conditions (n = 4) by use of a hypobaric chamber (7). In all cases, the inspired O₂ partial pressure was ~75-80 Torr, which corresponds to an inspired O₂ fractional concentration of ~10.5% in normobaria. At this inspired O₂ concentration, the arterial partial pressure of O₂ is 35-40 Torr, which approximately doubles pulmonary ventilation (12, 14). Exposures to hypercapnia (n = 4; 3% CO₂ in normoxia, arterial partial pressure of CO₂ ~45 Torr; Ref. 15) and hyperoxia (n = 4; 40% O₂ in normocapnia) were obtained by delivering the gases from pressurized tanks under continuous monitoring by appropriate gas analyzers for 1 wk, each under 12:12-h dark-light conditions. Switches between conditions occurred at ~10:00 AM.

Data were statistically analyzed by paired t-test or repeated-measurements ANOVA, with three post hoc Bonferroni limitations for the three comparisons of interest (air, intervention, recovery in air). In all cases, statistical significance was defined at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Synchronized, 12:12-h dark-light cycle. T_b averaged ~37.5°C and oscillated during the day with an amplitude of ~1°C, with the peak occurring during the dark phase. Similarly, O₂ and activity peaked during the dark phase. In fact, rats are nocturnal animals, and their T_b, activity, and metabolism cycles peak during the dark phase. On exposure to hypoxia, T_b decreased almost immediately, and in a few hours it averaged ~35°C. Subsequently, T_b gradually rose, reaching the lowest normoxic value of ~37°C in 24-36 h (Fig. 1). During the following days, the circadian oscillations were greatly depressed (Fig. 2A), with their amplitude averaging only ~0.2°C. This was due to the decrease in the dark-phase T_b, which averaged 37.3°C instead of the normoxic 37.9°C (P < 0.001), whereas the light-phase T_b in hypoxia (37.1°C) was as in normoxia. On return to normoxia, T_b increased toward the highest normoxic values even though the switch was performed during the light phase, and the normal circadian cycle was reestablished (Fig. 1, top).

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Top: body temperature (Tb) in air and hypoxia during synchronized conditions (12:12-h dark-light cycle). Curve represents average of 8 animals. Solid bars indicate duration of daily period of darkness, from 7:00 PM to 7:00 AM. Bottom: Tb of 1 rat in free-running conditions (constant light). Ultradian oscillations of Tb (13) are superimposed on circadian oscillations, which in this rat had a period of 25.3 h. Ultradian oscillations are not apparent in the record at top because of the averaging of 8 animals.

View larger version (37K): [in this window] [in a new window]   Fig. 2.   Tb (A, B) and activity [C, D; in arbitrary (arb) units] of rats during 12:12-h dark-light cycles. Horizontal bars indicate dark phase of cycle (7:00 PM to 7:00 AM). Traces represent average values of 8 rats studied for 7 consecutive days in air and hypoxia (A, C) and of 4 rats studied in air and hypercapnia (B, D). In all cases, the day of the change in inspired gases was excluded from the analysis. Error bars indicate SE. Hypoxia, but not hypercapnia, greatly reduced the amplitudes of the circadian oscillations of Tb and, to a lesser extent, of activity.

1

1

1

1

1

1

1

1

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Oxygen consumption of rats during 12:12-h dark-light cycles. Horizontal bars indicate dark phase of cycle (7:00 PM to 7:00 AM). Traces represent average values of 4 rats studied for 7 consecutive days in air and hypoxia. Error bars indicate SE. The day of the change from air to hypoxia was excluded from the analysis.

Free running, constant light. In normoxia the free-running period averaged 25-25.5 h, such that the T_b peak was progressively delayed by ~1-1.5 h every day. Because the period was not identical for all rats, each rat was analyzed individually. A representative recording is presented in Fig. 1, bottom, in which the ultradian oscillations in T_b, with periods of ~2 h (13), are superimposed on the circadian oscillation. On exposure to hypoxia, T_b dropped and then gradually rose with a pattern very similar to that observed in the synchronized condition. Because of the severely reduced amplitude, the period was not easily recognizable. On return to normoxia, the circadian pattern was readily restored, with approximately the same period of the prehypoxic condition and at the expected phase (Fig. 4).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Actogram-type double plot of Tb (hours of the day vs. days of study) in a rat during free-running conditions in air before (days 1-6) and after (days 14-20) 1 wk of hypobaric hypoxia. Horizontal bars join the hours of minimum and maximum Tb. During hypoxia, circadian oscillations were of small amplitude or absent (see Fig. 1, bottom). On rat's return to air breathing, the oscillations immediately resumed, at approximately the expected phase of the cycle.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have observed that hypoxia profoundly influenced the circadian oscillations of all variables and to different degrees. In fact, during hypoxia, the oscillation of T_b was almost entirely abolished, and those of activity and O₂ averaged, respectively, ~37 and 56% of the normoxic amplitudes. Different effects on the amplitude of these parameters is not surprising because, although activity influences the metabolic level and the latter can affect T_b, these cycles are not necessarily linked. Absence of activity, as in men during continuous bed rest, does not eliminate the circadian oscillations of T_b and O₂, and T_b changes during the day are not temporally linked to those of activity or O₂ (3, 13). Because thermoregulatory mechanisms are likely to be at the basis of the daily oscillations in T_b and O₂ (13), the fact that during hypoxia T_b and O₂ were closer to the minimal, rather than the maximal, values of the normoxic cycle is strongly suggestive that hypoxia depressed their circadian oscillations by inhibiting thermogenic mechanisms. In acute conditions, hypoxia is known to depress shivering, nonshivering, and behavioral thermogenesis, with a decrease in the set point of thermoregulation (11). In addition, an hypoxic depression of activity (Fig. 2C) would also contribute to the drop in O₂ and T_b; however, the depression of the oscillations of T_b and O₂ far exceeded those in activity.

To test whether the depression of the circadian oscillations of T_b and O₂ were specific to hypoxia or common to other conditions disturbing acid-base control and pulmonary ventilation, we repeated the measurements in hyperoxia or hypercapnia. Both conditions result in hyperventilation: hyperoxia with some increase in O₂ (8, 11) and hypercapnia with no changes in O₂ (15). Neither hyperoxia nor hypercapnia had any significant effect on the circadian patterns, although hypercapnia slightly reduced the dark-time activity to ~70% of the air value. The absence of effects during hypercapnia excluded an involvement of the peripheral chemoreceptors. This is in agreement with the notion that the effects of hypoxia on thermogenesis and metabolic rate do not depend on the function of the peripheral chemoreceptors (6).

On return to normoxia after 1 wk in hypoxia, T_b rapidly increased to the highest values, even if the return to normoxia occurred during the early part of the light phase when T_b was normally low. We did not measure O₂ during this posthypoxic recovery phase. It is known, however, that acute hyperoxia can increase O₂ (8, 11); hence, the T_b overshoot from the sudden release of the hypoxic inhibition could have represented a condition of relative hyperoxia.

To assess whether hypoxia also affected, in addition to the amplitudes, the period of the circadian oscillations, additional experiments were performed in free-running conditions, i.e., without external synchronizers. Under these conditions, the period of the cycle was close to 25 h; hence, the phase of the cycle was progressively delayed by ~1 h/day. We found that, on return to normoxia, the circadian pattern was clearly restored, not only with the same period of the prehypoxic condition but also at the expected phase (Fig. 4). Hence, during the hypoxic depression of the oscillations, the biological clock continued to click at the normal pace.

In conclusion, these experiments in adult rats indicated that hypoxia depresses the circadian amplitude of metabolism and T_b without affecting the period of the clock. The most likely action of hypoxia is, therefore, not at the level of the suprachiasmatic nucleus but at the hypothalamic centers of thermoregulation. Circadian oscillations in metabolism and T_b are likely determinants of many daily functions, including the duration and organization of sleep (2, 17). Sleep alterations are known to occur in situations of sustained hypoxia, as at high altitude (1), and in patients with cardiorespiratory diseases (4). Hence, the present results raise the possibility that some disturbances during chronic hypoxia could find their basis in the alterations of the most common circadian rhythms.