Metabolic and blood gas dependence on digestive state in the Savannah monitor lizard Varanus exanthematicus: an assessment of the alkaline tide

Large carnivorous reptiles may experience an increase in oxygen consumption(V̇_O2) during digestion that is as large as that achieved during activity(Andrade et al., 1997; Hicks et al., 2000; Secor and Diamond, 1995; Secor et al., 2000). While the metabolic level reached during these two physiological challenges is similar,there are considerable temporal differences in their development. During exercise, reptiles immediately respond with an increase in metabolic rate,which then returns to resting levels shortly after cessation of exercise(Garland et al., 1987; Wagner and Gleeson, 1997; Wang et al., 1997). This pattern differs from digestive metabolic responses which take hours to develop and have an extended duration lasting days(Coulson et al., 1950a; Coulson et al., 1950b; Hicks et al., 2000; Jorgensen, 1992; Secor et al., 2000). The increased V̇_O2 associated with digestion has been attributed primarily to the preparation of the gut for digestion via protein synthesis and secretion of digestive compounds(Andrade et al., 1997; Secor and Diamond, 1995),including acid secretion by the stomach(Secor, 2003).

Cardiopulmonary studies during digestion in two species of reptiles(Glass et al., 1979; Hicks et al., 2000; Secor et al., 2000) found a relative hypoventilation during digestion in comparison to either rest or activity.

The postprandial response is also associated with alkalization of the blood, known as the alkaline tide. Recent studies in alligators, pythons and boas reexamined the postprandial changes in pH(Andrade et al., 2004; Busk et al., 2000b; Overgaard et al., 1999) and found a significant increase in plasma [HCO₃^–], an increase in V̇_O2 but no significant change in arterial pH. However, these reptiles rely on anaerobic metabolism to a large extent, so to test the generality of the postprandial metabolic and acid–base response in reptiles it is necessary to examine a reptile species that has a relatively high aerobic capacity. The present study investigated the acid–base response during specific dynamic action(SDA) in another large carnivorous reptile, the Savannah monitor lizard, Varanus exanthematicus. Savannah monitor lizards belong to a group of reptiles with activity levels higher than most other reptilian groups, and they have the capacity to maintain relatively high rates of aerobic metabolism(Wood et al., 1978). We measured the magnitude and time course of the metabolic response, blood gases,strong ions and acid–base balance associated with the postprandial metabolic response in Savannah monitor lizards.

Savannah monitor lizards (Varanus exanthematicus Bosc) with body mass ranging from 350 to 1200 g were obtained from a commercial animal supplier (California Zoological Supply, Santa Ana, CA, USA) and housed in large plastic containers equipped with ceramic heat lamps at one end, creating a thermal gradient between 28° and 36°C. UVB (5%)-emitting fluorescent bulbs supplied light within the containers. Lights and heat lamps were maintained on a 12 h:12 h light:dark cycle. Animals were offered unlimited adult mice or rat pups and crickets once or twice weekly prior to the experimental period. Water was available at all times. This study was approved by University of California, Irvine Institutional Animal Care and Use Committee animal protocol number 2123.

Animals were fasted for at least 3 weeks prior to surgery and experimentation. Animals were anesthetized by exposure to isoflurane (Isoflo:Abbott Laboratories, North Chicago, IL, USA). Lizards were then artificially ventilated with 2% isoflurane via a vaporizer (Dräger, Lubeck, Germany). A femoral artery was occlusively cannulated with heparin-soaked, saline filled PE-50 or PE-60 tubing (Harvard Apparatus, Inc., Holliston, MA, USA). To reduce the risk of infection and post-surgical pain, antibiotics and analgesics were administered immediately after surgery and every second day post-surgery(Enrofloxacin; Baytril, Bayer Corporation, Shawnee Mission, KS, USA and Flunixin meglumine; Flunixamine, Fort Dodge, Madison, NJ, USA, respectively). Following surgery all animals resumed voluntary breathing and were allowed to recover for at least 18 h in an 8-l plastic box in a walk-in environmental chamber set at 35°C.

Fasting animals were placed in an 8-l metabolic chamber fashioned from a Rubbermaid box, which was placed in a temperature-controlled room at 35°C(preferred body temperature of this species)(Hicks and Wood, 1985). Room air was pulled serially through the metabolic chamber, a Drierite (anhydrous calcium sulfate; Xenia, OH, USA) column to remove water vapor, a flow meter,and oxygen and carbon dioxide gas analyzers (model S3A, Applied Electrochemistry, Inc., Sunnyvale, CA, USA and model LB2, Beckman, Schiller Park, IL, USA) at an average flow rate of 240±2 ml min^–1 with a vacuum pump. Oxygen consumption(V̇_O2) and carbon dioxide excretion (V̇_CO2) were monitored using AcqKnowledge data acquisition software (Biopac Systems, Inc. Goleta, CA, USA) and converted to stpd. A second group of lizards was used to measure arterial pH (pH_a), blood gases and strong ions. Fasting values were measured before each meal. Blood samples were taken every 8 h post-feeding, and measurements were made using a Nova blood gas analysis system (Waltham, MA, USA). Blood gases and pH values were corrected for temperature using correction curves that were generated for V. exanthematicus blood as follows. Blood from V. exanthematicuswas equilibrated over a range of oxygen and carbon dioxide tensions. Samples of this blood were then measured both by the Nova blood gas analysis system and by a Radiometer blood gas analysis system (Copenhagen, Denmark) that was thermostatted to the animal's body temperature (35°C). The differences in values between these two measurement systems were used to correct for the temperature difference between the animal's body temperature and the temperature at which the blood sample was measured in the Nova blood gas analysis system. Plasma bicarbonate concentration([HCO₃^–]) was calculated from the Henderson–Hasselbalch equation([HCO₃^–]=αP_CO2×10^(pH–pK′)) using simultaneously determined pH and P_CO2; α and pK′ were corrected for temperature and pH (Boutilier et al.,1984). Following the fasting period, the animals were removed from the chamber and offered 10% of their body weight in rat pups; each animal ate readily within 10 min. After feeding, the animal was returned to its chamber,and metabolic rate was measured continuously and 0.2 ml blood samples were taken every 8 h post-feeding for 5 days. After each measurement cannulae were refilled with a 100 IU ml^–1 heparin saline solution(Elkins-Sinn, Inc., Cherry Hill, NJ, USA).

Repeated measures analysis of variance (ANOVA; Dunnett Test for Multiple Comparisons vs Control Group) was used to analyze time course data(Graphpad Software, San Diego, CA, USA). Levels of significance were assumed with P values less than 0.05. Values are reported as means ±standard error of the mean (s.e.m.); N=6 unless otherwise noted.

V̇_O2 of V. exanthematicus increased from a fasting level of 0.95±0.19 ml min^–1 kg^–1 to 2.57±0.85 ml min^–1 kg^–1 24 h post-feeding(Fig. 1A). This increase in V̇_O2 began within 4 h post feeding, and reached a maximum at 24 h postprandial, remaining elevated above fasting values until 114 h postprandial. Carbon dioxide excretion exhibited a similar pattern to that of oxygen consumption with V̇_CO2 increasing from a fasting level of 0.53±0.31 to 1.63±0.58 ml min^–1 kg^–1 24 h postprandial, and remaining elevated until 114 h postprandial (Fig. 1B). Consequently, respiratory exchange ratio (R) did not change during the course of digestion (Fig. 1C), with a mean value of 0.62±0.21 prior to feeding and 0.73±0.18 24 h postprandial.

A significant increase in pH_a was measured at 8 h after feeding(7.43±0.03 to 7.51±0.03) with an apparent peak in the alkaline tide at 40 h postprandial (7.56±0.04). pH_a was no longer significantly elevated by 56 h postprandial(Fig. 2A). Bicarbonate levels in the arterial blood increased from 24.1±2.6 mmol l^–1fasted to 34.6±2.4 mmol l^–1 by 8 h postprandial with an apparent peak at 40 h postprandial (40.34±5.3 mmol l^–1). Bicarbonate was no longer significantly elevated by 64 h postprandial (Fig. 2B). The increase in pH_a resulting from the increase in bicarbonate was reduced by a significantly increased Pa_CO2(5.7±0.2, 5.7±0.2, 5.8±0.4 kPa at 24, 32 and 48 h,respectively) compared to the fasted state (4.6±0.2 kPa; Fig. 2C). Throughout the postprandial period, Pa_O2 did not significantly change(Fig. 2D). Hematocrit was significantly elevated at 24 and 32 h postprandial (0.23±0.01 and 0.23±0.01, respectively) compared to the fasted state(0.2±0.01), data not shown. A significant increase in plasma lactate concentration was measured at 24 and 32 h (1.2±0.2 and 1.3±0.3 mmol l^–1, respectively) compared to fast levels(0.3±0.1 mmol l^–1; Fig. 3A). Chloride significantly decreased between 24 and 48 h postprandial (115.7±1.9 mmol l^–1 at 24 h, 113.8±2 mmol l^–1 at 32 h, 115.6±2.9 mmol l^–1 at 40 h, 115.7±3.6 mmol l^–1 at 48 h) compared to fast (122±1.9 mmol l^–1; Fig. 3B). An increase in sodium was measured at 32, 40 and 48 h postprandial(152.5±2, 150.6±1.5 and 150.7±2.4 mmol l^–1, respectively) compared to fast (144.7±1.7 mmol l^–1; Fig. 3C). Potassium levels did not change during the postprandial period(Fig. 3D).

The postprandial metabolic response of large carnivorous reptiles typically has been reported to involve a 2- to 17-fold increase in resting V̇_O2(Busk et al., 2000b; Hicks et al., 2000; Overgaard et al., 1999; Secor and Diamond, 1995), and the postprandial rise in V̇_O2reported here is consistent with previous studies on V. exanthematicus (Fig. 1)(Hicks et al., 2000). While there was a clear metabolic elevation and an alkaline tide following feeding during this study, the alkalemia did not approach the magnitude previously reported in Alligator mississippiensis(Coulson et al., 1950b). The smaller increase in pH is similar to those reported in recent studies of Python molurus (Secor and Diamond, 1995) and Boa constrictor(Andrade et al., 2004), but higher than that reported for A. mississippiensis or Python molurus (Busk et al.,2000b; Overgaard et al.,1999), suggesting that the magnitude of the alkaline tide varies amongst carnivorous reptiles.

Elevation in metabolic state typically accompanies digestion in vertebrates with both the increment and duration of the increase varying among species,type of forager, and relative meal size(Andrade et al., 1997; Coulson et al., 1950b; Hicks et al., 2000; Preest, 1991; Secor and Diamond, 1995; Secor and Diamond, 1997; Secor and Phillips, 1997; Secor et al., 1994; Wang et al., 1995). Reptiles are extreme with regard to metabolic changes during digestion ranging from as high as 44 times standard metabolic rate(Secor and Diamond, 1997) to this study where the increase in V̇_O2 of V. exanthematicus was much less (threefold increase) than that seen in P. molurus. As indicated by the consistency of the respiratory exchange ratio, increases in V̇_CO2 paralleled those of V̇_O2(Fig. 1C). The time course for the postprandial metabolic response is consistent among reptiles with a peak in SDA at 24 h after feeding, lasting about 5 days(Andersen and Wang, 2003; Busk et al., 2000a; Busk et al., 2000b).

In addition to the metabolic response, pH_a increased during digestion in V. exanthematicus; the apparent highest mean value occurred 40 h post-feeding (Fig. 2). Arterial pH peaks before metabolic rate (40 h vs 24 h). In A. mississippiensis the highest postprandial pH_awas measured at 9 h (Coulson and Hernandez,1983). Overgaard et al.(Overgaard et al., 1999) also reported a peak in pH_a before 12 h while maximal V̇_O2 of P. molurusdid not occur until 48 h postprandial. In this study, pH_a was significantly increased from fasting values by 8 h postprandial and remained elevated until 56 h postprandial.

As illustrated in Figs 2 and 5, feeding elicits an increase in Pa_CO2 and[HCO₃^–], and no change in Pa_O2. An increase in Pa_CO2 has also been reported for A. mississippiensis(Busk et al., 2000b), P. molorus (Overgaard et al.,1999), B. constrictor(Andrade et al., 2004), and Bufo marinus (Andersen and Wang,2003). In mammals, an increase in Pa_CO2 is associated with a reduction in arterial Pa_O2. This inverse relationship is due to the tight coupling of ventilation and metabolic rate as expressed in the alveolar ventilation equation. However, reptiles do not necessarily follow this pattern (Wang et al., 1998). Pa_O2 could be increased or unchanged by improving ventilation perfusion(V/Q̇) inhomogeneity or by decreasing the intrapulmonary or intracardiac shunt. Increase in Pa_CO2 could be achieved by a relative hypoventilation as has been shown in a previous study of the cardiovascular and ventilatory response to digestion in V. exanthematicus(Hicks et al., 2000). This hypoventilation is seen as a decrease in air convection requirement for CO₂ and probably represents changes in postprandial ventilatory control.

Chloride and bicarbonate exhibited the most pronounced changes after feeding, although sodium and lactate also both increased. A significant increase in lactate concentration (Fig. 3A) during the postprandial period was also seen Scincella lateralis (Preest, 1991),and A. mississippiensis (Busk et al., 2000b), but not in Rana catesbeiana(Busk et al., 2000a), B. marinus (Andersen and Wang,2003) or P. molorus(Overgaard et al., 1999). Although this study measured a fourfold increase in lactate postprandially,the actual contribution of lactate to the total anion concentration is very small. The greatest changes in plasma electrolyte concentration were the decrease in chloride (secreted into the lumen of the stomach to form HCl) and the accompanying increase in bicarbonate, and, hence, the alkaline tide (Figs 2A,B, 3B). A similar decrease in chloride during the postprandial period was reported for B. marinus(Andersen and Wang, 2003), R. catesbeiana (Busk et al.,2000a) and A. mississippiensis(Coulson et al., 1950b). However, no change was reported in chloride concentration for A. mississippiensis in another study(Busk et al., 2000b), and an increase in chloride was reported for P. molorus(Overgaard et al., 1999).

This study has delineated the changes in metabolic function as well as blood gases and acid-base status that accompany feeding in V. exanthematicus. There is some variation in the acid-base response to feeding in reptiles and amphibians. Some studies have reported a large increase in pH (Coulson et al.,1950b; Secor and Diamond,1995), some a moderate increase in pH(Andrade et al., 2004; Busk et al., 2000a), and some no change in pH (Andersen et al.,2003; Andersen and Wang,2003; Busk et al.,2000b; Overgaard et al.,1999) during the postprandial period. While it appears that there is a different response to feeding, it is rather the fasted pH values that differ most strikingly between studies. For example, fasted pH was found to be 7.36 in an early study on alligators(Coulson et al., 1950b), and 7.51 in a more recent study (Busk et al.,2000b) whereas postprandial pH in both studies was approximately 7.57; similarly, fasted pH was 7.39 in an early study on the python(Secor and Diamond, 1995), and 7.52 in a more recent study (Overgaard et al., 2002) whereas postprandial pH was 7.49 and 7.53,respectively. Each of the studies discussed above showed a significant increase in bicarbonate during the postprandial period, but the impact of the alkaline tide on the acid–base status of the animal is varied by that animal's ventilatory response (Wang et al., 2001). The response to an increase in bicarbonate post-feeding in V. exanthematicus is an initial metabolic alkalosis followed by a relative hypoventilation, a return to fasting pH levels, and a return to fasting Pa_CO2(Fig. 4). The same response has been reported for B. constrictor(Andrade et al., 2004), and B. marinus (Andersen and Wang,2003); however, A. mississippiensis showed a relative hypoventilation prior to a small metabolic alkalosis(Busk et al., 2000b).

Many of the processes involved in feeding and absorption of food contribute to the cost of digestion, and it is of interest to consider the division of cost. Secor (Secor, 2003)proposed that 55% of the metabolic cost of digestion can be attributed to the production of HCl and other gastric functions related to breakdown of the food bolus. This study measured an increase in oxygen consumption up to 5 days after feeding (Fig. 1A). Part of this cost includes the secretion of hydrogen ions into the lumen of the stomach via the H⁺,K⁺-ATPase, however, in the present study it appears that H⁺ secretion does not continue past 3 days (Fig. 2). Blocking this H⁺,K⁺-ATPase(DiPalma, 2001) should give an estimate of the relative cost of H⁺ transport. However, while omeprazole successfully blocked the alkaline tide in B. constrictor,the postprandial increase in metabolic rate was indistinguishable from untreated animals (Andrade et al.,2004). This result suggests that the cost of H⁺transport is negligible in terms of the overall cost of digestion in B. constrictor. Other processes contributing to the cost of digestion include protein synthesis (Coulson and Hernandez, 1971; McCue et al.,2004), biosynthesis (mass) and nutrient transport(Secor and Diamond, 1995);however, the cost of increasing intestinal mass has been called into question by more recent studies (Overgaard et al.,2002; Starck and Beese,2001; Starck and Beese,2002) leaving the attribution of costs associated with digestion an open and interesting question.

It is concluded that carnivorous reptiles develop a metabolic alkalosis during digestion. Recent studies of alligators, pythons, toads and boas, as well as this study on monitor lizards, support the hypothesis that this metabolic alkalosis is partially or completely compensated by a respiratory acidosis (Andersen and Wang,2003; Busk et al.,2000b; Coulson et al.,1950b; Hicks et al.,2000; Overgaard et al.,1999; Wang et al.,2001). The degree of compensation may explain the variability of the pH change during digestion in different species; however, why there are different degrees of compensation remains unclear.

We thank Johnnie Andersen, Amanda Szucsik, and two anonymous reviewers for comments on this manuscript; also, we thank Björn Platzack and Dane A. Crossley II for helpful advice on an earlier version of this manuscript. This paper includes a portion of the work from the dissertation of L.K.H. This research was supported by NSF grant IBN-9727762 to A.F.B. and J.W.H.