Share this article on:

Anesthetic Properties of Carbon Dioxide in the Rat

Brosnan, Robert J. DVM, PhD*; Eger, Edmond I. II MD; Laster, Michael J. DVM; Sonner, James M. MD

doi: 10.1213/01.ane.0000265556.69089.78
Anesthetic Pharmacology: Research Report

BACKGROUND: Carbon dioxide decreases halothane minimum alveolar concentrations (MAC) in dogs when Paco2 exceeds 95 mm Hg. We sought to confirm these findings for several potent inhaled anesthetics in rats.

METHODS: Groups of eight rats were anesthetized with halothane, isoflurane, or desflurane. MAC was determined for each anesthetic alone, and then with increasing concentrations of inspired CO2. A fourth group was given CO2 alone to determine the MAC of CO2.

RESULTS: Increasing inspired CO2 concentrations produced a linear dose-dependent decrease in MAC of each potent inhaled anesthetic. With elimination of CO2, the MAC of isoflurane and desflurane returned to the original MAC. As determined by extrapolating these data to 0% of the inhaled anesthetic, the MAC of CO2 was approximately 50% of 1 atm. Given alone, CO2 proved lethal.

CONCLUSIONS: Unlike dogs, no threshold for the CO2-MAC response arose with halothane, isoflurane, or desflurane in rats. The ED50 for CO2 is also approximately 50% greater in rats than reported in dogs.

IMPLICATIONS: Increasing concentrations of CO2 rectilinearly decrease the MAC of halothane, isoflurane, and desflurane with no threshold to this effect. MAC of CO2 in the rat is approximately 50% of 1 atm.

From the *Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California, Davis; and †Department of Anesthesia and Perioperative Care, University of California, San Francisco, California.

Accepted for publication March 19, 2007.

Supported by NIH Grant 1PO1GM47818.

Dr. Eger is a paid consultant to Baxter Healthcare Corp. Baxter Healthcare Corp. donated the desflurane and isoflurane used in these studies.

The work is attributed to the Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California, Davis, and Department of Anesthesia and Perioperative Care, University of California, San Francisco.

Address correspondence and reprint requests to Robert J. Brosnan, DVM, PhD, Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California, Davis, CA 95616. Address e-mail to

The anesthetic effects of CO2 have long been recognized (1). Although CO2 acts as an “inhaled” anesthetic, the mechanism mediating CO2 narcosis may differ from that underlying anesthesia produced by conventional inhaled anesthetics. For example, the potent anesthetics isoflurane and halothane exhibit additivity; that is, administration of half an isoflurane minimum alveolar concentration (MAC) (the MAC of inhaled anesthetic required to suppress movement in response to noxious stimulation in 50% of subjects) plus half a halothane MAC produces the same percent immobilization as either anesthetic alone at MAC (2). In contrast, increasing Paco2 to 95 mm Hg in dogs does not significantly decrease halothane MAC (3). And hypocapnia in humans (4) or dogs (3) does not increase halothane MAC. Increases in CO2 exceeding 95 mm Hg decrease halothane MAC in a dose-dependent manner to zero at 245 mm Hg CO2 (3). The absence of a change in MAC with increasing CO2 partial pressures up to 95 mm Hg suggests a mechanism of action for CO2 that differs from that underlying halothane.

However, studies using end-points other than MAC have detected anesthetic effects from mild CO2 perturbations. In humans, hypocapnia increases the partial pressure of nitrous oxide required to produce incoordination and unconsciousness, and mild hypercapnia (Paco2 = 50–60 mm Hg) decreases the concentration of nitrous oxide needed to sustain consciousness (5). There is no CO2 threshold for either effect. Normal human subjects breathing CO2 concentrations up to 10% exhibit impaired dexterity and mental acuity; still larger inspired concentrations cause unconsciousness (6). Recently, unpublished observations from our laboratory suggest that moderate hypercapnia in rats might reduce the isoflurane concentration required for immobility. Therefore, the anesthetic-sparing effects of CO2 might depend on the species, phenotype, drug, and anesthetic end-point studied.

We hypothesized that if narcotic effects of CO2 were mechanistically similar in all species, then increasing Pco2 in rats should yield the threshold effect seen dogs, and that CO2 should not be additive with MAC for any volatile anesthetic. The present work tested these hypotheses.

Back to Top | Article Outline


We used 32 adult male Sprague-Dawley rats weighing 314 ± 48 g (mean ± sd) obtained from Charles River Laboratories (Hollister, CA). Husbandry before experiments included commercial rat chow and water available ad libitum and a 12-h light–dark cycle. We conducted four separate anesthetic (MAC) studies of the response to CO2, each with eight rats, using either halothane, isoflurane, desflurane, or CO2 alone as the anesthetic. Each rat was studied only once. The Animal Use and Care Committee at the University of CA, San Francisco approved this protocol.

A description of the anesthetic delivery apparatus and MAC determination techniques used in this study has been published (7). Briefly, 8.8 × 25 cm cylindrical Plexiglas chambers were connected in parallel each receiving gas through a one-holed rubber stopper from an agent-specific and temperature-compensated vaporizer. The fresh gas flow rate was ≥1 L · min−1 ·cylinder−1. A needle capped with a stopcock pierced each stopper, permitting gas sampling from the tube. The other end of each Plexiglas chamber was sealed with a two-holed stopper; rat tails were drawn through one hole and a passive gas scavenging system was attached to the second hole. A thermistor probe also traversed a sealed portion of the two-holed stopper to measure rectal temperature, which was maintained at 37°C ± 1°C using heating pads as needed.

Rats then were exposed to a sub-MAC concentration of the study anesthetic delivered in 100% oxygen for approximately 45 min, after which movement in response to 60-s tail clamping was assessed. Concentrations of inspired anesthetic were measured by gas chromatography (GOW-MAC Series 580, Bridgewater, NJ) using a 4.6-m, 0.22-cm internal diameter SF-96 packed column maintained at 100°C and a flame ionization detector maintained at 150°C. The chromatograph was calibrated before and several times during each experiment with anesthetic gas standards; linearity of the detector was confirmed using dilutions of the gas standards. After tail clamping and gas sampling, the anesthetic concentration was increased by 15%–20% of a normal MAC for that anesthetic and held constant for 30 min to allow equilibration between inspired and alveolar partial pressures (7). Response to tail clamp was reassessed and gas concentration was again measured by gas chromatography. This was repeated with 30 min equilibration for each change in anesthetic concentration until all rats were immobile in response to tail clamping. The arithmetic mean of the anesthetic concentrations that bracketed move/no move responses defined individual MAC values (8). At this point, a certified CO2 gas standard in oxygen replaced the O2 carrier gas (Airgas, Sacramento, CA) and the inspired anesthetic concentration was decreased until all animals moved. MAC was then determined for the anesthetic in CO2 as previously described. Tanks containing progressively higher certified concentrations of CO2 (10.8%, 20.9%, 31.4%, 41.8%, 51.6%, 63.2%) were used and MAC retested until immobility was achieved in all rats with no addition of a volatile anesthetic. Most CO2 carrier gas concentrations were studied in combination with all anesthetics. Finally, MAC for the anesthetic alone (no CO2) in 100% O2 carrier gas was remeasured to verify the reversibility of the effect of CO2 on anesthetic requirement.

In a separate experiment, direct CO2 MAC measurement was attempted in eight rats. After placement in the plexiglass cylinder as previously described, 51.6% CO2 in O2 was acutely administered to all rats in the absence of a volatile anesthetic. Subsequent MAC testing was not possible because of the development of pulmonary edema in most animals.

During CO2 inhalation, Pico2 was used to approximate Paco2 because rats were not instrumented to obtain end-tidal or arterial gas samples. However, in the absence of neuromuscular impairment or fatigue, alveolar ventilation increases in proportion to inspired CO2 fraction. The inspired-to-arterial CO2 difference is less than 10 mm Hg in rats breathing gas mixtures containing ≥5% CO2 (9). Even at light planes of anesthesia, Pico2 under these conditions provides a reasonable estimate of Paco2 (10).

Descriptive data are presented as mean ± sd. To compare CO2 effects among anesthetics, anesthetic ED50 for each rat was calculated as a MAC fraction by dividing the MAC for anesthetic plus CO2 by the MAC for the anesthetic alone for that rat. The effect of Fico2 on MAC fraction was analyzed using least squares linear regression (SPSS, v.11, Chicago, IL). A one-sided paired t-test was used to determine whether the post-CO2 MAC determination was less than the pre-CO2 MAC determination for any of the three volatile anesthetics (an indication of injury from prolonged hypercapnia). t-Tests with Dunn–Sidak corrections for multiple comparisons were used to determine significant differences among MAC for all three anesthetics administered in 100% O2 (no CO2) and 10% CO2. Differences were considered significant if P < 0.05.

Back to Top | Article Outline


MAC values for halothane, isoflurane, and desflurane (mean ± sd) were 1.14% ± 0.11%, 1.42% ± 0.10%, and 7.21% ± 0.63% atm, respectively. Addition of CO2 rectilinearly and significantly decreased MAC of all anesthetics. For each 10% increase in Fico2, anesthetic requirement for all anesthetics decreased approximately 20%. No threshold response was observed (Fig. 1). MAC values for each anesthetic administered in 10% CO2 were significantly different than MAC measurements made with 100% O2 carrier gas. The duration of each experiment was approximately 8–10 h.

Figure 1

Figure 1

Although CO2–MAC response curves are essentially colinear for the three anesthetics, there were differences did at higher Fico2. The halothane-CO2 response for the eight rats appears to deviate from the other anesthetics at the x-intercept (Fig. 1). However, two rats given halothane developed labored breathing and died after this measurement. MAC of halothane administered in O2 at the end of the study in the remaining six animals (repeat of control MAC) was 0.93% ± 0.18%, a statistically significant difference that was 18% lower than the measurement preceding CO2 exposure (P < 0.05). MAC values for isoflurane and desflurane at the end of each experiment were 1.37% ± 0.26% and 6.90% ± 0.84%, respectively; these values are statistically indistinguishable from measurements made before CO2 exposure.

Our attempt to measure MAC of CO2 in the absence of a volatile anesthetic was unsuccessful. The acute exposure of eight rats to 40% CO2 in O2 produced dyspnea and fulminant pulmonary edema within 10–15 min. However, direct measurement of CO2 MAC for rats in the desflurane group was 49% ± 5% atm. Extrapolation of the isoflurane data predicted a CO2 MAC of 54% atm.

A pooled regression analysis for all data points for all anesthetics, excluding the erroneous halothane x-intercept during the measurement of which rats developed pulmonary edema, is described by the following equation: %MAC = −1.9 × Fico2 + 1.0, where %MAC is the percent decrease in ED50 for the volatile anesthetic and Fico2 is the fraction inspired carbon dioxide. This relationship has a correlation coefficient of −0.99 and predicts CO2 MAC to be 53% atm.

Back to Top | Article Outline


In conflict with our hypotheses, CO2 linearly and progressively decreased the anesthetizing (MAC) concentrations of conventional inhaled anesthetics, abolishing the need for continued inhaled anesthetic administration to produce immobility at 0.53 atm CO2. For desflurane and isoflurane, the effect of increasing Fico2 was reversible; the MAC on removal of the increased Fico2 equaled that before administration of CO2.

The finding of a rectilinear decrease in MAC in rats with increasing Fico2 contrasts with results from dogs. In dogs, Paco2 <95 mm Hg did not significantly decrease halothane MAC, although the average MAC was less than the control MAC (3). Measurement of end-tidal halothane and CO2 versus the present study measurement of inspired concentrations would not appear to explain the difference in findings, particularly for desflurane. Given the poor solubility of desflurane, the inspired-to-end-tidal difference should be minimal (11). Further to this point, an increase in Fico2 increases minute ventilation, even in anesthetized rats (12,13), and reduces dead-space ventilation; both these factors should additionally minimize differences between inspired and alveolar CO2 and anesthetic tensions (14). Respiration of 10% inspired CO2 does not increase alveolar CO2 by 10% of an atmosphere. Were the application of 10% CO2 to quadruple ventilation in the absence of increased metabolism, PaCO2 would equal approximately 80 mm Hg. In evidence, measurements in anesthetized rats rebreathing CO2 up to 70 mm Hg demonstrate that Pico2 closely approximates Paco2 (10). Thus, the gas sampled (end-tidal versus inspired) does not explain the response differences between dogs and rats. Other reasons for this discrepancy include species differences to CO2 narcosis or underestimation of the control canine MAC. Given the rat responses reported here, it may be of value to verify dog CO2 threshold and MAC measurements using the identical methods from this study.

MAC data for the volatile anesthetics in this study confirm values previously reported (7,15). MAC for CO2 is approximately 50% higher in the rat than in the dog. Why the rat should be less sensitive to the anesthetic effects of CO2 than the dog is unclear. As burrowing animals, rats could be exposed in nature to CO2 inspired fractions approaching 10% (16); decreased sensitivity to the anesthetic effects of CO2 might therefore provide an important physiologic adaptation. Tolerance to CO2 when awake may also help explain the tendency for rats anesthetized at 1.0 times MAC to hypoventilate more than dogs that are similarly anesthetized (17,18).

Based on solubility in oil (19), CO2 is approximately three times more potent in the rat than predicted by Meyer–Overton (20,21) for conventional anesthetics (22). The basis for the greater narcotic effect is not clear, but Eisle et al., suggested that it was associated with the capacity of CO2 to decrease extracellular pH (3).

We could not measure MAC of CO2 without first anesthetizing the animals. CO2 increases arterial blood pressure indirectly by increasing circulating catecholamines and angiotensin II (23). Perhaps these effects of CO2, combined with excitement of anesthetic induction, caused pulmonary hypertension leading to pulmonary edema.

Halothane combined with CO2 proved toxic at larger CO2 concentrations. Compared to contemporary haloether anesthetics, halothane produces more depression of left ventricular output (24) and possibly more pulmonary vasoconstriction (25). These changes could increase mean pulmonary arterial pressure and exacerbate pulmonary hypertension caused by severe hypercapnia, contributing to the CO2 toxicity we observed.

The linear relationship between CO2 concentration and volatile anesthetic concentrations (Fig. 1) also describes an isobologram, with points for all anesthetics lying on the line of additivity. If additivity between conventional anesthetic MAC values is consistent with (albeit, not proof of) a common site of action (2), then the same may also hold true for combinations of conventional anesthetics and CO2.

Distinctions between dogs and rats with respect to CO2 potency and additivity with other conventional anesthetics suggest that either there may be different anesthetic-sensitive targets in different species, or the common anesthetic targets in different species may vary substantially in their sensitivity to inhaled anesthetics. In comparisons with previous studies, differences in anesthetic potency between dogs and rats appear much larger for CO2 than for halothane (8), isoflurane (26), or desflurane (27). Perhaps this reflects greater species variation in primary target site(s) for CO2 narcosis than for sites responsible for volatile anesthetic action.

Finally, one might ask whether humans resemble dogs or rats in their response to CO2. If the response of humans parallels that of dogs, then patients with chronically increased CO2 values <95 mm Hg may be awake at, and seriously discomfited by, such increased values. However, if patient responses are those of the rat, then the increased CO2 will produce an element of anesthesia, an effect that might decrease the discomfort of such elevated levels of CO2. This also implies that the increased concentrations of CO2 seen during anesthesia in humans could contribute to anesthesia. Yet, whether humans might be more like rats or dogs in this respect remains, at present, a topic for phylogenetic (28,29) and philosophic debate.

Back to Top | Article Outline


1. Thompson CJS. Henry Hill Hickman. A forgotten pioneer of anesthesia. BMJ 1912;1:843–5.
2. Eger EI II, Xing Y, Laster M, Sonner J, Antognini JF, Carstens E. Halothane and isoflurane have additive minimum alveolar concentration (MAC) effects in rats. Anesth Analg 2003;96:1350–3.
3. Eisle JH, Eger EI II, Muallem M. Narcotic properties of carbon dioxide in the dog. Anesthesiology 1967;28:856–65.
4. Bridges BE Jr, Eger EI II. The effect of hypocapnia on the level of halothane anesthesia in man. Anesthesiology 1966;27:634–7.
5. McAleavy JC, Way WL, Altstatt AH, Guadagni NP, Severinghaus JW. The effect of Pco2 on the depth of anesthesia. Anesthesiology 1961;22:260–4.
6. Seevers MH. The narcotic properties of carbon dioxide. NY State J Med 1944;44:597–602.
7. White PF, Johnston RR, Eger EI II. Determination of anesthetic requirement in rats. Anesthesiology 1974;40:52–7.
8. Eger EI II, Saidman LJ, Brandstater B. Minimum alveolar anesthetic concentration: a standard of anesthetic potency. Anesthesiology 1965;26:756–63.
9. Pepelko WE, Dixon GA. Arterial blood gases in conscious rats exposed to hypoxia, hypercapnia, or both. J Appl Physiol 1975;38:581–7.
10. Hayashi F, Yoshida A, Fukuda Y, Honda Y. CO2-ventilatory response of the anesthetized rat by rebreathing technique. Pflügers Arch 1982;393:77–82.
11. Holdcroft A, Bose D, Sapsed-Byrne SM, Ma D, Lockwood GG. Arterial to inspired partial pressure ratio of halothane, isoflurane, sevoflurane and desflurane in rats. Br J Anaesth 1999;83:618–21.
12. Lai YL, Tsuya Y, Hildebrandt J. Ventilatory responses to acute CO2 exposure in the rat. J Appl Physiol 1978;45:611–18.
13. Martin-Body RL, Sinclair JD. Analysis of respiratory patterns in the awake and in the halothane anaesthetized rat. Respir Physiol 1985;61:105–13.
14. Scheid P, Meyer M, Piiper J. Arterial-expired Pco2 differences in the dog during acute hypercapnia. J Appl Physiol 1979;47:1074–8.
15. Eger EI II, Johnson BH. MAC of I-653 in rats, including a test of the effect of body temperature and anesthetic duration. Anesth Analg 1987;66:974–6.
16. Kuhnen G. O2 and CO2 concentrations in burrows of euthermic and hibernating golden hamsters. Comp Biochem Physiol A 1986;84:517–22.
17. Imai A, Steffey EP, Farver TB, Ilkiw JE. Assessment of isoflurane-induced anesthesia in ferrets and rats. Am J Vet Res 1999;60:1577–83.
18. Steffey MA, Brosnan RJ, Steffey EP. Assessment of halothane and sevoflurane anesthesia in spontaneously breathing rats. Am J Vet Res 2003;64:470–4.
19. Lawrence JH, Loomis WF, Tobias CA, Turpin FH. Preliminary observations on the narcotic effect of xenon with a review of values for solubilities of gases in water and oils. J Physiol 1946;105:197–204.
20. Meyer HH. Theorie der Alkoholnarkose. Arch Exp Pathol Pharmakol 1899;42:109–18.
21. Overton E. Studien über die Narkose Zugleich ein Beitrag zur Allgemeinen Pharmakologie. Jena: Verlog von Gustav Fisher, 1901:1–195.
22. Smith RA, Porter EG, Miller KW. The solubility of anesthetic gases in lipid bilayers. Biochim Biophys Acta 1981;645:327–38.
23. Staszewska-Barczak J, Dusting GJ. Importance of circulating angiotensin II for elevation of arterial pressure during acute hypercapnia in anaesthetized dogs. Clin Exp Pharmacol Physiol 1981;8:189–201.
24. Kazama T, Ikeda K. The comparative cardiovascular effects of sevoflurane with halothane and isoflurane. J Anesth 1988;2:63–8.
25. Takemura M, Shiokawa Y, Okamoto S, Uno H, Futagawa K, Koga Y. Volatile anesthetics constrict pulmonary artery in rabbit lung perfusion model. J Anesth 2005;19:343–6.
26. Steffey EP, Howland D Jr. Isoflurane potency in the dog and cat. Am J Vet Res 1977;38:1833–6.
27. Doorley BM, Waters SJ, Terrell RC, Robinson JL. MAC of I-653 in beagle dogs and New Zealand white rabbits. Anesthesiology 1988;69:89–91.
28. Rat Genome Sequencing Project Consortium. Genome sequence of the Brown Norway rat yields insights into mammalian evolution. Nature 2004;428:493–521.
29. Lindblad-Toh K, Wade CM, Mikkelsen TS, Karlsson EK, Jaffe DB, Kamal M, Clamp M, Chang JL, Kulbokas EJ III, Zody MC, et al. Genome sequence, comparative analysis and haplotype structure of the domestic dog. Nature 2005;438:803–19.
© 2007 International Anesthesia Research Society