Measurement of Carbon Dioxide Production from Radiolabeled Substrates in Drosophila melanogaster

This protocol describes a method for measuring the oxidation of specific radiolabeled substrates in adult Drosophila melanogaster. This technique is straightforward, quantitative, reproducible and sensitive, and it complements existing methods that measure total CO₂ production and O₂ consumption because it can be used to identify the nutrient(s) being oxidized for ATP production.

Measurement of CO₂ liberated from the oxidation of specific radiolabeled substrates using the method described here is complementary to techniques that measure total CO₂ production. Total CO₂ production (V_CO2) allows estimation of metabolic rate and, when total O₂ consumption (V_O2) is known, the metabolic rate can be calculated and the fuel source for oxidation can be estimated based on the respiratory exchange ratio (V_CO2 / V_O2 = RER). When carbohydrates are the exclusive substrate, RER = 1, but when fatty acids are the exclusive source, RER = 0.7, owing to the stoichiometry of the reactions of glucose and lipid oxidation. In the absence of equipment to measure oxygen consumption in Drosophila¹⁴, the oxidative fuel source cannot be identified. However, by labeling flies with trace amounts of one radioactive substrate or another and measuring rates of radiolabeled CO₂ production, one can draw conclusions about the ability of flies to oxidize a given substrate at different time points during a stimulus or about the effects of a given mutation on fuel oxidation.

There are a number of critical steps in this protocol. First, care should be taken when using radioactive materials to avoid spills and contamination of research areas. Second, when anesthetizing flies during steps 2.1 and 3.4, the experimenter should work quickly to avoid effects of prolonged CO₂ exposure on metabolism and behavior. A group of 20 flies can be anesthetized and transferred to a new vial or into a fly pod in 20 sec or less. When transferred from an anesthesia apparatus to a food vial, the vial should be rested on its side to prevent anesthetized flies from getting stuck on the food surface. When flies are fed radiolabeled food over the course of several days as in step 2.2, the experimenter should insure that the food does not dry out as flies are sensitive to dehydration.

The fed or fasted state prior to labeling, the choice of the label, the length of labeling time, the length of time in the fly pod, and the method of anesthesia are parameters that can be subjected to troubleshooting and modification when using this technique. First, flies subjected to short labeling periods (2 – 3 hr) are unlikely to consume significant quantities of radiolabeled food unless subjected to a fasting period beforehand. Second, the choice of label can affect the conclusions one is able to draw about metabolic phenotypes. For example, radiolabeled CO₂ production from D-[1-¹⁴C]-glucose can reflect oxidation through the citric acid cycle or the pentose phosphate shunt, whereas radiolabeled CO₂ production from D-[6-¹⁴C]-glucose reflects oxidation through the citric acid cycle only^12,15,16. Feeding flies with a radiolabeled tracer for an essential amino acid^17,18 would be a good strategy for assessing oxidation of amino acids derived from proteins. Finally, lengthy labeling periods with radiolabeled glucose also raise the possibility of incorporation of this precursor into other classes of molecules such as triglycerides¹⁹ and amino acids such as alanine, which readily interconverts with pyruvate. This can be assessed by measuring radioactivity incorporated into these different classes of molecules⁶. Indeed, separate cohorts of flies can be fed radiolabeled substrates in the same manner and then used for fly pod experiments or measurement of radiolabel that is stored and remains in glycogen and triglycerides in fed and fasted conditions⁶, for example. Another strategy is to use short-labeling periods that are likely to reveal oxidation of the radiolabeled dietary nutrients themselves and not storage forms of these nutrients. Third, it is also possible that two groups of flies could have identical rates of ¹⁴CO₂ production over a given amount of time, but very different rates initially. Therefore, varying the amount of time that flies spend in the fly pod may be an important parameter to alter when assessing phenotypic variation. Finally, this protocol calls for the use of a short period of CO₂ anesthesia when putting flies into the fly pod apparatus. It is possible that the CO₂ anesthesia may affect metabolic function over the course of the fly pod experiment. An alternate approach is to use nitrogen anesthesia which does not affect metabolic rate ²⁰.

As with any technique that measures such a complex system as fuel oxidation and metabolic rate, the method described here has limitations. First, it is possible that flies in different groups could consume different amounts of radiolabeled food and thus would enter the CO₂ collection phase of the assay with different starting amounts of substrate. This can be controlled for to some extent by performing the experiment with large sample sizes and by feeding flies en masse. For short labeling periods, flies with similar blue-colored guts will have consumed the radiolabel and can be carried forward to the chase portion of the experiment. The amount of radiolabel remaining in a group of flies can also be measured at the end of the experiment and in separate cohorts of flies after the end of the feeding and initial chase period. Second, the activity levels between groups of flies in the fly pods may differ. This will have to be determined experimentally and taken into account when interpreting data. Third, this technique does not measure total CO₂ production, only production from a specific dietary nutrient or the storage form(s) of that nutrient. This protocol is not a replacement for but instead is complementary to measurements of VCO₂ because it provides different and additional information.

The method described here measures exhaled, radiolabeled CO₂ that is a product of the oxidation of trace amounts of specific metabolic substrates. By varying the label used – glucose, fatty acid, or amino acid – one can design increasingly sophisticated experiments to assess the contributions of different substrates to metabolism under different physiological conditions such as starvation and on different genetic backgrounds. Future applications of this technique include measurement of metabolism in the larval and pupal stages of development, two stages of the Drosophila life cycle that are characterized by extremes of nutrient storage and breakdown, respectively.