INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE USE AND MEASUREMENT of carbon isotopes have been critical to the understanding of nutrient utilization. Protein, carbohydrate, and lipid homeostasis has been assessed under a variety of conditions using ¹⁴C- and ¹³C-labeled substrates of leucine, glucose, and palmitate, respectively. These labeled substrates liberate labeled CO₂ when they are oxidized. Thus the whole body oxidative rate of these labeled substrates can be measured by collecting expired air and determining the amount of labeled CO₂ produced.

However, previous studies have suggested that the CO₂ produced is not completely recovered under all circumstances. Recovery of CO₂ has been assessed under a variety of conditions using different experimental techniques, and there is a wide variation in the reported values (~50-100%) (1-6, 10, 13, 15, 22, 24, 27-29). A portion of the incomplete recovery has been attributed to fixation of CO₂ through conversion of pyruvate to oxaloacetate (4, 7, 26, 28). Hence, organs with increased enzymatic activity for this metabolic process, such as the liver, should also have lower rates of CO₂ recovery. However, there are no published reports on the tissues and organs responsible for this incomplete recovery. Furthermore, it is unclear how much of the variation in reported CO₂ recoveries is due to differences in experimental conditions and techniques vs. the portion due to fixation of CO₂.

Therefore, in vivo studies with multicatheterized canines were conducted to elucidate the contributions of the liver, gastrointestinal tract (gut), muscle, and kidneys to the fixation of CO₂ and to test the hypothesis that the loss of CO₂ is greater in organs that are involved in gluconeogenesis. Because there are no published data that examine the contribution of these organs to CO₂ fixation in a comprehensive manner in vivo, appropriate fixation correction factors for measurements of organ oxidation of a ¹⁴C- or ¹³C-labeled substrate are unavailable. Therefore, studies also were conducted to determine the importance of appropriate correction factors for each organ that was tested.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals and surgical procedures. Seven mongrel dogs [20.6 ± 0.8 (SE) kg] of either sex were used in this study. Approximately 17 days before the metabolic study, a midline laparotomy was performed under general anesthesia, as previously described (31), to place chronic catheters in the left femoral artery, right common iliac vein, hepatic vein, portal vein, and, in two cases, the left renal vein. Doppler flow probes (Instrumentation Development Laboratory, Houston, TX) were placed around the left external iliac and renal arteries for the measurement of blood flow (12). After placement, the catheters were filled with heparinized saline (200 U/ml), and the free ends were knotted and placed in a subcutaneous pocket, along with the leads of the Doppler flow probes, until the study day. A tracheostomy was performed for the collection of breath, as previously described (31).

Experimental design. Each study (Fig. 1) consisted of an initial 15-min bolus tracer decay period (15 to 0 min), a 90-min equilibration period (0 to 90 min), and a 90-min sampling period (90 to 180 min). A stock solution of isotope was made by dissolving NaH¹⁴CO₃ (ICN Biomedicals, Irvine, CA) in 0.1 N NaOH (25 µCi/ml). The infusate for the constant infusion portion of the study was made immediately before the experiment by diluting stock solution with normal saline and placing it in infusion syringes. A portion of the infusate was sampled before and immediately after the study and counted for radioactivity. Radioactivity in the samples of infusate was not different between these two sampling times. Before bolus tracer injection, a baseline blood sample was taken. A bolus injection of NaH¹⁴CO₃ (1.76 ± 0.16 µCi/kg) was administered over 10 s to initiate the study. Subsequently, arterial blood was sampled at time (t) = 15 (immediately), 13, 11, 9, 6, 3, and 0 min. The constant infusion of NaH¹⁴CO₃ (0.022 ± 0.002 µCi · kg¹ · min¹) was initiated at t = 0 min.

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Experimental design. Each experiment consisted of 3 periods: bolus decay (A) and equilibration and experimental (B). The 15-min bolus decay period was initiated with a bolus injection of Na14CO2. This period was immediately followed by a 180-min constant infusion of Na14CO2. Steady state of arterial 14CO2 was reached after 90 min of equilibration. Arrows indicate blood and breath sampling time points.

Collection and processing of samples. Expired air for the measurement of ¹⁴CO₂ was collected into a 30-liter Douglas bag over a 2-min period. To trap CO₂ for scintillation counting, air in the Douglas bag was immediately bubbled through a solution of 2 ml absolute ethanol, 0.25 mg phenolphthalein, and 50 µl hyamine hydroxide (methylbenzethonium hydroxide; Sigma Chemical, St. Louis, MO), until the indicator turned from purple to clear. The trapping vials were prepared in one large batch, from which a subset was titrated with 1 N HCl to determine the microequivalents of CO₂ that would be extracted by the basic solution. After trapping the expired CO₂, 2 ml of 0.5 N NaOH and 19 ml of scintillation fluid [EcoLite⁽⁺⁾, ICN Biomedicals] were immediately added. The samples were left in the dark for 4 days to minimize chemiluminescence and were then counted in a liquid scintillation counter (LS-2800, Beckman Instruments, Palo Alto, CA).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Procedure for trapping blood 14CO2. A plastic well containing chromatography paper and 200 µl of hyamine hydroxide was suspended above 1 ml of 6 N HCl in scintillation vials before the study day. The vial, which was made airtight with a rubber stopper, had 2 ml of air evacuated to create a slight negative pressure within the vial. 1 ml of blood was pipetted through an 18-gauge needle into the HCl solution, and the samples were allowed to stand overnight so that the CO2 could be liberated from the blood and absorbed by the hyamine hydroxide. The next day, the chromatography paper and well were transferred to a second scintillation vial containing 2 ml of 0.5 N NaOH. Scintillation fluid was then added to fill the vial. The samples were counted for radioactivity after 4 days in darkness.

2

1

Calculations. Decay curves were analyzed by performing a multiple-regression ANOVA on the curve following the bolus injection to test for the best fit (Statistical Analysis System for Windows, 1996 Release 6.12, SAS Institute, Cary, NC). One-, two-, three-, and four-exponential models were tested. No additional advantage was noted beyond a two-exponential curvilinear model. Therefore, this model was used to analyze each individual study, and the results are reported as averages of these analyses.

(1)

(2)

(3)

(4)

(5)

(6)

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The dogs were in a normal overnight-fasted state, as was reflected by their circulating glucose and metabolic hormone concentrations. Arterial glucose concentrations remained constant throughout the time course of each study (108.8 ± 1.8 mg/dl). Insulin (10.5 ± 1.3 µU/ml), cortisol (6.0 ± 0.4 µg/dl), epinephrine (197 ± 43 pg/ml), and norepinephrine (180.8 ± 18.7 pg/ml) also remained steady for all dogs.

Pool size for the distribution of the CO₂ tracer was determined from the decay curve of the bolus tracer infusion (Fig. 3A, Table 1). The best-fit curve after the bolus tracer injection contained two components. Slopes of the rapidly perfused pool vs. the central pool were 4,626 ± 1,391 vs. 244 ± 53 dpm/min, respectively. It was assumed that the first component, which lasted ~2 min, represented the mixing phase for the rapidly perfused tissues. The second component, assumed to represent the central or primary pool, was measured from 13 to 0 min. The data fit this line with a correlation coefficient (r) of 0.9998. The pool size was calculated to be ~80% of the total body pool at 16.0 ± 1.7 liters.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Arterial CO2 radioactivity is depicted in disintegrations per min (dpm) per ml after the bolus infusion [time (t) = 15 to 0 min; A] and during the constant infusion (0-180 min; B). The values are means ± SE (n = 7 studies).

Steady state for the recovery data was achieved during the final 90 min, as illustrated in Fig. 3B. During this steady-state period, whole body CO₂ recovery was 97.2 ± 6.7% (Fig. 4). CO₂, based on measurements of tracer dilution, was 351 ± 18 µmol · kg¹ · min¹. This compared favorably with CO₂ measured by the respiration cart, which was 377 ± 28 µmol · kg¹ · min¹. Furthermore, the similarities of these two measures of CO₂ for the seven dogs tested are demonstrated by their significant correlation (r = 0.69) as well as the closeness of their relationship to identity. With perfect identity, the equation describing the relationship of these two measures would have a slope of 1.0 and a y-intercept of zero. However, the actual relationship was calculated to have a slope of 1.08 and a y-intercept of 2 µmol · kg¹ · min¹.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Whole body and organ recovery of 14CO2. Recoveries of CO2 for whole body, kidney, hindlimb, gut, and liver are presented in units of percentage. The values are reported as means ± SE for 7 studies, except for the kidney, where n = 2.

During the steady-state period, blood flows through the tissues of the gastrointestinal tract, liver, hindlimb, and kidney were 587 ± 32, 734 ± 46, 205 ± 10, and 186 ± 5 ml/min, respectively. Radioactivity in each milliliter of blood was between 1,670 and 1,770 dpm, and specific activities for CO₂ within the portal, hepatic, femoral, and renal veins were 75 ± 20, 76 ± 19, 70 ± 16, and 73 ± 14 dpm/µmol, respectively. Recoveries across the liver, gastrointestinal tract, leg, and kidney were 99.9 ± 1.3, 98.0 ± 1.4, 96.7 ± 2.6, and 99.9 ± 2.1%, respectively (Fig. 4). CO₂ by the gastrointestinal tract, liver, hindlimb, and kidney, as determined by tracer dilution measurements, were 12.4 ± 0.7, 15.4 ± 0.9, 4.6 ± 0.2, and 3.9 ± 0.1 mmol/min, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

There is considerable diversity reported for the measurements of CO₂ recovery in vivo. The aim of the present study was to determine CO₂ recovery across several major organs in addition to whole body CO₂ recovery in vivo. The data obtained from conscious, postabsorptive canines in this study suggest that whole body CO₂ recovery was slightly <100%. This is supported by the fact that measurements of whole body CO₂ were similar using either an indirect calorimeter or tracer dilution. Furthermore, recovery across the liver, gastrointestinal tract, leg, and kidney were also slightly <100%. Although CO₂ was considerably different between the four organs measured (liver > gastrointestinal tract > hindlimb > kidney), there were no significant differences between the gluconeogenic and nongluconeogenic organs in terms of CO₂ recovery. These data suggest that only minor corrections for incomplete recovery of CO₂ are needed when measuring oxidation of substrates across these organs.

Whereas CO₂ is the end product of substrate oxidation and is normally eliminated in the breath, there is a portion that theoretically can be reincorporated or "fixed" into macromolecules. Most higher animals have the ability to fix CO₂ as the carboxyl carbon of oxaloacetate. For example, the enzyme pyruvate carboxylase combines CO₂, ATP, and pyruvic acid in the formation of oxaloacetic acid, ADP, and inorganic phosphate. In a net sense, no additional glucose is formed from CO₂ fixation, as another CO₂ molecule is lost in the subsequent reactions when oxaloacetate is converted to phosphoenolpyruvate and, ultimately, glucose. However, even though there is not a net accumulation of total CO₂, labeled CO₂ formed from substrate oxidation can be taken up on a net basis. The incorporation of labeled CO₂ molecules into glucose molecules has been used as a basis for the estimation of gluconeogenesis (7, 20, 23).

The fact that CO₂ recovery was relatively high in the present study does not invalidate either the concept of CO₂ fixation or the aforementioned method of estimating gluconeogenesis. However, it does demonstrate the comparative magnitude of these events and suggests that CO₂ fixation would minimally impact measurements of substrate oxidation across the organs that were measured. A previous report of gluconeogenesis using radioactive bicarbonate infusion lead to similar conclusions (7). In this previous study, the radioactivity associated with CO₂ in 1 ml of blood (~180 dpm) was 10-fold greater than the simultaneous measurement of the radioactivity associated with glucose (~18 dpm in 1 ml blood). This necessitates either special counting techniques for radioactivity, greater quantities of blood collection, or larger doses of radioactivity during bicarbonate infusion. Although most of the tracer decay in the present study could be attributed to either rapidly perfused tissues or a central pool, the existence of a very slowly perfused pool is possible. The rate of entry into this slowly perfused pool would be tremendously smaller than the rates of entry into the rapidly perfused and central pools, thereby making measurements for such a pool difficult. Taken together, these findings demonstrate that the pathway of CO₂ fixation is proportionally small compared with the overall disposition of CO₂.

Although whole body CO₂ recovery in the present study approaches 100%, there is considerable variation in previously reported values. One potential factor that may contribute to this variation is that differing physiological conditions may increase fixation of CO₂. Prevailing physiological factors, such as abnormal acid-base balance or a hormonal milieu that promotes gluconeogenesis, have been reported to have significant effects on CO₂ recovery (10, 21, 25). If there were a perturbation that would result in a severalfold increase in the activity of enzymes that fix CO₂, recovery measurements may be increased, supporting the notion that recovery needs to be assessed in each particular experimental setting. However, the magnitude of this increase would have to be very large. The dogs in the present study were in a postabsorptive state; thus a significant portion of glucose production would already be coming from events of gluconeogenesis. However, even if the events of gluconeogenesis and the pathways of CO₂ fixation were doubled from this postabsorptive state, our data would suggest that the measurement of fixed CO₂ would still be minimal.

Differences in experimental designs may result in variations of measured recovery of CO₂. However, recovery does not appear to be species specific, as high and low recoveries have been reported in several species, including rats, dogs, and humans (1, 3-6, 9, 13, 14, 16, 22, 24, 27-29). Furthermore, there does not appear to be a relationship between study duration and CO₂ recovery. Studies from 3 to 36 h in length have produced similar variability. Likewise, the recovery of CO₂ is not affected by the use of either stable (¹³C) or radioactive (¹⁴C) isotopes, by the use of primed vs. unprimed continuous infusions, or by the amount of isotope infused (infusion rates ranging from 1.8 to 30 µCi · kg¹ · min¹ have produced similar variation).

Theoretically, the site of cellular oxidation or compartmentalization could play a significant role in the amount of CO₂ recovered. The incorporation of CO₂ into glucose via oxaloacetate or into urea via ureagenesis occurs within the mitochondria. [-¹⁴C]ketoisocaproate (KIC) is also oxidized within the mitochondria. In an experiment in which KIC was infused, secondary labeling of glucose and urea was similar, suggesting that mitochondrial compartmentalization did not exist (7). This is in agreement with previous in vivo and in vitro studies that concluded there is an absence of compartmentalization within hepatic mitochondria (18, 19). However, when either carbon-labeled KIC or bicarbonate was infused, mitochondrial ¹⁴CO₂ was greater in the KIC group (7). This suggests that ¹⁴CO₂ derived endogenously within the mitochondria (KIC) resulted in a greater mitochondrial ¹⁴CO₂ specific activity than when derived exogenously outside the mitochondria (NaH¹⁴CO₂). Furthermore, incorporation of ¹⁴CO₂ into glucose was ~50% greater in the KIC group. This result was similar to preliminary data from another laboratory (9). Together, these data suggest that compartmentalization within the cell may make an impact when quantifying the oxidation of a substrate; however, the results from the present study suggest that these events may have negligible effects on the overall measurements of CO₂ recovery.

CO₂ collection and measurement methods provide a likely basis for the variation in recoveries that have been reported. It is difficult to determine whether sample-processing errors were a factor in the discordant results of previous studies, and these errors could make significant contributions to decreased recoveries. For optimal assessment of whole body recovery, a system that accurately determines expiration of labeled CO₂ is necessary. In the present study, we collected and measured expired air continually via a permanent tracheostomy. This method assured that total CO₂ was measured accurately and completely. Whereas a tracheostomy is not required for complete measurement of labeled CO₂, mechanical factors are a source of recovery error that must be assessed in each experimental situation.

Another potential methodology error relates to the complete "trapping" of CO₂. The trapping process requires the formation of ionic bonds, which may dissociate. In the present experiments, excess base (1,000 µmol NaOH) was added to the scintillation vials, after the CO₂ was trapped, to prevent dissociation. Laboratory experiments before the in vivo studies demonstrated that >500 µmol of NaOH were required to decrease the loss of radioactivity over time (Fig. 5A). Additional tests also demonstrated the importance of using greater volumes of scintillation fluid and decreasing the air volume within the scintillation vial to complete recovery of ¹⁴CO₂ (Fig. 5B). When a vial with radioactive CO₂ in scintillation fluid without excess base sat for several hours, radioactive counts decreased (Fig. 5C). When the same scintillation vials were shaken to mix the phases of liquid and gas, radioactive counts increased significantly. Finally, radioactivity was irreversibly lost when the cover of the same vial was removed. Thus filling the vial with scintillation fluid is recommended to prevent CO₂ from moving into a layer of air above the fluid and away from the scintillate.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Importance of technique for 14CO2 recovery measurements. A: addition of NaOH blunts loss of 14CO2. Differing molar amounts of NaOH (, 20 µmol; *, 200 µmol; ×, 500 µmol;  750 µmol; , 1,000 µmol) were added to 23-ml scintillation vials containing radiolabeled CO2 (130,000 dpm) and 10 ml of scintillation fluid. Vials with <500 µmol NaOH had decreased dpm as time progressed. B: increased scintillation fluid volume blunts loss of 14CO2. Differing amounts of scintillation fluid (, 10 ml; , 15 ml; , 23 ml) were added to 23-ml scintillation vials containing radiolabeled CO2 (120,000 dpm). With 10 ml of scintillation fluid, dpm decreased over time, but, as the amount of scintillation fluid was increased to fill the vial, dpm was maintained. C: CO2 in scintillation fluid leaves the liquid phase and enters the gas phase when excess base is not present. NaH14CO3 (130,000 dpm) was added to 23-ml scintillation vials containing 10 ml of scintillation fluid. 1,000 µmol of NaOH were added (, scintillation fluid + NaOH) to one set of vials, and a second set had no NaOH added (, scintillation fluid). After 14 h, both vials were shaken (S) and, after 17 h, vials in both groups were opened (O) for 5 min. In the group with NaOH, dpm did not decrease, and neither shaking nor opening altered the counts. When NaOH was not added, dpm decreased with time. When the vials without NaOH were shaken, dpm increased, suggesting that the CO2 returned to the liquid scintillation layer. When the vial was left open for 5 min, dpm decreased, suggesting that the CO2 was released into the surrounding atmosphere.

In conclusion, although there is considerable diversity in the measurements of CO₂ recovery in vivo, whole body CO₂ recovery was slightly <100% in the present study of conscious, postabsorptive canines. Furthermore, recoveries across the gastrointestinal tract, leg, liver, and kidneys also were near 100%. This suggests that corrections for incomplete recovery of CO₂, when measuring oxidation of substrates across these organs under normal postabsorptive conditions, would be very minor.