Transpulmonary lactate shuttle

Animals and experimental preparations.

Twenty-one female Wistar rats (n = 7 per condition, body mass range, 240–290 g; Charles River Laboratories, Wilmington, MA) were used in these experiments, which were approved by the Animal Care and Use Committee at the University of California, Berkeley (AUP no. R017–1007). Prior to experimentation, rats were housed two per cage in a light- (light from 0700 AM to 0700 PM), temperature-, and humidity-controlled environment with unrestricted access to food and water.

Preparation for each experiment began with the induction of surgical anesthesia (isoflurane inhalation; 4% in 100% O₂ for induction, 2% for maintenance) followed by loose securing of the animal in supine position and connection to a system for continuous monitoring of arterial O₂ saturation (Sa_O2) by tail pulse oximetry. Arterial SO₂ remained between 94 and 99% for all experiments, and core temperature was maintained at ∼37°C by use of a thermostatically controlled heating pad. Two small skin incisions (∼5 mm) were made over the ventral thorax, and catheters inserted for blood sampling were positioned in the left common carotid artery and the right atrium (Intramedic PE-50; BD), as previously described (8). A third catheter was placed in the right iliac vein for infusion purposes. Surgical preparation of animals took ∼20 min.

Experimental protocol.

Immediately following surgical preparation, a blood sample of ∼200 μl was collected from the carotid artery for measurement of background isotope enrichments of lactate and pyruvate. Next, a primed continuous infusion of [U-¹³C]lactate, 99% enriched (Cambridge Isotope Laboratories, Andover, MA) was administered for 60 min via the iliac vein catheter using a syringe pump (Harvard Apparatus, South Natick, MA). Tracer was dissolved in 0.9% saline. For Con, lactate tracer was infused at 90 μg·kg⁻¹·min⁻¹ with a tracer prime corresponding to 15 times the minute infusion rate, an infusion rate and prime that we found to give steady-state values of ∼2% molar percent excess (MPE). Lactate tracer infusion rates corresponded to 288 and 577 μg·kg⁻¹·min⁻¹ during Epi, and LC conditions, respectively, again with the tracer prime corresponding to 15 times the minute tracer infusion rate. For Epi trials, the hormone cocktail was prepared immediately prior to infusion by dissolving stock epinephrine in 0.9% saline with ∼4 μg/ml sodium bisulfite to give a final epinephrine concentration of 4.35 μg/ml. Epinephrine was infused at 0.2 μg/kg/min. The total volumes of solutions administered (prime + constant infusion) were similar in all experiments (∼4 ml/kg).

Blood samples (150–200 μl) were collected simultaneously from arterial and right atrium sample sites (mixed central venous, v̄) after 40, 50, and 60 min of infusion. Cannulas were flushed with an equivalent amount of 0.9% saline after each collection. Animals remained anesthetized for the entire experiment, and after the last blood collection, animals were euthanized with an intravenous injection of pentobarbital sodium (150–200 mg/kg). Accurate catheter placement was confirmed by visual inspection during necropsy.

Processing and analysis of blood.

Blood lactate, pyruvate, and glucose concentrations were assayed enzymatically, as previously described (12). Briefly, sampled blood was immediately transferred to ice-chilled tubes containing 0.6 M of perchloric acid, shaken, and stored on ice until the end of the experiment. Within 1 h of collection, perchloric acid extracts were centrifuged [10 min at 3,000 rpm (g = 2,000), at 4°C], and the supernatants were transferred to separate tubes for storage at −20°C until further analysis.

Blood [lactate] was measured in neutralized perchloric extracts using an enzymatic method, as previously described (4). Blood lactate IE was determined using GCMS of the n-propylamide heptafluorobutyrate derivative (25). Briefly, neutralized perchloric extracts were lyophilized, resuspended in 200 μl of 2,2-dimethoxypropane and 20 μl 10% HCl in methanol, capped, and incubated at room temperature for 60 min. Following the addition of 50 μl of n-propylamine, the samples were heated at 100°C for 30 min, dried under a stream of N₂ and transferred to GCMS vials using ethyl acetate. Thereafter, the samples were dried under N₂, derivatized by adding 20 μl of heptafluorobutyric anhydride (5 min at room temperature), dried again under N₂, and resuspended in ethyl acetate for GCMS analysis. Methane was used for chemical ionization with selected ion monitoring for mass to charge ratios 328 (unlabeled lactate) and 331 (labeled lactate), respectively.

Blood [pyruvate] concentrations and IEs were determined using gas chromatography-mass spectrometry (GCMS: GC, model 6890 series; MS, model 5973N; Agilent Technologies, Santa Clara, CA) of the trimethylsilyl-quinoxalinol derivative, with α-ketovalerate as the internal standard for concentration measurements (12, 17). Briefly, 150 μl of perchloric extract was spiked with α-ketovalerate, mixed (1:1) with an ortho phenylenediamine solution (5 mg/ml in 3 M HCl) and heated for 60 min at 90°C. Pyruvate was subsequently extracted with methylene chloride, the aqueous layer discarded, and the remaining solution evaporated under a stream of N₂. The samples were subsequently derivatized with 50 μl of a pyridine-bis(trimethylsilyl)trifluoracetamide mixture (1:1). Chemical ionization (methane gas) was used with selected ion monitoring for mass to charge ratios 233 (unlabeled pyruvate), 236 (labeled pyruvate), and 261 (α-ketovalerate).

Blood [glucose] was measured in neutralized perchloric extracts using an enzymatic hexokinase kit (Pointe Scientific, Canton, MI).

Calculations.

The transpulmonary concentration gradient for lactate was calculated from concentrations in the right atrium (mixed central venous sample, v̄) and carotid artery [a]: transpulmonary concentration gradient (mmol/l) = [v̄] − [a] and net (lactate) balance (mmol/l) = ([v̄] − [a]) × Q, where [v̄] and [a] are the concentrations of lactate collected from the mixed central venous and arterial samples sites respectively, and Q is blood flow, as determined previously by Lin et al. (21). The value of Q used for the Con and LC conditions was 37 ml/min; and for the Epi condition, the estimated Q was doubled to 74 ml/min. Both positive transpulmonary concentration gradients and net balances reflect net pulmonary metabolite uptake, while a negative value reflects net metabolite release.

Concentration derived transpulmonary fractional extraction (FEXC) was calculated as FEXC = [v̄] − [a] / [v̄] × 100%

Tracer measured transpulmonary fractional extraction (FEXTM), tracer measured uptake, and total lactate release was calculated as FEXTM = [([¹³C]lactatev̄ IE)([lactate]v̄)] − [([¹³C]lactate_aIE)([lactate]_a)]/[([¹³C]lactatev̄ IE)([lactate]v̄)] × 100%, tracer measured uptake (mg/min) = Q × FEXTM × [lactate]v̄, and total lactate release (mg/min) = tracer measured uptake − net balance. Net rate of conversion of mixed central venous pyruvate to arterial lactate (P → L) was estimated as P → L, μmol/min = [(IEL_a/IEPv̄) × [lactate]v̄ − (IELv̄/IEPv̄) × [lactate]v̄] × Q, where IEL_a is IE of lactate in arterial blood, IEPv̄ is IE of pyruvate in the mixed central venous blood, and IELv̄ is IE of lactate in the mixed central blood.

While body lactate rate of appearance (Ra), rate of disappearance (Rd), and metabolic clearance rate (MCR) were calculated using the three blood sampling time points (40, 50, and 60 min into infusion) during each trial and the equations of Steele (25), as modified for use with stable isotopes.

where C₁ is the arterial lactate concentration at time 1 (t₁), and C₂ is the lactate concentration at time 2 (t₂). The percent that the lungs accounted for whole body lactate turnover (Ra) was calculated as (Ra_venous−Ra_arterial)/Ra_arterial×100% where Ra_venous used the enrichment values from the mixed central venous blood and Ra_arterial used the enrichment values from the arterial blood. The percent isotopic enrichment equilibration in the blood was calculated as IEL/IEP in simultaneously sampled arterial and mixed central venous blood samples, respectively.

Statistics.

Data are presented as means with deviation (SD). For statistical analysis, representative values of blood lactate concentrations and isotopic enrichments were obtained by averaging data taken during the last 20 min of each trial (i.e., from blood sampled at 40, 50, and 60 min) after infusion commenced. ANOVA indicated that ¹³C-isotopic enrichments of lactate in blood sampled prior to commencement of trials were not different among conditions and, therefore, values were pooled to obtain best estimates of background isotopic enrichments. Single one-way ANOVAs were used to test for treatment effects. Because concentrations did not always plateau during trials, a two-way ANOVA was used to determine the effect of treatment on lactate concentration. Following ANOVA, post-hoc testing to identify significant treatment effects was done using Tukey's honestly significant difference (HSD) tests; an α of 0.05 was used throughout. To evaluate the effect of the lungs on the lactate to pyruvate ratios, Student's paired t-tests with sequential Bonferroni correction factors were performed. Statistical tests were performed using SPSS Graduate Pack 11.0 software.

Mixed central venous and arterial lactate concentrations.

Preperfusion blood lactate [1.1 mM (SD 0.3)] and pyruvate [43.7 μM (SD 17.8)] concentrations were in the expected ranges (11, 12, 17, 19). Lactate concentration in the mixed central venous blood was lowest during the Con condition (Fig. 1A). Two-way ANOVA revealed an effect of treatment on lactate (P < 0.05; open triangles in Fig. 1) and multiple comparisons with Tukey's HSD showed that the LC significantly increased [v̄] lactate concentration (P < 0.05). Similarly, levels of [a] lactate were lowest during the control condition (Fig. 1B). Two-way ANOVA revealed an effect of treatment on arterial lactate concentration (P < 0.05), and multiple comparisons with Tukey's HSD showed that the lactate load significantly increased [a] lactate values (P < 0.05).

Transpulmonary lactate metabolism.

Transpulmonary lactate gradients ([v̄] − [a]) were positive during all three conditions (Table 1). One-way ANOVA revealed a treatment effect on the transpulmonary lactate gradient (P < 0.05), and multiple comparisons with Tukey's HSD showed that lactate loading significantly increased the transpulmonary lactate gradient above that during epinephrine stimulation (P < 0.05).

Net lactate balance was positive during all three conditions, indicating net uptake [1.2 mg/min (SD 0.7) for Con, 1.9 for LC, and 1.9 (SD 3.6) for Epi] (Table 1). One-way ANOVA did not reveal a treatment effect on net transpulmonary lactate balance (P < 0.05).

Concentration-measured fractional extraction of lactate calculated as a percent of net uptake averaged 13.1% (SD 16.6). One-way ANOVA did not reveal a treatment effect on FEXC (P < 0.05).

Tracer-measured fractional extraction averaged 16.6% (SD 23.8) for the Con and 8.2% (SD 15.3) for LC (Table 1). Epinephrine stimulation resulted in a negative FEXTM, −25.3% (SD 44.5), which indicated transpulmonary lactate production from mixed central venous pyruvate [12.8 μmol/min (SD 24.8)]. One-way ANOVA revealed a treatment effect on FEXTM (P < 0.05), and multiple comparisons with Tukey's HSD showed epinephrine stimulation significantly decreased FEXTM.

For both control and lactate load conditions, there appeared to be an association between FEXTM and net balance. However, when epinephrine stimulation was considered, any possibility of a correlation across conditions disappeared, because FEXC from net balance was positive [8.2% (SD 15.3)], whereas FEXTM was negative [−25.3% (SD 44.5)].

Tracer measured uptake was positive for Con and LC conditions, whereas epinephrine stimulation resulted in a negative tracer measured lactate uptake (Fig. 2). One-way ANOVA revealed a treatment effect on tracer measured lactate uptake (P < 0.05), and multiple comparisons by Tukey's HSD showed that epinephrine stimulation significantly decreased tracer measured lactate uptake.

Lactate-to-pyruvate ratios were significantly lower in the mixed central venous compared with arterial blood under all three conditions (P < 0.05). The transpulmonary change in [L]/[P] ratio provides further evidence for lactate formation in the pulmonary bed and averaged 39 (SD 19.7) to 77 (SD 36.9) in mixed central venous and arterial blood pools, respectively (Fig. 3).

The rate at which lactate appeared and disappeared from the blood (Ra and Rd) increased significantly during epinephrine stimulation and lactate load conditions (Table 1, P < 0.05). Lactate Ra during the Con condition [1.2 mg/min (SD 0.3)] was lower than that for pyruvate [2.0 ± 0.8mg/min (SD 0.8)] under similar conditions (16). The rate at which lactate was cleared from the blood, i.e., metabolic clearance rate, or MCR, showed very different patterns during the lactate load and epinephrine stimulation (Table 1). The Con condition produced the lowest MCR at 8.6 ml/min (SD 2.7), while Epi stimulation increased MCR slightly [12.9 ml/min (SD 4.1)], and the lactate load significantly increased lactate clearance to 21.4 ± 2.7 ml/min (SD 2.7) (P < 0.05). The lungs accounted for 5.1% of whole body lactate turnover during Con, and significantly more during LC and Epi stimulation conditions, 12.9% and 43%, respectively (Table 1, P < 0.05).

Total transpulmonary lactate release was not significantly different from zero for Con [−0.3 mg/min (SD 1.6)] and lactate load [−0.9 mg/min (SD 3.2)] conditions. However, total lactate release was negative during epinephrine stimulation [−6.9 mg/min (SD 10.2), P < 0.05], indicating pulmonary lactate production.

Mixed central venous isotopic equilibration ratios (i.e., IEL/IEP) averaged 2.6 (SD 1.9), 3.6 (SD 2.5), and 2.2 (SD 0.9) during Con, LC, and Epi conditions, respectively. One-way ANOVA did not reveal a treatment effect on mixed central venous isotopic equilibration. The arterial isotopic equilibration, calculated from arterial lactate IE and arterial pyruvate IE following the infusion of tracer lactate was higher during the lactate load [2.9 (SD 1.1)] and epinephrine stimulation [3.2 (SD 1.6)] compared with Con [1.3 (SD 0.9)]. One-way ANOVA revealed a treatment effect and multiple comparisons and Tukey's HSD analysis showed that the lactate load significantly increased the arterial IEL/IEP compared with Con (P < 0.05). The relationships among the isotopic enrichments of central venous and arterial lactate and pyruvate are shown for each condition in Fig. 4, A and B. There was a higher correlation between lactate and pyruvate IE in central venous (R² = 0.7) than arterial blood (R² = 0.4). As well, slopes and y-intercepts of the relationships were significantly different (P < 0.05). Clearly, epinephrine had profound effects on metabolism in the pulmonary parenchyma. Arterial blood MPEs are shown in Fig. 4C to highlight the effect of epinephrine on increasing lactate MPE as blood transits through the lungs.

Mixed central venous and arterial glucose concentrations.

Glucose concentration in the mixed central venous blood was lowest during the control condition (Table 2). One-way ANOVA revealed an effect of treatment on mixed venous glucose concentration [v̄] (P < 0.05), and multiple comparisons with Tukey's HSD showed that Epi significantly increased [v̄] glucose concentration (P < 0.05). Mean levels of arterial glucose concentration [a] were lowest during the control condition, but there was no evidence for an effect of treatment. Transpulmonary glucose concentration differences ([v̄] − [a]) were positive during all three conditions (Table 2). One-way ANOVA revealed a treatment effect on the transpulmonary glucose gradient (P < 0.05), and multiple comparisons with Tukey's HSD showed that Epi significantly increased the [v̄] − [a] difference above that during Con (P < 0.05). Glucose uptake was highest in the Epi condition due to the estimated elevation in cardiac output.