INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

AFTER AN INITIAL "ebb phase" lasting 24-36 h, major burn injury elicits a prolonged and pronounced period of hypermetabolism and catabolism ("flow phase") (36). This catabolic flow phase is characterized by increased glucose and oxidative metabolism, hepatic urea synthesis, gluconeogenesis, net nitrogen loss, muscle proteolysis, and subsequent efflux of amino acids, primarily glutamine and alanine (8). As the primary center of glucose and ammonia homeostasis in the body, the liver has been shown to display marked changes in metabolism after burn injury, including accelerated amino acid flux through gluconeogenic (9) and urea synthetic (45) pathways. The glutamine released from skeletal muscle after burn plays a key role in accelerated hepatic metabolism, serving as a substrate in both gluconeogenesis (33) and ureagenesis (16). As part of the well-characterized "intercellular glutamine cycle" (19) along the liver acinus, glutamine is hydrolyzed to glutamate and ammonia via hepatic glutaminase in the large population (>90%) of hepatocytes containing urea cycle enzymes. By virtue of its matrix location, glutaminase acts as an intramitochondrial ammonia amplification system, helping to efficiently drive the low-affinity rate-limiting urea cycle enzyme carbamoyl phosphate synthetase I (31). Therefore, flux through glutaminase plays a critical role after burn injury, because it is estimated that 80-90% of urinary nitrogen loss in burn patients occurs as urea (36). During accelerated metabolism, the transport of glutamine across the hepatocyte plasma membrane has been shown to constitute a rate-limiting step in its hydrolysis via glutaminase (22, 23, 29), an observation that served as the basis for these studies.

The uptake of glutamine across the plasma membrane of hepatocytes occurs primarily via a Na⁺-dependent amino acid transporter termed system N (6, 25) for its selectivity for glutamine, histidine, and asparagine only (amino acids bearing nitrogenous side chains). Previous studies from our laboratory (27) showed that burn injury stimulated glutamine transport rates, as measured in hepatic plasma membrane vesicles (HPMV) isolated from liver homogenates. Maximum transport stimulation was proportional to the size of the injury but was transient (peaking at 24 h but waning by 72 h) and mirrored the onset and rectification of burn-induced liver damage. These temporal effects at the plasma membrane level, however, were inconsistent with the advent of sustained hypermetabolism beginning 48 h after burn. This raised the possibility that other forms of transporter regulation may be necessary in vivo to support the markedly accelerated hepatic nitrogen metabolism resulting from this trauma. To address this issue, we chose to study isolated hepatocytes, which allow a more comprehensive assessment of burn-influenced transporter physiology, in which the effects of transporter influences lost in subcellular fractionation such as transmembrane electrochemical potentials, intracellular amino acid effects, and signal transduction pathways can be observed. In light of the importance of glutamine transport in supporting hepatic ureagenesis and gluconeogenesis, the studies presented here were undertaken to examine the impact of burn injury on system N activity in hepatocytes and to investigate potential signal transduction pathways that may participate in its regulation after this trauma.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

L-[³H]glutamine was obtained from DuPont NEN (Boston, MA). Collagenase was from Boehringer Mannheim (Indianapolis, IN), and chemicals, perfusion media [Calcium-free minimal essential medium for suspension cultures (S-MEM)], glucagon, and unlabeled amino acids were from Sigma (St. Louis, MO). Tissue culture medium (RPMI 1640) and SelectAmine kits were from GIBCO BRL Life Technologies (Gaithersburg, MD), tissue culture medium additives were from Biofluids (Rockville, MD), and supplies and chemicals for scintillation spectrophotometry were from Packard Instruments (Meriden, CT). Wortmannin and LY-294002 were obtained from BioMol (Plymouth Meeting, PA).

Burn model. Male Sprague-Dawley rats (150-200 g) were obtained from Charles River Laboratories (Wilmington, MA). Animals were housed in the Massachusetts General Hospital animal facility under controlled conditions of 12:12 h light-dark cycles and ad libitum access to chow and water. All experimental procedures were approved by the Massachusetts General Hospital Institutional Animal Care and Use Committee/Subcommittee on Research Animal Care, according to the guidelines in the Guide for the Care and Use of Laboratory Animals. Briefly, rats were anesthetized intraperitoneally (75 mg/kg ketamine and 1 mg/kg xylazine; Henry Schein, Port Washington, NY). The animals' backs were shaved, and full-thickness scald burns [~20% of total body surface area (TBSA)] were administered by dorsal immersion in 95°C water for 10 s. The full-thickness nature of the burn renders the injury anesthetic (9, 38). The major and minor axes of the resulting elliptical thermally injured areas were measured and used to calculate the size of the injury in square centimeters (0.7854 × major diameter × minor diameter). Burn size in percent TBSA was calculated via the empirical formula for total animal surface area: 11 × weight (grams)^0.631 (4). Control (sham) animals were immersed in water at room temperature. After scald, the animals were immediately fluid resuscitated by intraperitoneal injection with 12 ml of warm (37°C) lactated Ringer solution. Primary hepatocytes were obtained from these animals at 24, 48, and 72 h as well as 9 and 18 days after burn injury.

Hepatocyte and HPMV isolation. Primary hepatocytes were isolated by a modified collagenase digestion procedure described previously (11). To minimize nutritional influences on experimental results, all animals were subjected to an overnight fast before surgery the following morning. In brief, laparatomy was performed and the portal vein and inferior vena cava were cannulated with 16- and 14-gauge Teflon angiocatheters, respectively. The liver was cleared of blood by antegrade perfusion (20 ml/min) with 150 ml of warm (37°C) Ca²⁺-free S-MEM. Thereafter, the flow rate was reduced to 10 ml/min and the liver was digested with collagenase (0.5 mg/ml) in 100 ml of Ca²⁺-containing S-MEM. The digested liver was removed, and cells were released by gentle agitation into 50 ml of S-MEM. The suspension was filtered over a 50-mesh metal tissue sieve to remove gross particulate matter, and the filtrate was subjected to three centrifugations at 50 g for 2 min. Between each centrifugation, the supernatant was aspirated and the pellet subsequently resuspended in 30 ml of fresh S-MEM. After the final centrifugation, the pellet was resuspended in 20-30 ml of a chemically defined medium (RPCD) (44) composed of RPMI 1640 supplemented with 2 mM L-glutamine, 10 mM HEPES (pH 7.4), 3.7 mg/ml bovine serum albumin, 2.2 µg/ml insulin, 5 µg/ml transferrin, 5 ng/ml selenium, 500 nM phosphorylethanolamine/ethanolamine, 1 µM dexamethasone, 10 nM glucagon, trace elements (manganese, silicate, molybdenum, vanadium, nickel, and tin salts), 100 U/ml penicillin, and 100 µg/ml streptomycin. The hepatocytes were quantified and assessed for viability (typically >85%) by trypan blue exclusion on a hemacytometer and diluted in RPCD to a density of 5.4 × 10⁵ cells/ml. The cells (0.5 ml/well) were placed in 24-well culture plates (Costar, Cambridge, MA) previously coated with type I rat tail collagen (Collaborative Biomedical Products, Bedford, MA) and allowed to attach for 2 h in a humidified atmosphere of 5% CO₂-95% air at 37°C. In some experiments designed to test the role of amino acids in system N regulation, hepatocytes were incubated in amino acid-free RPCD (AAFRPCD) made with Selectamine Kits (GIBCO BRL Life Technologies), which permitted the formulation of RPCD with all its normal constituents except amino acids.

Amino acid transport measurement in primary rat hepatocytes. Primary rat hepatocyte amino acid transport was carried out via the cluster-tray method (12) as reported previously (11). After an initial two rinses with warm Na⁺-free Krebs-Ringer phosphate buffer (choline KRP), all transport measurements were carried out at 37°C in the presence of 50 µM L-glutamine and 4 µCi/ml L-[³H]glutamine in either choline KRP or Na⁺-containing KRP. Transport was terminated after 30 s by three rapid washes with 2 ml/well of ice-cold wash buffer [in mM: 119 NaCl, 25 Na₂HPO₄ (pH 7.5), 5.9 KCl, 0.5 CaCl₂ · 2H₂O, and 1.2 MgCl₂]. Intracellular radiolabeled glutamine was extracted with 0.2 ml/well of 0.2% SDS + 0.2 N NaOH; after 1 h, 0.1 ml of the lysate was neutralized with 10 µl of 2 N HCl and added to 1 ml of Microscint 20 for determination of trapped radioactivity by scintillation spectrophotometry (TopCount, Packard Instruments). The remaining cellular lysate was measured for protein content by the bicinchoninic acid procedure (Pierce Chemicals, Rockford, IL).

1

1

Plasma membrane vesicle amino acid transport assay. Initial-rate amino acid transport in HPMV was evaluated by a rapid mixing-filtration technique described previously (27) in the absence or presence of Na⁺. Uptake was initiated by mixing 20 µl of plasma membrane vesicles with 20 µl of Na⁺- or K⁺-containing transport buffer containing amino acid tracer in 1.5-ml centrifuge tubes using an electronic timer-vortexer apparatus. Final concentrations in the reaction mixture were 50 mM NaCl or KCl, 1 mM MgCl₂, 10 mM HEPES (pH 7.5), 50 µM L-glutamine, and 5 µCi/ml L-[³H]glutamine. After 10 s, amino acid uptake was terminated by addition of 1 ml ice-cold wash buffer followed by immediate low-pressure vacuum filtration of the mixture over a 0.45-µm nitrocellulose filter to separate intravesicular from extravesicular radiolabeled amino acid. The filter was rapidly washed twice with 2 ml of ice-cold wash buffer and subjected to extraction in the same manner as that used for the hepatocytes noted above. Na⁺-dependent transport rates were calculated similar to that for hepatocytes, are expressed as nanomoles of L-glutamine per milligram of protein per 10 s, and are the average ± SD of at least four separate determinations.

Glucagon and insulin measurements. The levels of glucagon and insulin were measured in plasma from portal vein blood in burned and sham-burned animals after 24, 48 and 72 h, using RIA kits (Linco Research, St. Charles, MO) according to the manufacturer's instructions.

Statistical analysis. Differences in specific measured parameters between experimental conditions were evaluated for statistical significance by a paired two-tailed t-test (Microsoft Excel) or by ANOVA with post hoc Fisher, Scheffé, and Dunnett tests (StatView Student, Abacus Concepts, Berkeley, CA) where multiple comparisons were performed. Relative differences were considered significant at P < 0.050.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of burn injury on hepatocyte glutamine transport rates. Glutamine transport rates were measured in hepatocytes isolated from animals at specific times after thermal or sham injury. Preliminary transport measurements were conducted in the presence and absence of unlabeled 5 mM -(methylamino)isobutyric acid (MeAIB; system A-specific substrate) to assess potential contributions of system A to glutamine uptake, which can occur under certain conditions (14). However, MeAIB failed to affect Na⁺-dependent glutamine uptake rates in both control and burn-influenced hepatocyte preparations, suggesting that all measured values were attributable to system N activity (data not shown). This is consistent with the lack of system A involvement in HPMV glutamine uptake reported in our previous study (27) using the small burn model employed here. As shown in Fig. 1, at 24 h after burn hepatic glutamine transport was slightly enhanced but not statistically significant compared with control hepatocytes (0.335 ± 0.038 vs. 0.400 ± 0.038 nmol · mg protein¹ · 30 s¹ in sham vs. burn, P > 0.050). This small difference in system N velocity in hepatocytes is consistent with the modest increase previously reported in HPMV after 20% TBSA burn at 24 h (27). After 48 h, however, hepatocyte system N activity was markedly enhanced after burn injury, where a nearly threefold increase was noted (0.267 ± 0.043 vs. 0.783 ± 0.076 nmol · mg protein¹ · 30 s¹ in sham vs. burn, P < 0.050). This burn-enhanced system N activity persisted after 72 h (0.308 ± 0.030 vs. 0.888 ± 0.097 nmol · mg protein¹ · 30 s¹ in sham vs. burn, P < 0.050).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Na+-dependent glutamine (GLN) transport rates in hepatocytes isolated from animals at specific times after burn injury. Hepatocytes were isolated from sham-burned and scald-injured rats at times listed after burn injury and placed in culture for 2-3 h before glutamine transport assay described in MATERIALS AND METHODS. Na+-dependent rates shown are average ± SD of 4 separate determinations in at least 2 hepatocyte preparations (n = 8). Inset: Na+-dependent glutamine transport rates (in nmol · mg1 protein · 10 s1) in hepatic plasma membrane vesicles (HPMV) isolated from hepatocyte preparations from control and burn-injured animals 48 h after burn injury. *P < 0.050 vs. sham-burned rates.

1

1

1

1

Mechanism of burn-dependent system N activation. We previously reported (27) that burn injury (20 ± 1% TBSA) elicited an increase in Na⁺-dependent glutamine transport activity of 35% after 24 h and 29% after 72 h compared with activities in HPMV from sham-burned animals; however, HPMV system N activity 48 h after burn had not been investigated. Given our current report of markedly stimulated hepatocyte glutamine uptake beginning 48 h after burn, we initially wanted to determine whether this could be attributable to increased system N activity in the plasma membrane per se. Therefore, glutamine transport activity in HPMV isolated from hepatocyte preparations 48 h after sham or burn injury was determined. As shown in the inset to Fig. 1, burn injury elicited an increase in plasma membrane glutamine transport activity of 25% after 48 h (0.032 ± 0.003 vs. 0.040 ± 0.004 nmol · mg protein¹ · 10 s¹ in sham vs. burn, P < 0.050). However, in the context of the threefold increase in hepatocyte system N activity elicited 48 h after thermal injury, the data indicated that a mechanism involving activation of existing transporters must account for the marked stimulation of glutamine uptake.

1

1

1

1

1

1

1

1

1

1

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Amino acid activation of system N in hepatocytes from sham-burned and scald-injured animals. Hepatocytes were isolated from sham-burned and scald-injured rats at times listed after burn injury and placed in culture for 2-3 h in RPCD medium before initial glutamine transport assay described in MATERIALS AND METHODS. Afterwards, medium was changed to amino acid-free RPCD (AAF) for 60 min, followed by a second measurement. Finally, medium was replenished with amino acid-containing RPCD (AAR) and glutamine transport was measured after an additional 60-min incubation. Na+-dependent rates shown are average ± SD of 4 separate determinations in at least 2 hepatocyte preparations (n = 8). *P < 0.050 vs. rates in hepatocytes from sham-burned animals.

1

1

1

1

1

1

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Suppression of amino acid-dependent system N activation by hyperosmotic culture medium. Hepatocytes were isolated from sham-burned or thermally injured animals after 9 or 18 days and placed in culture for 2 h before initial glutamine transport rate measurements (RPCD). Amino acid activation was assessed exactly as described in Fig. 2, except that on amino acid repletion with RPCD, hyperosmotic medium (made by addition of 300 mM sucrose) was included to suppress amino acid transport-induced cell swelling (AAR-hyperosmotic). Na+-dependent rates shown are average ± SD of 4 separate determinations in at least 2 hepatocyte preparations (n = 8). Sham-burned values depicted are from animals 9 days after thermal injury procedure but are not statistically different from those at 18 days after burn. *P < 0.050 vs. rates in hepatocytes from sham-burned animals; §P < 0.050 vs. RPCD; ¶ P < 0.050 vs. AAR.

1

1

1

1

Signal transduction pathways of amino acid-dependent system N activation. Changes in cellular hydration not only influence system N activity but also elicit global changes in hepatic amino acid, carbohydrate, and fatty acid metabolism (17). Although this relationship has been established for nearly ten years, the signal transduction mechanisms that underlie these effects remain poorly defined. Recently, Krause and colleagues (26) reported that wortmannin, a phosphatidylinositol-3-kinase (PI3K) inhibitor, attenuated the activation of hepatic glycogen synthase and acetyl-CoA carboxylase by cell swelling. To test whether a similar link existed between amino acid-dependent system N activation and PI3K, hepatocytes isolated 48 h after thermal injury were subjected to the amino acid activation assay in the absence or presence of either of two PI3K inhibitors, wortmannin (1) or LY-294002 (37). Cells previously maintained in AAFRPCD for 30 min were subjected to an additional 30-min pretreatment with or without 0.3 or 3.0 µM wortmannin or 50 µM LY-294002, followed by amino acid repletion with RPCD in the absence or presence of either of these PI3K inhibitors. Glutamine uptake rates were measured at 5 and 60 min thereafter. Pretreatment of both control and burn-influenced hepatocytes with wortmannin decreased basal system N rates by ~20% [0.116 ± 0.015 vs. 0.149 ± 0.020 nmol · mg protein¹ · 30 s¹ (P > 0.050) in control hepatocytes and 0.310 ± 0.021 vs. 0.386 ± 0.036 nmol · mg protein¹ · 30 s¹ (P < 0.050) in burn-influenced hepatocytes in presence and absence of wortmannin, respectively]. Likewise, pretreatment with LY-294002 decreased basal activity in burn-influenced hepatocytes by 30% [0.138 ± 0.014 vs. 0.196 ± 0.017 nmol · mg protein¹ · 30 s¹ (P < 0.050) in presence and absence of LY-294002, respectively]. As shown in Fig. 4, on repletion with amino acids (RPCD), both 0.3 and 3.0 µM wortmannin completely abolished the rapid 80% amino acid-dependent stimulation of system N in burn-influenced hepatocytes after 5 min (0.323 ± 0.028, 0.573 ± 0.057, and 0.279 ± 0.035 nmol · mg protein¹ · 30 s¹ in AAFRPCD, RPCD, and RPCD + 0.3 µM wortmannin, respectively; P > 0.050 between AAFRPCD and RPCD + wortmannin), as did 50 µM LY-294002 (0.241 ± 0.010, 0.490 ± 0.065, and 0.215 ± 0.012 nmol · mg protein¹ · 30 s¹ in AAFRPCD, RPCD, and RPCD + 50 µM LY-294002, respectively; P > 0.050 between AAFRPCD and RPCD + LY-294002). After 60 min, 0.3 µM wortmannin inhibited the induction by 53% (0.496 ± 0.018 vs. 0.686 ± 0.051 nmol · mg protein¹ · 30 s¹ in presence and absence of wortmannin, respectively, P < 0.050), and 3.0 µM wortmannin by 44% (0.708 ± 0.074 vs. 0.947 ± 0.100 nmol · mg protein¹ · 30 s¹ in presence and absence of wortmannin, respectively, P < 0.050). The effects of the more specific inhibitor LY-294002 also waned to a 43% suppression after 60 min (0.397 ± 0.024 vs. 0.524 ± 0.021 from 0.228 ± 0.025 nmol · mg protein¹ · 30 s¹ in presence and absence of LY-294002, respectively, P < 0.050). Collectively, the data suggest that the suppression of rapid system N activation by these two separate PI3K inhibitors is acutely effective but wanes with time.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Effect of phosphatidylinositol-3-kinase (PI3K) inhibition on amino acid-dependent activation of system N in burn-influenced hepatocytes. Hepatocytes were isolated from animals 48 h after thermal injury and placed in culture in RPCD for 2 h. Culture medium was changed to AAFRPCD (AAF) for 60 min, and "basal" glutamine transport rates were measured as described in Figs. 2 and 3. Thereafter, medium was replenished with amino acid-containing RPCD in absence (AAR) or presence (AAR + inhibitor) of PI3K inhibitors wortmannin (at 0.3 or 3.0 µM) or LY-294002 (at 50 µM); after 5 or 60 min, system N activity was again measured. AAR-treated cells received 0.1% DMSO as a vehicle control for inhibitor-treated cells, whereas cells treated with PI3K inhibitors were subjected to a 30-min pretreatment with these compounds before repletion of amino acids. Where not visible, SD lies within the bar. Na+-dependent rates shown are average ± SD of 4 separate determinations. *P < 0.050 vs. AAF; §P < 0.050 vs. AAR.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Time course of amino acid-dependent system N activation in absence or presence of wortmannin. Hepatocytes were isolated from sham-burned (A) and thermally injured (B) rats after 48 h and placed in culture in RPCD for 2 h. Medium was changed to AAFRPCD (AAF) for 60 min, and basal glutamine transport rates were measured at time zero. On amino acid repletion in absence (AAR) or presence (AAR + W) of 3 µM wortmannin, glutamine transport was measured at indicated times after medium change. AAR-treated cells received 0.1% DMSO as a vehicle control for wortmannin-treated cells, which were subjected to a 30-min pretreatment with this compound before repletion with amino acids, as in Fig. 4. Where not visible, SD lies within the symbol. Na+-dependent rates shown are average ± SD of 4 separate determinations. *P < 0.050 vs. AAF and AAR + W; §P < 0.050 vs. AAF.

Portal insulin and glucagon levels after burn. Finally, on the basis of the data it is clear that the unmasking of the latent system N activation pathway requires an in vivo "conditioning period" of >24 h. This is consistent with what is known about the latency of the amino acid-dependent regulatory pathway in general, where prior starvation of the donor animal for >24 h is required to maximize its function (7, 39). Although it is difficult to assess the quantitative contributions of individual hormones to this response in vivo, the evidence suggests that glucagon may play a role in eliciting this pathway, because both starvation and burn injury are known to alter the plasma insulin-to-glucagon ratios (43). To determine the effects of our model on these parameters, portal blood was obtained from burned and sham-burned animals after 24, 48, and 72 h. The levels of both pancreatic hormones were measured in the plasma by RIA, and the results are shown in Fig. 6. Portal insulin levels were decreased by ~50% in the burned animals compared with controls (1,187 ± 238, 846 ± 135, and 1,009 ± 209 pg/ml for control plasma vs. 490 ± 140, 501 ± 150, and 444 ± 180 pg/ml in burned animal plasma at 24, 48, and 72 h after burn, respectively, P < 0.050) on all 3 days. In contrast, portal glucagon levels in burned animals were elevated after 48 h (440 ± 50 pg/ml) vs. 24 h (216 ± 70 pg/ml, P < 0.050) and further increased after 72 h (572 ± 10 pg/ml, P < 0.050), compared with 48 h. The collective result of these burn-induced hormonal alterations was a progressive decrease in the insulin-to-glucagon ratio and thus an increasingly more "catabolic" profile over the first 72 h.

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Effect of burn injury on portal pancreatic hormone levels. At times indicated after thermal injury, portal blood was obtained from sham-treated or injured animals, and insulin (A) and glucagon (B) levels were measured by RIA as indicated in MATERIALS AND METHODS. *P < 0.050 vs. sham value; §P < 0.050 vs. 24-h burn value; ¶ P < 0.050 vs. 48-h burn value.

1

1

1

1

1

1

1

1

1

1

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Effect of glucagon on amino acid-dependent system N activation in hepatocytes from fed rats. Male Sprague-Dawley rats fed ad libitum received an intraperitoneal injection of 2 mg/kg glucagon or an equal volume of saline (control), and hepatocytes were isolated 4 h later. After culture in RPCD for 2 h, hepatocyte glutamine transport rates were measured. After a subsequent 60-min incubation in AAFRPCD (AAF), transport was again measured and medium was changed to amino acid-containing RPCD for an additional 60 min followed by a final glutamine transport assay. Na+-dependent rates shown are average ± SD of 4 separate determinations. *P < 0.050 vs. saline-injected RPCD; §P < 0.050 vs. RPCD; ¶ P < 0.050 vs. AAF.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The results presented here are significant for several reasons. First, the elucidation of the latent amino acid activation pathway for system N in burn-influenced hepatocytes resolves the paradoxical observation that only modest and transient increases in the corresponding liver plasma membrane activity occur (27) during the advent of the hypermetabolic phase (9) and the associated increase in hepatic amino acid extraction reported in vivo (41). Second, although amino acid activation of system N has been recognized for several years (7, 39), these studies provide the first report of a significant role for this pathway in regulating glutamine flux during a catabolic state. Finally, this work provides the first insights into signal transduction pathways linking amino acid-induced cell swelling to hepatic system N activation. Each of these points are discussed in the context of hepatic physiology during burn injury.

Among the various forms of trauma, severe burn injury gives rise to the most hypermetabolic state, resulting in appreciable increases in muscle proteolysis (35) and glutamine efflux (2) and a net negative nitrogen balance (8), the extent of which can be monitored by hepatic urea production (9, 10, 36). Both hepatic gluconeogenesis and ureagenesis rates are enhanced as a result of thermal trauma (9, 10), as is amino acid extraction by the hepatic bed (41). Given that hepatic system N activity constitutes a rate-limiting step in glutamine metabolism when intracellular utilization rates are enhanced (22, 29), our initial studies into the effect of burn injury on HPMV transport activity yielded interesting yet paradoxical results (27). In those studies, it was found that burn injury causes acute hepatic damage and stimulation of glutamine uptake, proportional to the size of the trauma. A larger (31% TBSA) injury elicited a 130% increase in HPMV glutamine transport activity after 24 h, 55% of which involved a system A component. After 72 h, the response waned to a 67% increase in HPMV glutamine uptake rates, 25% of which was system A mediated. This transient increase in HPMV Na⁺-dependent glutamine transport activity corresponded with the rectification of the hepatic damage, which was no longer evident after 72 h. In contrast, the small burn injury (20% TBSA) model utilized in the present studies elicited modest increases of 35 and 29% in HPMV glutamine uptake rates after 24 and 72 h, respectively, with no discernable system A component (27). The 25% increase in glutamine uptake in HPMV isolated 48 h after burn (Fig. 1, inset) and lack of visible system A involvement corroborate the results from the previous study. However, this modest increase in plasma membrane system N activity and the early and transient nature of transporter stimulation after a larger burn were inconsistent with the well-established onset of the hypermetabolic phase 48 h after thermal injury in both humans (36) and rodents (9). Demonstration of the latent amino acid-dependent system N activation pathway in hepatocytes beginning 48 h after burn injury (Fig. 2) resolves this issue, because the effects of this form of regulation are not retained in isolated HPMV (7).

With respect to the coordinated integration of hepatic physiology and systemic nitrogen metabolism after burn injury, the amino acid-dependent activation of system N may serve several purposes. Primarily, amino acid stimulation of glutamine uptake may serve to maintain adequate cytoplasmic pools of this amino acid when intracellular utilization rates for ureagenesis and gluconeogenesis are enhanced (9, 22, 29). In rats, plasma concentrations of glutamine are increased during the hypermetabolic phase of burn injury compared with normal values, reflecting an augmented output from muscle (24), whereas in humans with major burn injury, plasma glutamine levels are often depressed (10). In either case, a compensatory mechanism for enhanced glutamine uptake may be necessary, as hepatic levels of this amino acid are decreased during the hypermetabolic phase (24), underscoring its rapid metabolism. The affinity of system N for glutamine is ~0.6 mM to 1 mM in humans and rats, respectively (6, 25), suggesting that this transporter operates at or below its Michaelis constant in vivo, and is therefore responsive to changes in circulating glutamine levels. From the data presented in Figs. 2-4, it is apparent that amino acids elicit a more profound activation of system N activity in burn-influenced hepatocytes compared with controls and therefore enhance the inherent efficiency of hepatic glutamine transport during this catabolic state. Thus, even in the face of a slight diminution of glucogenic plasma amino acids during the hypermetabolic phase (10), hepatic glutamine uptake would remain sufficiently stimulated via this regulatory pathway.

Secondly, the transporter may itself elicit metabolic changes via the effects of its activity on cellular hydration. Among hepatic amino acid transporters, system N exhibits the highest activity (5, 7). Thus changes in glutamine transport rates exert considerable effects on the transmembrane flux of ions and metabolites. Glutamine has been known for some time to regulate the hepatic metabolism of not only amino acids but carbohydrates and fatty acids as well (3). It has since been demonstrated that its transport-dependent effects on hepatocellular hydration underlie its regulatory properties (21, 26, 40). During the hypermetabolic phase, amino acid-dependent activation of system N may serve to increase cellular hydration, which is known to enhance flux through the glutaminase and the urea cycle (20, 21). In summary, we conclude that enhanced system N activity may serve as an autostimulatory mechanism for driving the hypermetabolic flux through the urea cycle in the burn-influenced hepatocyte, via both increased substrate delivery and transport-dependent increases in hepatocellular hydration status.

The results presented here also provide the first insights into the signal transduction mechanisms that link amino acid-induced cell swelling (Fig. 3) to hepatic system N activation. Although the association among transporter activity, cell volume, and compensatory K⁺ movements during the regulatory volume decrease has been established (7), the signal transduction pathways that underlie these processes have remained poorly understood. Recently, Low and colleagues (28) demonstrated that the cell volume modulation of muscle glutamine transport system N^m activity was mediated by a PI3K-dependent pathway. Also, studies by Krause and colleagues (26) have shown that the well-established activation of glycogen synthase and acetyl CoA carboxylase by glutamine-induced hepatocyte swelling is PI3K dependent. We show here that the rapid amino acid-dependent system N activation in burn-influenced hepatocytes is mediated similarly, as evidenced by the complete inhibition of the response over the first 20 min by the PI3K inhibitors wortmannin and LY-294002 (Figs. 4 and 5). The direct activation of PI3K activity by amino acid-induced cell swelling should, however, be demonstrated in future studies to confirm the inhibitor-based results presented here and in other reports (26, 28). For example, it is unclear why this complete inhibition by both concentrations of wortmannin and LY-294002 wanes to ~50% after 30-60 min. This could reflect an inherent "leakiness" in xenobiotic PI3K inhibition with time (Fig. 4). Alternatively, this observation could be attributable to the induction of compensatory collateral pathways for transporter activation. In contrast to results in burn-influenced hepatocytes, there is only a marginal inhibitory effect of wortmannin on the amino acid induction in control hepatocytes (Fig. 5), when the 20% decrease in basal system N rates (in the absence of amino acids) by wortmannin pretreatment is taken into account (at time zero). Together, the results indicate that under normal conditions, PI3K partially regulates basal hepatic glutamine uptake rates, but this role is magnified and extended to amino acid stimulation beginning 48 h after burn injury. These findings also underscore the dynamic and complex regulation of this hepatic transport activity.

The role of specific systemic mediators in eliciting this latent PI3K-dependent regulatory pathway during the initial 48 h "catabolic priming period" remains unclear. However, several pieces of evidence suggest that glucagon may play a role. Both system N (13, 30) and glutaminase (18, 32) activities are accelerated by glucagon, which is known to increase in the plasma after burn injury (8, 34, 42). Our studies confirmed that the plasma insulin-to-glucagon ratios progressively decreased over the first 72 h after burn, attributable to an acute drop in plasma insulin levels accompanied by an incremental increase in glucagon (Fig. 6). The impact of this progressively more catabolic hormonal profile on the amino acid-dependent system N activation might have been predicted, as it is known that starvation of donor animals for >24 h before hepatocyte isolation is required to fully manifest this pathway (7, 39). This observation is underscored by the relatively modest stimulation observed in hepatocytes from overnight-fasted sham-burned animals shown in Figs. 1-3 and 5. The effects of exogenous glucagon on the ability of hepatocytes to respond to amino acids are shown in Fig. 7, and they corroborate results obtained in earlier studies (13). Here, we extend those studies and show that both system N transport rates as well as the ability of this carrier to respond to amino acids are enhanced by glucagon in these otherwise anabolic animals. As with any in vivo event, however, it is difficult to draw conclusions on the quantitative contribution of glucagon alone to the latent activation of this regulatory pathway. Given the complex and dynamic hormonal changes that occur after burn and the interactions between potential mediators (46), more detailed and focused in vivo studies will be required to specifically address this issue. Nonetheless, these data collectively suggest that glucagon may play a role in unmasking the latency of this pathway after burn injury.

In summary, the studies presented here provide the first mechanistic insights into the accelerated hepatic glutamine uptake observed after burn injury. Through its effects on both substrate delivery and cellular hydration, the system N carrier may play a pivotal role in governing accelerated ureagenesis and nitrogen loss during the hypermetabolic phase of burn injury. Enhanced activation of this glutamine transporter by extracellular amino acids, driven by cell swelling and mediated largely by PI3K, appear to be the collective result of a 24- to 48-h "hepatic conditioning period" after burn injury by a complex mixture of catabolic hormones, including glucagon. More detailed studies on the relationships among acute-phase cytokines, catabolic hormones, and hepatic PI3K isozyme expression and function will be required to better understand the complex pathophysiology of burn injury as it relates to liver glutamine and nitrogen metabolism.