From the Department of Surgery, University ofFlorida College of Medicine, Gainesville, Florida, and the Health Sciences Division, Monsanto Company, St. Louis, ...
ANNALS OF SURGERY Vol. 218, No. 3, 350-363 C 1993 J. B. Lippincott Company
Arginine Transport in Human Liver Characterization and Effects of Nitric Oxide Synthase Inhibitors Yoshifumi Inoue, M.D., Barrie P. Bode, Ph.D., Dale J. Beck, Ph.D., Al P. Li, Ph.D., Kirby 1. Bland, M.D., and Wiley W. Souba, M.D., Sc.D.
From the Department of Surgery, University of Florida College of Medicine, Gainesville, Florida, and the Health Sciences Division, Monsanto Company, St. Louis, Missouri, and the Department of Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts
Objective Arginine transport was characterized and studied in human liver.
Summary Background Data Plasma arginine uptake may regulate hepatocyte intracellular availability and the subsequent biosynthesis of nitric oxide (NO), but little is known about arginine transport across the human hepatocyte plasma membrane.
Methods The authors characterized plasma membrane transport of 3[H]-L-arginine in hepatic plasma membrane vesicles (HPMVs) and in hepatocytes isolated and cultured from human liver biopsy specimens. They also studied the effects of the NO synthase inhibitors w-nitro-L-arginine methyl ester (L-NAME) and N-methylarginine (NMA) on arginine transport in HPMVs and in cultured cells.
Results Arginine transport was saturable, Na+-independent, temperature and pH sensitive, and was inhibited by the naturally occurring amino acids lysine, homoarginine, and ornithine (System y+ substrates). Arginine transport by both vesicles and cultured hepatocytes was significantly attenuated by NO synthase inhibitors, suggesting that the arginine transporter and the NO synthase enzyme may share a structurally similar arginine binding site. Dixon plot analysis showed the blockade to occur by competitive, rather than noncompetitive, inhibition. In vivo treatment of rats with lipopolysaccharide (LPS) resulted in a twofold stimulation of saturable arginine transport in the liver. This LPS-induced hepatic arginine transport activity was also inhibited by L-NAME. These data indicate that arginine transport by human hepatocytes is mediated primarily by the Na+-independent transport System y+.
Conclusions Besides inhibition of the NO synthase enzyme, the ability of arginine derivatives to block NO production may also be due to their ability to competitively inhibit arginine transport across the hepatocyte plasma membrane. The use of selective arginine derivatives that compete with arginine at the plasma membrane level may be a metabolic strategy that can be used to modulate the septic response.
350
Vol. 218 -No. 3
In recent years, the important role of the multifunctional bioregulatory molecule nitric oxide (NO) in regulating cellular metabolism has become appreciated. First identified in vascular endothelial cells,'"2 NO production has been demonstrated in a variety of cell types including human hepatocytes.3 It is now apparent that NO is derived exclusively from the N-terminal guanidino nitrogen of arginine, a dibasic, cationic, amino acid that also occupies a key position as an intermediate in the urea cycle. The source ofarginine to support NO synthesis in the liver is unclear, particularly in consideration of the low hepatic intracellular arginine pool.4 The intracellular concentration of arginine in hepatocytes is quite low (5 pM) compared to the plasma concentration (50 to 100 uM), suggesting that the blood may be an important source of arginine for the hepatocyte and that the activity of the plasma membrane carrier for arginine may control arginine availability. NO is intimately involved in a number ofbiologic systems and pathophysiologic states. Although the exact role that NO plays in the liver is unclear, it has been suggested that the liver may be a major source of circulating NO during septic states. NO production by liver parenchymal cells is markedly stimulated by bacteria, endotoxin, and cytokines and may be involved in the hepatocellular damage associated with sepsis.3 5'6 There is also growing evidence to suggest that, under certain circumstances, increased production of NO is responsible for the hemodynamic compromise seen during sepsis.7 Consequently, strategies to diminish NO production during sepsis have employed NO synthase inhibitors.8 Clinical reports suggest that the hypotension associated with severe infection can be reversed with arginine analogs9 such as w-nitro-L-arginine methyl ester (LNAME) and N-methylarginine (NMA), which decrease NO production by competitively binding to the NO synthase enzyme. The availability of the extracellular arginine pool to the hepatocyte may be controlled, under certain circumstances, by the inherent plasma membrane transport activity, which forms a kinetic barrier between extracellular arginine and the cytoplasmic space. As a consequence of the expanding metabolic roles played by arginine, particularly in reference to its interaction with sepsis and hepatic NO production, we have characterized, for the first time, arginine transport in the human liver using hepatic plasma membrane vesicles (HPMVs) and culPresented at the 11 3th Annual Scientific Session of the American Surgical Association, Baltimore, Maryland, April 1-3, 1993. Supported by NIH grant CA 45327 to Dr. Souba. Address reprint requests to Wiley W. Souba, M.D., Sc.D., Division of Surgical Oncology, Department of Surgery, Massachusetts General Hospital, Boston, MA 02114. Accepted for publication April 8, 1993.
Arginine Transport in Human Liver
351
tured hepatocytes. We also examined the ability of NO synthase inhibitors to block arginine transport across the cell membrane. Our results indicate that arginine transport by human hepatocytes is mediated primarily by the Na+-independent transport System y+. In addition, we have demonstrated that compounds that have traditionally been viewed as NO synthase inhibitors also block arginine uptake across the hepatocyte plasma membrane.
MATERIALS AND METHODS Studies were done in liver tissue from surgical patients and in laboratory rats. Human studies were approved by the Institutional Review Board (IRB) at the University of Florida Shands Teaching Hospital and by the Subcommittee for Clinical Investigation at the Gainesville Veterans Administration Hospital. Informed consent was obtained from each patient. Similar IRB approved protocols were performed in obtaining liver biopsy specimens from patients at Saint Louis University Hospital. Animal experiments were approved by the Committee for the Use and Care of Laboratory Animals at the University of Florida. For the HPMV studies, intraoperative liver biopsy specimens were obtained from six healthy adults undergoing elective abdominal operations. Biopsy specimens were obtained from patients without evidence of liver dysfunction. All patients consumed a regular diet up to the day before surgery and were made nulla per os (NPO) for 12 to 24 hours before operation. Similarly, biopsy specimens were obtained through the Saint Louis University School of Medicine, Department of Surgery, from a 38-year-old woman and a 56-year-old man, and hepatocytes were isolated as described below. Studies were also done in Sprague-Dawley rats (200 to 225 g, Harlan Sprague-Dawley, Indianapolis, IN). The rodents were maintained in the animal care facility in the laboratories for Surgical Metabolism/Nutrition. They were exposed to 12-hour light-dark cycles and given access to standard rat chow and water ad libitum. Animals were fasted overnight (access to water was maintained) and separated into two groups before study. Rats received an intraperitoneal injection ofeither Esche-
richia coli lipopolysaccharide (LPS) (7.5 mg/kg intraperitoneally for one dose, LPS, E. coli 01 27:B8; Sigma Chemical Co., St. Louis, MO) (n = 5) or saline (controls) (n = 5). HPMVs and isolated whole cells were prepared as described below.
Preparation of HPMVs HPMVs from human liver biopsy specimens were prepared using a method previously described.'0"' At oper-
352
Inoue and Others
ation, a 5- to 10-g wedge biopsy of liver was obtained, placed on ice, and immediately taken to the laboratory to begin processing for vesicles without prior freezing. All procedures were performed at ice-cold temperatures. Liver samples were perfused using a syringe with approximately 50 mL of ice-cold phosphate-buffered saline (PBS) (150 mM of NaCl, 10 mM of Na2HPO4; pH 7.4) until blanched free of gross blood. Samples were then minced with scissors in approximately 35 mL of sucroseEGTA buffer (SEB) (250 mM of sucrose, 1 mM of EGTA, 10 mM of HEPES; pH 7.5) and homogenized for 20 seconds on setting 6 using a Polytron (Kinematica, Switz., Brinkman Instruments, Westbury, NY), followed by an additional 15-second homogenization. The homogenate was diluted to 6% (w/v) with SEB and centrifuged at 150 X g for 2 minutes to remove gross particulate matter. The supernatant was then centrifuged at 1464 x g for 10 minutes and the resultant pellets were pooled, brought to approximately 60 mL with SEB, and resuspended with a Dounce homogenizer by 10 passes of a loose-fitting pestle followed by 4 passes with a tight-fitting pestle. The suspension was filtered through 8-ply gauze, added to 13.7 mL of Percoll (Pharmacia, Piscataway, NJ), and brought to a final volume of 115 mL with SEB (11.9% v/v final [Percoll]). The Percoll suspension was thoroughly mixed, transferred to three 50-mL, clear polycarbonate test tubes, and centrifuged at 34,540 x g for 30 minutes. The resultant plasma membrane bands were harvested with a 3-cc syringe, pooled, and diluted 1:6 (v/v) with sucrose-magnesium buffer (SMB) (250 mM of sucrose, 1 mM of MgC12, 10 mM of HEPES; pH 7.5). The plasma membrane suspension was centrifuged at 34,540 x g for 30 minutes and the membrane pellets resuspended and vesiculated in SMB by three passes through a 22-g needle to an approximate concentration of 3 to 6 mg of membrane protein/mL. Vesicles were aliquoted into Nunc cryotubes (InterMed, Denmark) and stored at -70 C until studied. For studies using rat liver, vesicles were prepared essentially as described above with minor modifications.'0 The entire rat liver (-8 g wet weight) was used to prepare vesicles, and a single homogenization step using the Dounce hand held homogenizer was employed before initial centrifugation in place of the dual Polytron and Dounce homogenization employed for human liver. Vesicle yield ranged from 2.0 to 3.0 mg and 3.0 to 5.0 mg of membrane protein/g of wet weight liver for human and rat liver, respectively. HPMV purity and integrity of selected samples was routinely evaluated in our laboratory by determining the relative enrichment of the plasma membrane enzyme marker 5'-nucleotidase,'2 the microsomal enzyme marker glucose-6-phosphatase,'3 and the mitochondrial marker NADPH:cytochrome c reductase,14 as well as by evaluation of membrane trans-
Ann. Surg. * September 1993
port and permeability characteristics. Inorganic phosphate was determined according to a previously described method.'5
Measurement of Arginine Transport by Membrane Vesicles Transport was assayed as previously described.'0" ''6 Human or rat HPMVs were thawed at room temperature and resuspended by three passes through a 22-g needle/ 1-cc syringe. Vesicles were diluted to a protein concentration of 2.5 to 3.5 mg/mL with SMB and kept on ice. Transport activity was determined in both the presence and absence of sodium. Final assay concentrations in the Na+-uptake buffer were 100 mM of NaCl, 1 mM of MgCl2, 10 mM of HEPES (pH 7.5), and varying concentrations of L-[3H]-arginine (Amersham, Arlington Heights, IL). The K+-uptake buffer was identical with the exception that NaCl was replaced by 100 mM of KCI. Transport was initiated by mixing 30-.uL vesicles (75 to 100,ug ofmembrane protein) with 30,ul of Na+- or K+-uptake buffer in 12 X 75-mm polystyrene tubes using an electronic timer/vortexer apparatus. At various time points the uptake of arginine was terminated by the addition of 1 mL of ice-cold PBS. The mixture was immediately passed over a 0.45-,um nitrocellulose filter under low pressure vacuum filtration. The filter was rapidly washed twice with 4 mL of ice-cold PBS/wash. Filters were then dissolved in 10 mL of Aquasol (New England Nuclear, Boston, MA) and trapped radioactivity was determined in a liquid scintillation counter (Beckman LS 7800, Beckman Scientific Instruments, Irvine, CA). Sodium-dependent transport was determined by subtracting uptake in the presence of potassium (Na+-independent uptake, triplicate or quadruplicate samples) from that observed in the presence of sodium (total uptake, quadruplicate samples). In order to evaluate transport kinetics, assays were undertaken as described above with the final initial extravesicular amino acid concentration varied from 50 ,M to 10 mM. Osmotic adjustments for the varying concentrations of amino acid were made with sucrose. Kinetic experiments were performed at the 10-second time point under initial rate conditions. Saturable Na+-independent kinetic parameters were examined by varying the concentration of [3H]arginine in various experiments over the range of 10 ,M to 20 mM and subtracting the nonsaturable component obtained in the presence of excess unlabeled arginine. Amino acid competition-inhibition transport assays were performed as described in the presence or absence of the designated amino acids at a concentration of 10 mM. Except as indicated, transport assays were performed at 25 C and blank values (no vesicles present) were subtracted out from uptakes in all ex-
Vol. 218 -No. 3
periments. Osmotic compensations for varying concentrations of arginine in the uptake reaction mixtures were made with sucrose where appropriate. All protein determinations were performed by a modified Lowry method'7 in order to obviate the confounding influence of interfering materials (i.e., lipid, sucrose) and all data were normalized to membrane protein. Transport velocities were calculated from radioactive counts and protein values using a computer software program and are expressed as pmol of arginine/mg of protein/unit time.
Isolation of Human Hepatocytes Freshly obtained biopsy specimens of human liver were placed in ice-cold buffer, immediately transported to the laboratory, and hepatocytes were isolated using a previously described method.'8 Briefly, biopsy specimens were placed on moistened sterile gauze in a Buchner funnel on top of a sterile waste beaker. Once major blood vessels on the open face of the biopsy specimen were identified, perfusion was initiated in a singlepass manner at 37 C and a flow rate of 20 mL/min through a peristaltic pump outfitted with silicon tubing and a 0.2 mL pipet tip. Calcium-free buffer, perfusion buffer I (PFI) (100 mM of NaCl, 5.5 mM of D-glucose, 5.5 mM of KCI, 0.4 mM of KH2PO4, 0.3 mM of Na2HPO4, 50 mM of HEPES [pH 7.4], 15 mM of NaHCO3, 0.5 mM of EGTA, 5 mU/mL of porcine insulin, 10 mg/L of phenol red, and 84 mg/L of gentamycin sulfate) was used to clear the biopsy specimen of gross blood. Upon observation of adequate blanching of the liver tissue and a clear effluent (typically after 100 mL of PFI), perfusion was stopped briefly and the funnel containing the biopsy specimen was transferred to the top of a sterile flask containing a 1 mg/mL collagenase (Boehringer-Mannheim Corp., Indianapolis, IN) solution in PFII (100 mM of NaCl, 5.5 mM of D-glucose, 5.5 mM of KCI, 0.4 mM of KH2PO4, 0.3 mM of Na2HPO4, 50 mM of HEPES [pH 7.6], 15 mM ofNaHCO3, 5 mM of CaCl2, 5 mU/mL of porcine insulin, 10 mg/L of phenol red, and 84 mg/L of gentamycin sulfate). Inlet tubing was inserted into the flask and perfusion was resumed in a recirculating manner. After the parenchyma within the liver capsule appeared digested and was soft to the touch, perfusion was ceased and the tissue was transferred to a sterile beaker containing 50 mL of ice-cold PFII. The capsule was cut with a scalpel and the tissue was gently teased to release the dissociated hepatocytes into the buffer. The cell suspension was transferred to a 50-mL tube and centrifuged at 50 X g for 5 minutes; the resulting cell pellet was resuspended in 40 mL of PFII and washed a total of three more times. Under these wash conditions, only the parenchymal cells sediment and the nonparenchymal cells remain in the discarded super-
Arginine Transport in Human Liver
353
natant. Viability of the final cell suspension was determined by trypan blue exclusion, and was greater than 85%.
Hepatocyte Culture The human hepatocyte suspension was diluted to a final concentration of 0.8 x 106 cells/mL in Dulbecco's Modified Eagles Medium/Hams Nutrient Mixture F12 (DME/F12) (Sigma Chemical Co., St. Louis, MO) supplemented with 4 mM of L-glutamine, 29 mM of NaHCO3, 84 mg/L of gentamycin sulfate, and 5% fetal bovine serum (Hyclone Laboratories, Logan, UT). Cell suspension (0.5 mL) was added to each well of a 24-well culture tray (Costar Corp., Cambridge, MA) previously coated with Type I acid-soluble calf skin collagen (Sigma Chemical Co.). The cells were maintained at 37 C in a humidified atmosphere of 5% C02/95% air. After allowing the cells to attach for 4 hours, the medium was aspirated and the cell monolayer was rinsed and replenished with serum-free DME/F1 2. The cells were maintained in serum-free medium until initiation of the transport assays.
Transport of 3H-Arginine by Hepatocyte Monolayers Transport of radioactively labeled arginine by cultured cells was assayed using a modification of the cluster tray method.2032 The uptake buffers are prepared from either Na+-free (choline-containing) or Na+-containing Krebs-Ringer phosphate buffer, cholKRP and NaKRP, respectively. For transport inhibition studies, osmotic compensation for excess amino acids was provided by an equimolar substitution of sucrose in the control transport buffer. Before the transport assays, the hepatocyte monolayers are rinsed twice with 2 mL/well of 37 C cholKRP to remove extracellular amino acids and sodium. Arginine transport activity was monitored by measuring the the uptake of 50 uM of L-arginine (5 ,Ci/ mL of 3H-L-arginine) for 1 minute at 37 C in both cholKRP and NaKRP. After discarding the transport buffer into a dish pan, the cell monolayer was washed three times with ice-cold cholKRP to remove extracellular radioactivity. After allowing the wells to air dry, the cells were solubilized by the addition of 0.2 mL of 0.2% sodium dodecyl sulfate (SDS) in 0.2N NaOH to each well. After a 30-minute incubation at room temperature, 0.1 0-mL aliquots of the cell extract were neutralized with 0.1 mL of 0.2N HCI and transferred to scintillation vials for determination of intracellular radioactivity in a Beckman model LS-6800 scintillation spectrophotometer (Beckman Scientific Instruments). The remaining 0.1 mL in each well was used to assay the protein content
354
Inoue and Others
using a modification of the Lowry procedure.32 Transport rates were calculated from protein values and specific activities using a BASIC computer program, and are expressed as pmol of arginine transported per mg protein per minute. Sodium-dependent arginine transport was determined by subtracting the transport velocity in cholKRP from that in NaKRP. Similarly, saturable arginine transport was determined by subtracting the transport velocity in the presence of 25 mM of unlabeled arginine from that in the absence of excess unlabeled argi-
nine.
Statistical Analysis In all vesicle studies, transport data were derived from two to four plasma membrane isolations from a minimum of three different livers from control and tumorbearing rats/experiment. Each individual assay was performed in triplicate. Kinetic parameters were calculated as described above. For cultured hepatocyte studies, each experiment was performed at least in quadruplicate. Data points represent means ± standard errors of the mean (SEM) of three to four determinations done in
triplicate. Values for transport rates (pmol/mg of protein/time) are expressed as mean ± SEM. Differences between means are considered significant at the p < 0.05 level. Data are analyzed using analysis of variance (with Dunnett's test when indicated) for comparison of multiple groups, or the Student's t test is used when two groups are compared.
RESULTS Integrity of Human and Rat HPMVs We have previously demonstrated the purity, functionality, and transportability of our rat HPMVs.'0 HPMVs exhibited time- and substrate concentration-dependent accumulation of amino acids in both the absence and presence of sodium, uptake overshoots in the presence of sodium, and uniformity of vesicle size. Electron microscopy revealed well-formed vesicles of plasma membrane origin with little contamination from endoplasmic reticulum and the absence of mitochondria.2I In order to evaluate the purity of HPMVs prepared from human liver biopsy specimens, selected marker enzymes from multiple vesicle preparations were measured to evaluate enrichment of vesicles in plasma membrane and impoverishment of endoplasmic reticulum. The specific activity (mean ± SEM) of the plasma enzyme marker 5'-nucleotidase was enriched 19-fold in vesicle preparations over the crude liver homogenates (Fig. 1). In addition to plasma membrane enrichment, the specific activities of
Ann. Surg. * September 1993
Cytochrome-C Reductase
0 Homogenate U Vesicles
Glucose-6-
Phosphatase e 5'-Nucleotidase
_^~~~~~~~~~C I
_lI
o1
,
/
I,
I
13
-,
is
,I
.I
I
17
;9
,
Relative Specific Activity (vs. Crude Homogenate ) Figure 1. Specific activities of marker enzymes in plasma membrane and endoplasmic reticulum. The specific activities of marker enzymes in human vesicles were compared to those in crude liver homogenates. Activities in vesicles are expressed relative to those in crude homogenates. Vesicles from control and endotoxin-treated rats were enriched 19-fold in the specific activity of the plasma membrane enzyme marker 5'-nucleotidase, respectively, over the crude liver homogenates. Preparation of vesicles resulted in a 25% to 35% impoverishment in the specific activity of the microsomal enzyme marker glucose-6-phosphatase and the mitochondrial marker NADPH: cytochrome c reductase relative to the crude homogenates.
the endoplasmic reticulum marker enzyme glucose-6phosphatase and the mitochondrial marker NADPH:cytochrome c reductase were diminished 25% to 35% compared to the crude liver homogenate. Collectively, these activities indicate highly enriched plasma membrane vesicles in the preparation.
Characterization of Arginine Transport in Human HPMVs Studies were initially performed to evaluate the transportability of human HPMVs (Fig. 2). To determine whether vesicle arginine uptake was due to actual transport of the substrate into an osmotically sensitive (intravesicular) space and not due to binding to the membrane surface, we examined the effect of increasing incubation medium osmolarity (by adjusting the sucrose concentration) on the uptake of 100 ,uM of arginine at equilibrium. Substrate binding was calculated by extrapolating the linear relationship between uptake and l/osmolarity to infinite osmolarity (Fig. 2A). From these data, the contribution nonspecific binding to total uptake was determined to be less than 10%, indicating transport into an osmotically sensitive space. We also evaluated the effect of temperature (O C, 22 C, 37 C) on the 10-second Na+-independent uptake of 100 ,uM of arginine (Fig. 2B). Transport was 10-fold higher at 22 C and at 37 C
Arginine Transport in Human Liver
Vol. 218 -No. 3
_ll @
*a
O
150
C6 10
4 0
1
Of
200
w
30
2
3
(lIOsmolarity) x 1000
B
*
*-r
355
linear for at least 60 seconds. In the presence of 1O mM of L-NAME, transport was diminished at all time points studied except at 30 and 60 minutes. The similarity of the 60-minute equilibrium value indicates that incubation with L-NAME did not change vesicle size. Figure 4 shows that arginine transport by human HPMVs increased steadily between pH values of 7 and 8, indicative of pH sensitivity in the physiological range. In addition, arginine transport exhibited trans-stimulation (Fig. 5). Vesicles were equilibrated at room temperature in the presence or absence of 40 mM of unlabeled arginine for 45 minutes. Extravesicular amino acid was removed by centrifugation and resuspension in fresh buffer. The uptake of 100 ,uM of 3[H]-arginine was determined in control unloaded vesicles and in vesicles preloaded with un-
aU -
ev cs
U, o 30
*Eb
10 > o20 v 0.c
2?C
3rc
W
1remperature
O,9:
Figure 2. Osmolarity plot and temperature sensitivity of arginine transport in human HPMVs. (A) Effect of changing the osmolarity of the incubation medium on arginine transport by human HPMVs. Mannitol was added to adjust the osmolarity of the incubation medium. The 30-minute transport of 3H-arginine was assayed as described in Methods. Each point represents the mean ± SEM of four determinations. When not shown, error bars lie within the symbol. (B) Effect of temperature on arginine transport.
0
20
40
60
8o
100
Sodium Concentration (mM)
IDw
0
IV E:
compared to 0 C, further indicating carrier-mediated transport into an osmotically active space. Amino acid transport can be broadly subdivided into Na+-dependent (requiring sodium ion for co-transport) and Na+-independent processes. While the membrane transport of arginine has largely been observed to occur through an Na+-independent pathway in rodent liver, we wished to examine the possibility of the involvement of a Na+-dependent transport component in human liver. The effects ofsodium on arginine transport and the time course of Na+-independent arginine transport are seen in Figure 3. As shown in Figure 3A, -80% of carrier-mediated arginine transport was accomplished by Na+-independent pathways. Because of the minor contribution of sodium-dependent transport to total carriermediated arginine transport, subsequent transport assays were performed in the absence of sodium ion. The time course of Na+-independent arginine transport is shown in Figure 3B. Transport in vesicles was rapid and
10
v.
-i;
0
Z a i _4
._.04 _.
CA
15
30
Time (min) Figure 3. Effect of sodium on and time course of arginine transport by human HPMVs. Human HPMVs were prepared from human liver biopsy specimens as described in Methods. (A) The 10-second uptake of 100 AM of [3H]-L-arginine was determined in either the presence of NaCI or KCI. Actual values for total uptake (mean ± SEM, n = 3 preparations) were 21.6 ± 1.1 pmol of arginine transported/mg of protein/10 sec when no sodium was present and 28.9 ± 1.5 at 100 mM of NaCI. *p < 0.05 versus KCI. (B) HPMVs were prepared and the vesicular uptake of 100 ,uM of [3H]-L-arginine was measured as a function of time. The saturable component of transport was determined in the presence and absence of 10 mM of L-NAME as described in Methods. Initial (60-second) rates of uptake are linear.
356
*S:, Q 2500
0;
< Et >g;E.>sG .....~~~~ .00|. ~ ~ ~ . . . .. :>~~~..:..a...:.....
..,:.,:..
':>.:.:>, >
8v ._s
1500
D
1000
s
10
5000
to ,. -E
t0OP0g.g g E%~4 0
= 2000
6
o
5.0
0.0
6.5
7
75
15.0
10.0
Arginine Concentration (mM)
8
pH Figure 4. Effect of pH on Na+-independent arginine transport in human HPMVs. The 1 0-second uptake of 100 AM of [3H]arginine was determined at the indicated pH values. Data represent the mean + SEM of three assays performed in quadruplicate on separate vesicle preparations. *p < 0.05 versus other values.
V
o s ut
10
v
g_ w
>c CL, 04
WC
labeled arginine. Vesicles preloaded with arginine exhibited -70% higher transport activity. Arginine transport kinetics were determined to further characterize Na+-independent arginine transport in human HPMVs (Fig. 6). When transport was assayed as a function of increasing extravesicular arginine concentrations, carrier-mediated arginine transport by vesicles was demonstrated to be saturable (Fig. 6A). Eadie-Hofstee linear transformation of the saturable arginine transport data suggested the presence of both a high affinity and a low affinity Na+-independent carrier, which was confirmed when the data were analyzed by nonlinear regression (Fig. 6B). The total Na+-independent velocity (V) is
*
;so%25
uz
4
04
Oe+O
2e+5
4e+5
6e+5
8e+5
le+6
Velocity/[Arginine] Figure 6. Na+-independent arginine transport as a function of extravesicular arginine concentration. (A) The 1 0-second uptake of [3H]arginine by human HPMVs was determined as described in Methods. (B) Eadie-Hofstee plot of saturable Na+-independent arginine transport. Eadie-Hofstee linear transformation of kinetic data was performed by plotting transport velocity as a function of velocity/[arginine]. A representative plot is shown. Regression analysis from multiple kinetic experiments indicated a high affinity and a low affinity component. Data points represent the mean + SEM of quadruplicate determinations. Error bars not apparent lie within the symbol. See text for actual kinetic values.
the sum of the velocities of the two systems (n = 17), as described mathematically below:
W
-.0 20
.X .
-
z.
Vtotai
15* -
20
Viow + Vh
Vmaxiow [arginine] Kmiow + [arginine]
Vmaxhigh [arginine] Kmhigh + [arginine] The curved line in Figure 6B is a plot of the empirical data, which is a composite of the two systems, indicated by the solid lines that are labeled as low affinity and high affinity components. Analysis of data from three separate kinetic studies revealed the low affinity carrier to have a Km = 5340 ± 532 uM and a Vmax = 999 ± 43 pmol/mg of protein/30 sec, while the high affinity carrier had a Km = 39 ± 9 ,uM and a Vmax = 37 ± 6 pmol/mg of protein/10 sec. Such a heterogeneous population of Na+-independent arginine transporters has been recently reported.8 +
@vat10'
::
=
o 5.5
00
Unloaded
Loaded
Figure 5. Trans-effects of preloading on saturable Na+-independent arginine transport. Vesicles were equilibrated at room temperature with either buffer and 40 mM of unlabeled arginine or buffer alone for 45 minutes. Extravesicular amino acid was removed by centrifugation and resuspended in fresh buffer. The uptake of 100 pM of [3H]-arginine was determined in control unloaded vesicles and in vesicles preloaded with unlabeled arginine. *p < 0.01 versus Unloaded.
Arginine Transport in Human Liver
Vol. 218 - No. 3
In other experiments, competition-inhibition assays performed using selected amino acids in order to further characterize arginine transport by human HPMVs and to define the specific systems that mediate transport. Two points are apparent from the inhibition data shown in Figure 7. First, alanine was ineffective at inhibiting arginine transport, indicating that System asc was not involved in carrier-mediated Na+-independent arginine uptake in human HPMVs. Second, the cationic amino acids ornithine, lysine, homoarginine, and arginine itself, were far more efficacious inhibitors of arginine transport than were the neutral amino acids glutamine and leucine. In these studies, the transport buffer contained 50 ,uM of arginine; therefore, transport through the high affinity carrier (System y+) rather than the physiologically insignificant low affinity system was evaluated. were
357
a.-
't uO-
._.
20
0
40
60
8o
Hours ~-b
500
B
-0 400- -
400
---
--O
Total Saturable
Na-dependent DiffusionI
300-
Characterization of Arginine Transport in Human Hepatocytes
200
In human hepatocytes, arginine uptake increased over time (Figs. 8A and 8B). Saturable arginine transport was 31 ± 7 and 57 ± 7 pmol/mg of protein/min at 8 hours compared to 78 ± 9 and 168 ± 16 pmol/mg of protein/ min at 72 hours in hepatocytes from the female and male, respectively. Although the absolute magnitude of the arginine transport activity in hepatocytes from the
Arginne
-r *
Homoarginine
Lysine MEW"
*
2~**
Oniffiine
Alanine Glutamine A
0
20
40
60
80
Hours Figure 8. Arginine transport activity by cultured human hepatocytes as a function of time in primary culture. Human hepatocytes were isolated from human liver biopsy specimens and cultured in DME/F12 as described in Methods. The culture medium was replaced every 24 hours. The 60-second uptake of 50 ,uM of [3H]-L-arginine was determined at the indicated times after culture initiation. (A) Hepatocytes were isolated from a liver biopsy specimen from a 38-year-old woman and placed in primary culture. Transport assays were performed in NaKRP in the absence or presence of 25 mM of unlabeled arginine, and saturable arginine transport was defined as the difference in the resulting transport velocities. (B) Hepatocytes were isolated from a liver biopsy specimen from a 56-year-old man and placed in primary culture. Transport assays were performed in cholKRP in the absence and presence of 25 mM of unlabeled arginine to determine Na+-independent saturable transport, and in NaKRP to determine Na+-dependent transport. Na+-dependent transport is defined as the difference in transport velocities in NaKRP and cholKRP. All data are presented as mean ± SEM (n = 4) and, where not indicated, the error bar lies within the symbol.
H~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-
Leucine
100
a
I
I
None I,.unu
ru
0
20
40
.
.
60
80
Arginine Uptake
(%
100
M
of Control)
Figure 7. Inhibitor profile of saturable Na+-independent arginine transport in human HPMVs. The 10-second uptake of 100 MM of [3H]arginine transport was determined in the absence of sodium and in the absence (control) or presence of the indicated L-isomer amino acids at a concentration of 10 mM. Osmotic compensation for inhibitor was made with sucrose. Saturable arginine uptake was obtained by subtracting the nonsaturable component (i.e., that portion of uptake not inhibited by 10 mM of arginine). Data represent mean ± SEM from four inhibition assays performed on three separate vesicle preparations. Final tyrosine concentration = 2.0 mM. *p < 0.05, **p < 0.01 versus control.
female and male patient were different, the fold induction from 8 to 72 hours was similar: 2.5-fold in the female cells and 2.9-fold in the male cells. The causes of this increase over time are unclear, but it may occur secondary to dedifferentiation of hepatocytes to a more fetal phenotype, and a loss of urea synthetic capacity (Beck, Bode, and Li, unpublished observations), leading to a reduction in endogenous arginine synthesis. Because the arginine transport assays in hepatocytes from the female were performed in only NaKRP ± excess unlabeled
358
Ann. Surg. - September 1993
Inoue and Others
arginine (i.e., total saturable uptake), the transport system(s) responsible for the increase in saturable arginine uptake were not determined. To further elucidate the nature of the transporters responsible for increased arginine uptake after 24 hours in culture, both sodium-dependent transport arginine transport (System B°'+) and sodium-independent saturable arginine transport (System y+) were monitored over 72 hours in hepatocytes from the male patient. Figure 8B illustrates that System y+ is responsible for nearly all of the carrier-mediated arginine transport during the first 18 hours of culture. A similar observation was made in hepatocytes from the female patient (data not shown). However, at 24 hours a sodium-dependent agency emerges that mediates arginine uptake, albeit at approximately 50% of the activity of System y+. The steady increase in saturable arginine transport activity observed between 24 and 72 hours in culture is due to parallel increases in both System y+ and the sodium-dependent system (System B°'+). Collectively, the data suggest that arginine transport in freshly isolated human hepatocytes is mediated by System y+, consistent with the HPMV data presented above. It is well established that freshly isolated cells more closely reflect the in vivo state than long-term cultures.33 Figure 9A shows the inhibitor profile of arginine transport in cultured human female hepatocytes. The results are similar to those observed in vesicles and demonstrate marked inhibition by arginine, lysine, and ornithine. A lesser degree of inhibition was observed in the presence of histidine and glutamine. Because the inhibition assays were performed in NaKRP, marked inhibition by the System N substrates asparagine, glutamine, and histidine was observed (31%, 42%, and 55%, respectively). In the presence of sodium, glutamine and other neutral amino acids have been shown to inhibit arginine transport,4 where it is postulated that the Na+ ion substitutes for the positive charge on substrate amino acids. To further elucidate the sodium-independent system(s) responsible for nearly all of the arginine transport early in culture, inhibitor studies were performed in cholKRP in the male hepatocytes (Fig. 9B). Again, consistent with System y+ mediation, the positively charged amino acids arginine, lysine, and ornithine exerted the greatest inhibitory effects at 5 mM. A small but significant amount of inhibition was observed in the presence of the neutral amino acids alanine, serine, and leucine (26%, 28%, and 22%, respectively), indicating that a small (-25%) amount of arginine transport may be mediated by System asc or b°'+. In contrast to the female hepatocyte inhibition profile, no significant inhibition was observed in the presence of the System N substrates asparagine or glutamine, probably due to lack of Na+ ions in the transport buffer. However, there was significant (35%) inhibition by histidine, even in the absence of sodium. This
Asparagine r; =;
Glutamine
,
Ornithine ;
I
A ***
Histidine
Lysine AArginine
* i * * i
A
None (Control)
20
0
I 80
60
40
* ! * 100
1220
% of Control Transport Velocity
I~~7=
,-
Glutamate Phenylalanine Leudne Serine Alanine w IMeAIB "; < Glycine c Asparagine w Glutamine Histidine X; = Ornithine Lysine Arginine
B
I ~~~~~~ I **
. .
**
I **
I I
I
**
I*** I
***
I.,i .,i.,I .,I.
None (Control)
20
0 %
40
60
80
.I 100
1220
of Control Transport Velocity
Figure 9. Inhibitor profile of saturable Na+-independent arginine transport in cultured human hepatocytes. Hepatocytes were isolated from the liver biopsy specimens from (A) a 38-year-old woman and (B) a 56-yearold man. After 23 hours (female) or 10 hours (male) in culture, the hepatocyte monolayers were rinsed twice with warm (37 C) cholKRP and the 60-second transport of 50 gM of L-arginine was determined as described in Methods in the absence or presence of the indicated unlabeled amino acids at 5 mM. Transport assays were performed in NaKRP in the female cells and in cholKRP in the male hepatocytes. Data are expressed as per cent of control arginine transport velocity, which was 53 ± 3 pmol of arginine/mg of protein/min in the female cells and 56 ± 3 pmol of arginine/ mg of protein/min in the male cells. All inhibitor studies were performed in quadruplicate. *p < 0.05, **p < 0.01, ***p < 0.005 versus control. may be due to the partially charged nature of histidine at neutral pH. Finally, model substrates from other transport systems such as glutamic acid (System X-AG), glycine (System gly), phenylalanine (System L), and MeAIB (System A) had no effect on arginine transport.
Effects of L-NAME and NMA on Arginine Transport in HPMVs and Cultured Cells Na+-independent arginine transport in human HPMVs in the presence of 10 mM of L-NAME was attenuated by 70% (Fig. OA). Similarly, arginine transport by cultured human hepatocytes was diminished by 60% in the presence of 5 mM of NMA (Fig. lOB). HPMVs prepared from the livers of rats treated with LPS
Arginine Transport in Human Liver
Vol. 218 d No. 3
359
DISCUSSION
A
HPMVs and isolated hepatocyte preparations provide
°O12
two convenient and complementary models for the study of amino acid transport in the liver. Vesicles are
I~~~~
spheres of plasma membranes prepared from whole orthat have been used by our laboratory0",' and by others22 to examine amino acid transport. They offer unique advantages over other approaches. Transport activity representative of that occurring in intact cells is adequately preserved in vesicles.22 Alterations in transport reflect the changes that occur in vivo and thus may be secondary to direct or indirect effects of mediators. By employing vesicles, it is possible to clearly discriminate the intrinsic level of membrane transport activity free from other confounding influences (e.g., metabolism). In contrast, employing isolated hepatocytes allows the investigator to examine transport in individual liver cell types, as opposed to liver plasma membrane vesicles, which contain membranes not only from parenchymal cells, but also nonparenchymal cells. Use of isolated hepatocytes also allows the study of regulation of transport by putative individual effectors in vitro. We23 and others20 have used cultured hepatocytes to investigate the regulation of amino acid transport. In the current studies, membrane vesicles and isolated cells were prepared immediately from fresh liver biopsy specimens obtained intraoperatively from healthy patients. The simigans
~ (-) L-NAME
(+)
L-NAME
W _ 60-
I=
.-
40'
20
_J C x 00 o
1.2
U) _
-
0 0.025mM Arginine 0.O75mM Arginine
O3
0.15 mM
4,-
Cont
Endotoxin
Figure 10. Effects of NO synthase inhibitors on arginine transport. Details are provided in the Methods section. (A) The 10-second transport of [3H]arginine by human HPMVs was assayed in the presence of 10 mM of L-NAME. (B) The 60-second transport of [3H]arginine by human female hepatocytes in the presence of 5 mM of NMA was measured. (C) The 10-second transport of [3H]arginine by rat HPMVs from control and LPStreated rats was assayed in the presence of 10 mM of L-NAME.
*~0.80~
~
~
0.40
0.0
exhibited an 80% increase in arginine transport activity that retained sensitivity to inhibition by L-NAME (Fig. OC). Dixon plot analysis (Fig. 11) of the effect of increasing concentrations of L-NAME on the transport of three different concentrations of arginine by human HPMVs revealed a linear relationship (intersection above the X-axis) consistent with competitive inhibition of the transporter by L-NAME with an apparent Ki of -0.8 mM.
Arguine
-2.0
T
r
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
L-NAME Concentraton (mM) Figure 11. Dixon plot of L-NAME inhibition of saturable Na+-independent arginine transport. The 10-second uptake of arginine by human HPMVs was determined at various [3H]arginine concentrations in the absence or presence of increasing concentrations of L-NAME. Nonsaturable uptake was determined in the presence of 10 mM of arginine and subtracted from total uptake to obtain the saturable component, the reciprocal of which was plotted as a function of inhibitor concentration. The apparent Ki value for L-NAME was 0.8 mM.
360
Inoue and Others
larity in the inhibition profile and initial uptake rates for arginine in both preparations confirms the reliability and functionality of both transport models. The transport of amino acids across the plasma membrane of mammalian cells is a dynamic process having the potential for controlling intracellular metabolism.9'"1"19-21,24 Distinct hepatic amino acid transport systems have been identified, each system presumably relating to a discrete, putative, membrane-bound transporter protein. Included among these agencies is System y+, an Na+-independent transport system that mediates arginine and related cationic amino acid transmembrane movement.4 In the current report, we have demonstrated that arginine transport by the human hepatocytes is mediated primarily by System y+. Assignment of this agency as the principal system responsible for accomplishing the transport of arginine into the hepatocyte was based on kinetic parameters (Fig. 6) and the inhibition profile present in both vesicles (Fig. 7) and cultured cells (Fig. 9). Inhibition of arginine transport was greatest in the presence of other cationic amino acids such as lysine, ornithine, and homoarginine. In contrast to previous studies in rat hepatic plasma membrane vesicles that indicated a Km for arginine transport of -0.8 mM,'6 our data in human liver vesicles indicate the presence of a higher affinity (-40 ,uM) arginine transporter. This suggests interspecies differences or the possibility of several different families of the y+ carrier. This is consistent with previous work that characterized arginine transport by endothelial cells,'9 which was shown to be mediated by System y+(Km 300 ,tM) and, to a lesser extent, by the Na+-independent carrier System B,'+. Further evidence that arginine transport in hepatocytes is mediated by System y+ is derived from previous studies that demonstrated that the tumor necrosis factor-stimulated, sodium-independent arginine transport activity was unaffected in the presence of neutral amino acids, while it was abrogated in the presence of the cationic amino acids. 16 In the current study, kinetic evaluation of Na+-independent arginine transport indicated the presence of a second low affinity carrier. The Km of this agency (-5 mM) is such that its contribution to total carrier-mediated arginine transport is negligible at physiologic circulating arginine concentrations (- 50 uM). Based on the small amount of inhibition observed in the presence of leucine in both intact cells and vesicles, this low affinity Na+-independent carrier is most likely System b°'+. Arginine is a cationic, dibasic, semi-essential amino acid with numerous roles in cellular metabolism.26 It serves as an intermediate in the urea cycle, a precursor for polyamine biosynthesis, and has both immunomodulatory and secretory functions. Recently, the N-terminal guanadino nitrogen in arginine has been identified as the exclusive precursor for the pleiotropic molecule
Ann. Surg. * September 1993
NO'. First identified in endothelial cells, NO2 is a potent, short-lived molecule that is produced by a variety of cell types, has an expanding role in cellular communication, and is thought to be a common final mediator of a host of physiologic and pathophysiologic responses. Although the constitutive calcium-dependent NO synthase isoenzyme present in vascular endothelium is absent in the liver,27 hepatocytes do possess an inducible calciumindependent NO synthase and NO production by liver parenchymal cells is markedly stimulated by endotoxin and cytokines5l8 Knowledge concerning the factors that regulate arginine transport to support NO biosynthesis and other arginine-dependent biosynthetic activity is scant. Intracellular arginine availability may be rate limiting for NO production, particularly during critical illness when serum arginine levels are decreased. Moreover, arginine availability to the hepatocyte intracellular space is markedly restricted under ordinary circumstances due to the low basal activity of System y'.4"16 Therefore, it is interesting to consider the implications of the increase in transport activity elicited by endotoxin on the metabolic pathways involving arginine in the hepatocyte. Certainly, the clinical implications ofthese findings regarding endotoxemia and sepsis are provocative in view of the fact that elevated circulating levels ofNO have been reported in septic patients.28 The intracellular fate of arginine is diverse but includes introduction into the urea cycle, polyamine and creatine biosynthesis, and generation of nitric oxide. NO is probably responsible for the hemodynamic compromise associated with septic shock,9 but Billiar et al. have suggested that NO production may exert a protective effect on sepsis-related hepatocellular damage.6 The source of arginine to support sepsis-induced NO synthesis in the liver is unclear, particularly when one considers the very low levels of hepatic intracellular arginine. Based on the results of the current work, one might postulate that circulating arginine is used, in part, for NO biosynthesis. However, the marked increase (10- to 20fold) in hepatic NO synthase gene expression observed in hepatocytes isolated from the livers of septic rats29 is far greater than the modest increase in arginine transport we observed in vesicles from LPS-treated rats. This suggests that the primary source of arginine for NO synthase biosynthesis may be derived from arginine produced in the urea cycle, not from arginine in the plasma. Nonetheless, is is intuitively apparent that the bloodstream may represent an important source of arginine for the hepatocyte. Of interest to us was that arginine transport by both vesicles and cultured hepatocytes was significantly attenuated by the NO synthase inhibitors L-NAME and NMA, suggesting that the arginine transporter and the NO synthase enzyme may share a structurally similar arginine binding site. One might argue that the ability of
Vol. 218 * No. 3
these compounds to block membrane transport in cultured cells was due to rapid transport and binding to the analog to the NO synthase active site, which acts as a signal to switch off the y+ carrier. However, this is unlikely to occur within the time frame we used to assay transport in cells. Moreover, the studies in vesicles, which examine only plasma membrane events independent of intracellular metabolism, indicate that L-NAME and NMA act directly on the transporter. Dixon plot analysis showed the blockade to occur by competitive, rather than noncompetitive, inhibition. The LPS-induced stimulation of hepatic arginine transport that was observed in vesicles prepared from endotoxemic rats was also inhibited by L-NAME, most likely due to inhibition of System y'.'6 The relatively high K, value of 0.8 mM for L-NAME versus the Km of 39 MM for arginine may reflect a lower binding efficiency of L-NAME by the arginine site on System y+. Despite the lower affinity of the carrier for the NO synthase inhibitor, it does competitively block the uptake of arginine into the cell (Fig. 1 1). Our data indicate that arginine transport by human hepatocytes is mediated primarily by the Na+-independent transport System y+. The mechanism by which arginine derivatives block NO production is due not only to their ability to block intracellular NO production, but also to their ability to competitively inhibit arginine transport across the hepatocyte plasma membrane. The arginine (y+) transporter and NO synthase enzyme may be closely associated in the plasma membrane and this may result in coupled regulation of both proteins. The use of selective arginine derivatives that compete with arginine at the plasma membrane level may be a nutritional/metabolic strategy that can be used to modulate the septic response. Further studies are necessary to better define the regulation of arginine metabolism as it relates to NO production in the liver during critical illness. The availability of the gene for the System y+ transporter30 as well as the cloning of NO synthase3' will allow study of this regulation at the molecular level.
Acknowledgement The authors thank Donald Kaminski, M.D., Department ofSurgery, St. Louis University, for providing some of the liver biopsy specimens.
References 1. Palmer RMJ, Ferrige AG, Moncada S. Vascular endothelial cells synthesize nitric oxide from L-arginine. Nature 1988; 333:664666. 2. Ignarro U, Buga GM, Wood KS, et al. Endothelium-derived relaxing factor produced and released from artery and vein is nitric oxide. Proc Natl Acad Sci U S A 1987; 84:9265-9269. 3. Curran RD, Billiar TR, Stuehr DJ, et al. Hepatocytes produced nitrogen oxides from L-arginine ion response to inflammatory products from Kupffer cells. J Exp Med 1989; 170:1769-1774.
Arginine Transport in Human Liver
361
4. White MF. The transport of cationic amino acids across the plasma membrane of mammalian cells. Biochim Biophys Acta 1985; 822:355-374. 5. Curran RD, Billiar TR, Stuehr DJ, et al. Multiple cytokines are required to induce hepatocyte nitric oxide production and inhibit total protein synthesis. Ann Surg 1990; 212:462-469. 6. Billiar TR, Curran RD, Harbrecht BG, et al. Modulation of nitrogen oxide synthesis in vivo: NG-monomethyl-L-arginine inhibits endotoxin-induced nitrite/nitrate biosynthesis while promoting hepatic damage. J Leukoc Biol 1990; 48:565-569. 7. Kilbourn RG, Griffith OW. Over production of nitric oxide in cytokine-mediated and septic shock. J Natl Cancer Inst 1992; 84:827-831. 8. Kilbourn RG, Jurban A, Gross SS, et al. Reversal of endotoxinmediated shock by NG-methyl-L-arginine, an inhibitor of nitric oxide synthesis. Biochem Biophys Res Commun 1990; 172:11321138. 9. Petros A, Bennet D, Vallance P. Effects of nitric oxide synthesis inhibitors on hypotension in patients with septic shock. Lancet 1991; 338:1557-1558. 10. Pacitti AJ, Inoue Y, Souba W. Tumor necrosis factor stimulates amino acid transport in plasma membrane vesicles from rat liver. J Clin Invest 1993; 91:474-483. 11. Pacitti AJ, Plumley DA, Inoue Y, et al. Growth hormone regulates amino acid transport in human and rat liver. Ann Surg 216:353362, 1992. 12. Moore DJ. Enzyme purification and related techniques. Methods Enzymol 1971; 22:130-148. 13. Swanson MA. Glucose-6-phosphatase from liver. Methods Enzymol 1955; 2:541-543. 14. Kilberg MS, Christensen HN. Electron transferring enzymes in the plasma membrane of the Ehrlich ascites tumor cell. Biochem J 1979; 18:1525-1530. 15. Fiske CH, Subbarow Y. The colorimetric determination of phosphorus. J Biol Chem 1925; 66:375-400. 16. Pacitti AJ, Copeland EM III, Souba WW. Stimulation of hepatocyte system y+-mediated L-arginine L-arginine transport by an inflammatory agent. Surgery 1992; 112:403-411. 17. Bensadoun A, Weinstein D. Assay of proteins in the presence of interfering materials. Anal Biochem 1976; 70:241-250. 18. Li AP, Roque MA, Beck DJ, et al. Isolation and culturing of hepatocytes from human livers. J Tissue Cult Methods 1992; 14:139146. 19. Greene B, Pacitti AJ, Souba WW. Characterization of L-arginine transport by pulmonary artery endothelial cells. Am J Physiol 1993; 264:L35 1-L356. 20. Bode BP, Kilberg MS. Amino acid dependent increase in hepatic system N activity is linked to cell swelling. J Biol Chem 1991; 257:345-348. 21. Pacitti AJ, Austgen TR, Souba WW. Adaptive regulation of alanine transport in hepatic membrane vesicles (HPMVs) from the endotoxin-treated rat. J Surg Res 1991; 51:46-53. 22. Schenerman MA, Kilberg MS. Maintenance of glucagon-stimulated system A amino acid transport activity in rat liver plasma membrane vesicles. Biochim Biophys Acta 1986; 856:428-436. 23. Dudrick PS, Copeland EM, Souba WW. Hepatocyte glutamine transport in advanced malignant disease. Surg Forum 1992; 43:13-15. 24. Souba WW, Pacitti AJ. How amino acids get into cells: mechanisms, models, menus, and mediators. JPEN J Parenter Enteral Nutr 1992; 16:569-579. 25. Van Winkle LI, Christensen HN, Campione AL. Na+-dependent transport of basic, zwitterionic, and bicyclic amino acids by a broad-scope system in mouse blastocysts. J Biol Chem 1985; 260:12118-12123.
362
Inoue and Others
26. Barbul A. Arginine: biochemistry, physiology, and therapeutic implications. JPEN J Parenter Enteral Nutr 1986; 10:227-238. 27. Knowles RG, Merrett M, Salter M, et al. Differential induction of brain, lung and liver nitric oxide synthase by endotoxin in the rat. Biochem J 1990; 270:833-836. 28. Ochoa JB, Udekwu AO, Billiar TR, et al. Nitrogen oxide levels in patients following trauma and during sepsis. Ann Surg 1991;
214:621-626. 29. Geller DA, DiSilvio M, Nussler AK, et al. Nitric oxide synthase gene expression is induced in hepatocytes in vivo during hepatic inflammation. J Surg Res (in press). 30. Kim JW, Closs El, Albritton LM, et al. Transport of cationic amino acids by the mouse ecotropic retrovirus receptor. Nature
1991; 352:725-728. 31. Lamas S, Marsden PA, Li GK, et al. Endothelial nitric oxide synthase: molecular cloning and characterization of a distinct constitutive enzyme isoform. Proc Natl Acad Sci U S A 1992; 89:6348-
6352. 32. Kilberg MS. Measurement of amino acid transport by hepatocytes in suspension or monolayer culture. Methods Enzymol 1989; 173:564-575. 33. Gurr JA, Potter VR. The significance of differences between fresh cell suspensions and fresh or maintained monolayers. Ann N Y
Acad Science 1980; 349:57-66.
Discussion DR. RICHARD L. SIMMONS (Pittsburgh, Pennsylvania): It's hard to know what to say about such a brilliantly clear study. Even the 30 pages of materials and methods of this paper read like poetry. Arginine is the substrate for urea synthesis and nitric oxide synthesis-the two waste products of nitrogen metabolismand therefore arginine is vitally important in nitrogen balance. The message of this paper is that access of arginine to the cell is regulated by the relative concentration of other cationic amino acids, including ones that might result from methylation of arginine in vivo (L-NAME). So my first question to Dr. Souba is: Is there any evidence that such regulation of arginine transport actually is operative in vivo under physiologic or pathophysiologic circumstances? Are there circumstances in which arginine is actually kept out of the cell so that arginine is kept from entering the urea cycle. If this were so, arginine could not contribute to the manufacture of polyamines that come from ornithine. The polyamines, in turn, are so important in immune function and other synthetic functions? Furthermore, if you block arginine's access to the cell by physiologic or pathophysiologic circumstances, nitric oxide is not made. Now, if nitric oxide isn't made, you can't regulate your blood pressure and many aspects of neurotransmission will be inhibited. In fact, life would be impossible because nitric oxide is essential for penile erection. This paper also makes it evident that we can pharmacologically or nutritionally manipulate access of arginine to the cell and decrease both urea and nitric oxide production. We can do so with normal nutritional components like lysine or we can do so with the synthetic ones like L-NAME. In this regard, it's interesting that certain arginine derivatives not only competitively inhibit nitric oxide synthase production directly, but inhibit access of arginine to the enzyme nitric
Ann. Surg. * September 1993
oxide synthase and therefore prevent arginine passage into the cell where it can gain access to this important enzyme. One of the most important things that Dr. Souba has shown is that you no longer can assume the nitric oxide synthase is the only thing being inhibited by L-NAME. Since many investigators have thought that L-NAME is a specific nitric oxide synthase inhibitor, it is important to find that transport is also affected. This was a wonderful paper. DR. THOMAS A. MILLER (Houston, Texas): President MacLean, Members and Guests, I would like to congratulate Dr. Souba and his colleagues on a very elegant series of experiments that have helped to characterize, in a fashion not previously known, the mechanism by which arginine is taken up by human hepatocytes and thereby regulates intracellular availability ofthis substance and subsequent nitric oxide synthesis. I have three questions that I would like to direct to Dr. Souba. The first one is as follows: Using both vesicles and cultured hepatocytes, you demonstrated that nitric oxide synthase inhibitors blocked arginine transport into the cell, suggesting that the arginine transporter and the nitric oxide synthase enzyme may share a structurally similar arginine binding site. Are we to conclude from these findings that nitric oxide synthase inhibitors work primarily by blocking transport of arginine and thereby nitric oxide synthesis rather than directly preventing such synthesis of nitric oxide intracellularly? Second, your studies with endotoxin in rats stimulated saturable arginine transport twofold and this effect could be inhibited by L-NAME. Again, are we to conclude that endotoxin's primary ability to promote nitric oxide production is through enhancement of arginine transport rather than any direct effects of endotoxin on nitric oxide synthesis intracellularly? Finally, you have proposed that enhanced hepatic nitric oxide synthesis in septic states is bad. Is that necessarily so? The reason I raise the question is that others have suggested the opposite, that nitric oxide production may be protective to the liver in septic states and may allow perfusion to be maintained so that it can carry out its important metabolic functions in the context of sepsis. Would you comment further in this regard and how you view this? Again, I would like to congratulate you and your colleagues on a splendid piece of work and for the opportunity to discuss it. DR. DONALD TRUNKEY (Portland, Oregon): Chip, I really enjoyed this paper. I guess I would like to follow up on Tom Miller's question. It seems to me that you've shown that you can block nitric oxide synthetase in liver, but is that truly good? Joe Fischer had shown earlier that nitric oxide is necessary for protein synthesis. The other question has to do with any studies you may have done on brain tissue. We know that nitric oxide may act like a Kevorkian cell because when there is injury or stroke nitric oxide synthetase is produced by these cells and nitric oxide kills the surrounding cells. Thus, you may want to block nitric oxide in the brain, but not in the liver. Is there any substance that you have identified that can do this?
