|Home | About | Journals | Submit | Contact Us | Français|
The effect of cyclosporin A on the hepatic energy status and intracellular pH of the liver and its response to a fructose challenge has been investigated using in vivo phosphorus-31 nuclear magnetic resonance spectroscopy in dogs. Three experimental groups were studied: (a) control dogs (n = 5), (b) dogs 4 days after the creation of an end-to-side portacaval shunt (n = 5), and (c) dogs 4 days after portacaval shunt and continuous infusion of cyclosporin A (4 mg/kg/day) by way of the left portal vein (portacaval shunt plus cyclosporin A, n = 5). The phosphorus-31 nuclear magnetic resonance spectra were obtained at 81 MHz using a Bruker BIOSPEC II 4.7-tesla nuclear magnetic resonance system equipped with a 40-cm horizontal bore superconducting solenoid. The phosphomonoesters (p < 0.01), inorganic phosphate and ATP levels (p < 0.05) were decreased significantly in portacaval shunt–treated and in portacaval shunt-plus-cyclosporin A–treated dogs compared with unshunted control dogs. After a fructose challenge (750 mg/kg body wt, intravenously), fructose-1-phosphate metabolism was reduced in portacaval shunt–treated dogs compared with either the normal or portacaval shunt-plus-cyclosporin A–treated dogs (p < 0.05). Both portacaval shunt– and portacaval shunt-plus-cyclosporin A–treated dogs demonstrated a reduced decline in ATP levels after fructose infusion when compared with the controls (p < 0.05). Immediately after the fructose challenge, the intracellular pH decreased from 7.30 ± 0.03 to 7.00 ± 0.05 in all animals (p < 0.01) and then gradually returned to normal over 60 min. These data, obtained in vivo using phosphorus-31 nuclear magnetic resonance spectroscopy of the liver after a portacaval shunt, suggest that: (a) the energy status of the liver is reduced in dogs with a portacaval shunt compared with that of normal controls and (b) cyclosporin A treatment ameliorates the reduction in hepatic metabolism normally observed after a fructose challenge to the liver with a portacaval shunt.
The intrinsic hepatotoxicity of cyclosporin A (CsA) is well documented in the literature (1-10). Clinically, it can be manifested either as an increase in one or more liver injury parameters or as a cholestatic syndrome. Hepatic regeneration may be crucial during and immediately after transplantation, when the liver recovers from the injury experienced during the period of ischemic cold preservation and reperfusion. The acute effects of CsA on liver regeneration have been investigated recently by several different groups (11-16). These studies have shown that CsA may actually enhance the hepatocyte's regenerative response after partial hepatectomy.
In this study, the dog with portacaval shunt (PCS) (Eck's fistula) was used. PCS usually is associated with hepatic atrophy and low-grade hepatocyte hyperplasia (16-21). We have previously shown in this model that CsA prevents the atrophy and augments the hyperplasia (21). In this report, the effect of CsA treatment on the metabolic events caused by PCS is examined in vivo using phosphorus-31 nuclear magnetic resonance (31P-NMR) spectroscopy. Various in vivo studies using 31P-NMR spectroscopy (22-24) have reported on the intracellular pH (pHi), the hepatic content of ATP, sugar phosphates and inorganic phosphate (Pi) under basal conditions and after a physiological challenge with fructose. The only other NMR study relevant to hepatic regeneration or hepatocyte growth control was performed in vitro and used perchloric-acid extracts of tissue samples to calculate phosphate metabolites (25).
Fifteen female purebred Beagle dogs weighing between 8.5 kg and 15 kg were studied after care and conditioning at the University of Pittsburgh Animal Research Facility. Three groups of animals were studied: (a) normal dogs (control, n = 5); (b) animals subjected to PCS (n = 5); and (c) animals subjected to PCS and treated with a continuous infusion of CsA (4 mg/kg/day) into the left lobes of the liver through the left portal vein that had been cannulated at the time of the PCS procedure (PCS + CsA, n = 5). The basic design of the experiment using other test substances has been described previously (19, 20). Four days after the procedure and after general anesthesia induced with 20 mg/kg sodium pentobarbitone intravenously and ventilation with 2% halothane in 50% O2/50% N2O, femoral arterial and venous catheters were inserted for blood sampling. The proximal airway pressure, arterial pressure and blood gases were monitored continuously during the experiments using a Gould RS3600 recorder and a Radiometer ABL2 instrument (Radiometer America, Westlake, OH), respectively. Body temperature was maintained at 37° C with a heated water pad. Laparotomy was performed, and a surface coil was placed over the liver for NMR measurements. The left and right hepatic lobes were studied separately.
31P-NMR spectra were acquired on a Bruker BIOSPEC II 4.7-tesla NMR system equipped with a 40-cm horizontal bore superconducting solenoid. A 3.5-cm surface coil tuned to the resonance frequency of 31P nucleus (81 MHz at 4.7-tesla) was used both for 31P-NMR spectroscopy and for shimming on the water signal. The typical line width of the water resonance was 100 Hz. Generally, 31P-NMR spectra were acquired in 4-min blocks of 64 transients with an approximate 3-sec pulse delay (respiratory cycle) and 90 degree pulses. Gating was accomplished with a gating device that was armed by the respirator and triggered before each respiratory cycle. A 5-mm spherical bulb containing a 0.15 mol/L solution of methylenediphosphonic acid (MDPA) (23 ppm downfield from phosphocreatine [PCr] at 37° C) in D2O at pH 9 was placed in the center of the surface coil to serve as chemical shift and signal strength references for pHi and relative concentration measurements.
A calibration curve for the determination of pH from the 31P chemical shifts of Pi and fructose-1-phosphate (F-1-P) was obtained as follows. Liver specimens (4 gm wet wt) obtained from a normal dog were homogenized and diluted 1:4 (wt/vol) in physiological saline. The homogenate was centrifuged (2,500 rpm) at 4° C for 15 min, and the supernatant (12 ml) was transferred to a 30-mm NMR sample tube. A spherical glass bulb containing MDPA (0.15 mol/L in D2O, pH 9.0) was centered in the sample as an external standard. After the addition of 10 mmol/L F-1-P (Sigma Chemical Co., St. Louis, MO) and 10 mmol/L PCr (Sigma), a titration curve (from pH 4.24 to pH 9.45) was generated at 37° C. For each pH studied, the solution was positioned in the magnet, and 10-min acquisitions were accumulated using a 3.5-cm diameter solenoid coil, 90-degree pulses and a 5-sec interpulse delay. The 31P chemical shifts of Pi and F-1-P in the 31P--NMR spectra were referenced to the internal signal of PCr. Titration curves were obtained by plotting the 31P chemical shifts of Pi and F-1-P as a function of pH. A least-squares fit was performed to establish the titration parameters for each curve. The pHi was determined using the following equations (26):
for Pi, and
for F-1-P, where pKPi and pKF-1-P are 6.76 and 6.03 and σPi and σF-1-P are the observed values of the chemical shift in ppm of Pi and F-1-P, respectively, error ± 0.03 pH unit.
The content of phosphate metabolites (ATP, F-1-P, and Pi) in the liver was calculated by cutting out the corresponding peaks in the 31P-NMR spectra from the recording paper, weighing each peak and normalizing the data obtained using the MDPA peak area as a reference. This agreed satisfactorily with computer integration after baseline correction. After an injection of fructose (750 mg/kg body wt in 30% saline administered intravenously over 2 min), hepatic metabolism was monitored sequentially with a time resolution of 4 min.
The statistical analysis of the experimental data for intergroup variations was performed using a one-way ANOVA combined with the Fisher test. Student's t test for paired data was used to compare the change in parameters after fructose within each group, A p value < 0.05 was considered significant. All results are expressed as mean value ± S.E.M.
Three typical 81-MHz 31P-NMR spectra obtained in vivo from normal control, PCS-treated and PCS-plus-CsA–treated dogs are shown in Figure 1. The phosphomonoester (PM) peak (peak 2) represents the sum of signals from various sugar phosphates, glycolytic intermediates and AMP. The chemical shift of the Pi peak (peak 3) relative to the external standard (peak 1) permitted us to determine the pHi of the liver. This value was 7.30 ± 0.05 for all three groups of animals studied. The phosphodiester (PD) resonance (peak 4) is the sum of glycero-3-phosphocholine, glycerol phosphoethanolamine and other related compounds. The γ-ATP-phosphate and β-ADP-phosphate resonances (peak 5) occur at about −2.4 ppm upfield from PCr; the α-phosphates of both ATP and ADP (peak 6) occur at −7.5 ppm upfield from PCr, and the β-ATP resonance (peak 7) has a single peak at −16 ppm upfield from PCr. Because the right and left hepatic lobes yielded identical 31P-NMR spectra, only the left lobes were studied after the first two experiments.
The relative concentrations of the various phosphate metabolites in the liver as determined by 31P-NMR spectroscopy in vivo are shown for the three groups of animals studied in Figure 2. The PM (p < 0.01), Pi (p < 0.05) and ATP (p < 0.05) decreased in the presence of PCS-treated and in the PCS-plus-CsA–treated dogs with respect to the control. The (Pi/ATP) ratio that correlates inversely with the energy state of the liver did not significantly change in the three groups of dogs studied (Fig. 2).
After the administration of a fructose challenge (Fig. 3), major changes were observed in the 31P-NMR spectra of all three groups. An increase in the sugar phosphates, caused by the accumulation of F-1-P, and a reduction in the Pi and ATP content of the liver was evident immediately.
The time course of the F-1-P peak for the three groups of animals is shown in Figure 4. In the control dogs, the F-1-P resonance appeared within the first 4 min after the fructose infusion and reached a maximum value in the second 4 min. It then decreased rapidly as a result of the metabolism of the fructose load. This process was complete within 20 min, and the sugar phosphates region in the 31P-NMR spectrum returned to prefructose levels. In PCS-treated dogs, the metabolism of F-1-P was greatly reduced, as shown by a larger increase and a slower decline of the F-1-P resonance in the sugar phosphate region (p < 0.05). In contrast, the metabolism of the fructose load by the PCS-plus-CsA–treated dogs was normal and equivalent to control.
The ATP level (Fig. 5) decreased significantly (p < 0.01) to about 50% of the initial value 8 min after the fructose challenge. It began to recover at 12 min and reached a level of about 65% of the prefructose value at 40 min. The recovery of the hepatic ATP content was never complete even 2 hr after the fructose challenge, when the level is approximately 90% of the basal value (results not shown). The hepatic ATP level decreased less rapidly and to a smaller extent in the PCS-treated and PCS-plus-CsA–treated dogs (p < 0.05) after the fructose challenge. However, the subsequent recovery in ATP content paralleled that noted in the controls. The difference between shunted and control dogs was statistically significant (p < 0.05) (Fig. 5).
The Pi changes noted after fructose are shown in Table 1. The Pi was depressed early after the fructose challenge (p < 0.01) and returned toward normal values as the F-1-P peak disappeared. No statistically significant differences among groups were observed.
After the fructose challenge, the pHi fell from a value of approximately 7.3 to 7.0 (p < 0.01) in all three animal groups within 12 min and returned to a value close to baseline values thereafter (results not shown), confirming the findings of an earlier study performed in rat liver (28). Forty minutes after the fructose challenge, the pHi of the PCS-plus-CsA–treated dogs was not significantly different from that noted before the challenge. In contrast, the pHi of control and PCS-treated dogs was still reduced (p < 0.01 and p < 0.05, respectively).
The advantages of using 31P-NMR spectroscopy to study liver metabolism and function have been reviewed recently (23). In vivo 31P-NMR spectroscopy allows a nondestructive monitoring of tissue pH and a study of the levels of hepatic phosphorus-containing compounds associated with energy metabolism (23, 24, 27). In this study, this method was used to investigate the consequences of PCS and PCS in conjunction with cyclosporin treatment on hepatic function and basal energy status. The response of the liver to a fructose challenge was monitored by examining the changes in ATP, Pi, and F-1-P levels within the liver as a function of time (28).
The dog with a portacaval shunt was chosen as a well-established model for studying hepatic atrophy and contemporaneous hyperplasia (17-20). The typical injury pattern after PCS consists of a severe reduction in hepatocyte cell size (atrophy) plus organelle disruption and a moderate but persistent stimulation of cell renewal (hyperplasia). Insulin and cytosolic extract of the regenerating liver are able to restore the hepatocyte to normal size and to further stimulate hyperplasia (19, 20). A disadvantage of earlier studies was that the animals had to be killed, and in vitro assays of the metabolites of interest could be obtained only at selected times.
Interestingly, despite its recognized potential for intrinsic hepatotoxicity (1-10), CsA has been shown to have “hepatotrophic” effects after partial hepatectomy (11-16) or a PCS (21). Furthermore, CsA does not appear to either prevent or limit the hepatic regenerative response observed when a small-for-size liver is transplanted into a larger recipient (29) or when a reduced-size liver is used in pediatric recipients (30).
This study provides new data obtained by in vivo examination of the energy status and the metabolic responsiveness of the liver in an experimental situation that could have clinical relevance. The decline in the hepatic-phosphate metabolites observed after partial hepatectomy by in vitro studies (25) has been confirmed in the PCS model and in vivo. The effect of portal CsA infusion on these parameters was also determined. Using a fructose challenge as a test of metabolic responsiveness, the clearance of F-1-P in PCS dogs treated with CsA was found to be similar to that observed in the control and occurred considerably more rapidly than that observed in PCS dogs not receiving CsA treatment.
An apparent discrepancy arises when one considers the total balance of phosphates in the data. After fructose infusion, negative changes in ATP and Pi and positive changes in F-1-P did not seem to add up to zero in all single time points. In particular, when phosphorylation occurred (0 to 12 min after fructose), more phosphate is seen by 31P-NMR spectra in the PM region of the PCS-treated animals. In the same time period, the concomitant drop of ATP is larger only in the control group, and the decrease in Pi is not significantly different among the three groups. One possible explanation could lie in the slightly greater decrease in Pi in the PCS-treated dogs at most time points. Another possibility is that the low signal/noise ratio of the in vivo experiments and the proximity of the PM and Pi peaks in the 31P-NMR spectra limit an accurate measurement of relative changes in the two areas. Finally, the rapid changes in Pi could be underestimated in the 4-min time frame of our study. Under the present 31P-NMR protocol (3-sec repetition), the phosphate resonances are not likely to be significantly saturated (31), allowing us to relatively quantify the changes of the various phosphate metabolite concentration over time. These findings suggest that CsA preserves the capacity of the liver to metabolize a fructose load after a PCS treatment. This may reflect the fact that intracellular levels of the Pi that inhibits AMP degradation (32) are better preserved. Also, it is interesting that the changes for pHi noted in this study are consistent with data obtained using the perfused rat liver (28), suggesting that the fall in pHi after fructose is caused mainly by production of hydrogen ions accompanying the formation of lactate from fructose.
How these biochemical changes relate to prevention of atrophy by CsA and the augmentation of hyperplasia (21) is not clear. A simple explanation would be that the better response to a fructose load simply reflects the increased function under CsA of more healthful and more plentiful hepatocytes. However, a more sophisticated hypothesis may be forthcoming now that information exists concerning the action of drugs like CsA and FK 506 on cis-trans peptidyl-prolyl isomerase, an enzyme at the cytosolic binding sites for these agents. Peptidyl-prolyl isomerase inhibition or activation seems to have a wide-ranging effect on cell physiology, including responsibility for the so-called hepatotrophic mechanisms and control of carbohydrate metabolism (33).
Meanwhile, it must be remembered that cyclosporine is not a panacea for the liver. The drug can also be hepatotoxic even at therapeutic doses (1-10), impair hepatocyte transport processes (34, 35) or have an adverse effect on intracellular membranes (36, 37) after having gained entrance to the cells by passive diffusion (38). It is probable that CsA can be either hepatotrophic or injurious to the liver, depending on the dose.
We thank Drs. Susan R. Dowd, Alan P. Koretsky and Mark R. Busch (Department of Biological Sciences, Carnegie Mellon University) and Ms. Maryann Butowicz (Pittsburgh NMR Center for Biomedical Research) for their helpful discussions and expert technical assistance; and Ms. Cynthia Davis (Pittsburgh NMR Center for Biomedical Research) for typing the manuscript. We are also grateful to Drs. José Trejo-Bellido, Alejandra Imventarza, Michele Barone and John Prelich (Department of Surgery, University of Pittsburgh) for surgical and technical assistance; to Prof. Remo Naccarato (University of Padova, Italy) and Dr. G. Braga (Regione Veneto, Italy) for their continual encouragement and support. We are grateful to the Richard King Mellon Foundation, the Lucille P. Markey Charitable Trust, the Ralph M. Parsons Foundation and the Ben Franklin Partnership Program of the Commonwealth of Pennsylvania for providing financial support for the establishment of the Pittsburgh NMR Center for Biomedical Research.
This research was supported by research grants (DK-29961 to Thomas E. Starzl and HL-24525 to Chien Ho) from the National Institutes of Health, Bethesda, Maryland.
The Pittsburgh NMR Center for Biomedical Research is supported by a grant from the National Institutes of Health (RR-03631).
Lorenzo Rossaro was supported by grants from the C.N.R. (Italy 1988, No. 203.4.11), the Council of International Exchange of Scholars (U.S.A., 1986-1988 Fulbright Fellowship), and the Regione Veneto (Italy 1987, Piano Sanitario Finalizzato “II Trapianto di Fegato”).
A preliminary report of this work was presented at the Eighth Annual Meeting of the Society of Magnetic Resonance in Medicine, August 12-18, 1989, Amsterdam, The Netherlands; and at the Normal and Neoplastic Growth in Hepatology: Interface Between Basic and Clinical Science, Pugnochiuso-Foggia, Italy, June 21-24, 1989.