- Email: [email protected]
Interspecies differences in hepatic Ca2+-ATPase activity and the effect of cold preservation on porcine liver Ca2+-ATPase function
Interspecies Differences in Hepatic Ca2⫹-ATPase Activity and the Effect of Cold Preservation on Porcine Liver Ca2⫹-ATPase Function Piotr K. Janicki,*† Paul E. Wise,† Andrey E. Belous,† and C. Wright Pinson† The accumulation of intracellular calcium ([Ca2ⴙ]i) caused by ischemia-reperfusion during liver transplantation has been implicated as a factor leading to primary graft nonfunction. Plasma membrane (PM) and endoplasmic reticulum (ER) Ca2ⴙ-adenosinetriphosphatases (ATPases) are the primary transporters that maintain [Ca2ⴙ]i homeostasis in the liver. We hypothesized that the porcine liver is better than the rat liver as a model for the study of human liver Ca2ⴙ-ATPase activity. We also hypothesized that cold preservation would depress Ca2ⴙATPase activity in the porcine liver. Pig and rat livers were harvested, and human liver samples were obtained from surgical resection specimens. All were preserved with University of Wisconsin solution, and porcine livers were also preserved on ice for 2 to 18 hours. Ca2ⴙ-ATPase activity was measured after incubation with 45Ca2ⴙ and adenosine triphosphate in the presence of specific Ca2ⴙ-ATPase inhibitors. Porcine PM and ER Ca2ⴙ-ATPase activities were 0.47 ⴞ 0.03 and 1.57 ⴞ 0.10 nmol of Ca2ⴙ/mg of protein/ min, respectively. This was not significantly different from human liver, whereas rat liver was significantly greater at 2.60 ⴞ 0.03 and 9.2 ⴞ 0.9 nmol of Ca2ⴙ/mg of protein/min, respectively. We conclude that the Ca2ⴙATPase activity in the pig liver is equivalent to that of human liver, and thus, the pig liver is a better model than the rat liver. Cold preservation studies showed a significant decrease in porcine hepatic PM Ca2ⴙ-ATPase activity after 4 hours of storage and near-total inhibition after 12 hours. Porcine hepatic ER Ca2ⴙ-ATPase activity showed a 45% decrease in activity by 12 hours and a 69% decrease by 18 hours. We conclude that cold ischemia at clinically relevant times depresses PM Ca2ⴙ-ATPase more
From the Departments of *Anesthesiology and †Surgery, Division of Hepatobiliary Surgery and Liver Transplantation, Vanderbilt University Medical Center, Nashville, TN. Supported in part by Individual National Research Service Award no. 1 F32 DK09959-01 from the National Institutes of Health (P.E.W.). Presented in part at the Digestive Disease Week Meeting, May 21-24, 2000, Orlando, FL; the American Association for the Study of Liver Disease 50th Annual Meeting, November 5-9, 1999, Dallas, TX; and the American Association for the Study of Liver Disease 51st Annual Meeting, October 27-31, 2000, Dallas, TX. Address reprint requests to Piotr K. Janicki, MD, PhD, Department of Anesthesiology, VUMC, 504 Oxford House, 1313 21st Ave S, Nashville, TN 37232-4125. Telephone: 615-936-2800; FAX: 615-9362801; E-mail: [email protected] Copyright © 2001 by the American Association for the Study of Liver Diseases 1527-6465/01/0702-0018$35.00/0 doi:10.1053/jlts.2001.21459
132
than ER Ca2ⴙ-ATPase activity in pig liver homogenates. (Liver Transpl 2001;7:132-139.)
P
rimary graft nonfunction (PGNF) is an important and persistent problem in liver transplantation, occurring at the rate of up to 8%.1 Its cause remains unclear, but it appears to be related to ischemia-reperfusion injury, which is likely mediated at least in part by intracellular calcium ([Ca2⫹]i) accumulation in hepatic cells.2,3 These cells contain 2 families of Ca2⫹-adenosinetriphosphatases (ATPases; pumps) that maintain the optimal [Ca2⫹]i concentration in physiological conditions by transporting Ca2⫹ out of the cytosol externally (plasma membrane Ca2⫹-ATPase [PMCA]) or internally into the endoplasmic reticulum (sarcoendoplasmic Ca2⫹-ATPase [SERCA]).4 These 2 pumps are the primary means by which hepatic cells decrease cytosolic Ca2⫹ levels given that the Na⫹-Ca2⫹ exchanger does not have a significant role in the regulation of Ca2⫹ in the liver.5,6 PMCA and SERCA activity have been shown previously in rodent liver preparations,7,8 but the isoforms of PMCA differ somewhat between humans and rats.9 Conversely, peptide sequencing of some porcine PMCA and SERCA isoforms show that they have greater homology to human isoforms than the rat (Swiss Institute of Bioinformatics peptide sequence databases search using BLAST network service, August 2000). For this reason, we chose to test whether the porcine liver model is a closer equivalent to human liver than the rat model for the in vitro study of Ca2⫹ATPase activities. The first aim of this study is to show that our preparative techniques were sufficient to adequately separate plasma membrane (PM) and microsomal or endoplasmic reticulum (ER) fractions from porcine liver, as well as to show PMCA and SERCA activity in these fractions. The second aim is to compare the activity of PMCA and SERCA in membrane preparations from rat, pig, and human liver samples. After substantiating that the porcine model is closer to human liver than the rat, we used the porcine model to study the effects of cold preservation on the activity of hepatic PMCA and SERCA. The activity of all Ca2⫹-ATPases is temperature- and adenosine triphos-
Liver Transplantation, Vol 7, No 2 (February), 2001: pp 132-139
2⫹
Cold Ischemia Inhibits Hepatic Ca
-ATPases
133
Figure 1. List of putative factors involved in the modulation of activity of the Ca2ⴙ-ATPase pumps.
phate (ATP)-dependent10 and susceptible to modulation by a number of factors (Fig. 1). It has been shown that hypothermic storage produces disturbances in several cellular systems,11,12 including Ca2⫹-regulating systems.3,13 Therefore, the final aim of this study is to investigate whether cold preservation of clinically significant duration could produce changes in PMCA and SERCA activity in pig liver membrane fractions.
Materials and Methods Liver Procurement and Storage The Vanderbilt University Institutional Animal Care and Use Committee approved all protocols for animal use in this study. Experiments using the porcine model were performed on healthy cross-bred farm pigs (20 to 25 kg) aged between 9 and 12 weeks. Animals were housed in pens, kept on a 12:12hour light-dark cycle, fed once daily with pig chow, and given water ad libitum. The pigs were acclimatized at our facility for 7 to 10 days before the operation. After a 24-hour fast, the animals were anesthetized with ketamine (30 mg/kg intramuscularly [IM]), atropine (0.05 mg/kg IM), and acetylpromazine (0.1 mg/kg IM); the trachea was intubated; and the endotracheal tube was connected to the ventilator. Anesthesia was maintained with isoflurane, 0.5% to 1.25% (endtidal concentrations) in oxygen. The liver was harvested after midline celiotomy and used immediately for the preparation of membranes (see Preparation of Membrane Fractions). For the cold ischemia studies, aortic and portal cannulas were placed in situ, and the liver was perfused with ice-cold (4°C) University of Wisconsin (UW) lactobionate preservation solution (2 L). The livers were then stored in a plastic bag placed in ice slush for 2 to 18 hours before preparation of the membranes. For the rat model experiments, male Sprague-Dawley rats (250 to 300 g) were allowed food and water ad libitum until the morning of the experiment. Animals were anesthetized with isoflurane in oxygen and killed by decapitation. The liver was quickly harvested after midline celiotomy and processed as described (see Preparation of Membrane Fractions). Samples of fresh human liver from patients under isoflurane anesthesia undergoing partial liver resection for tumor were quickly obtained from the surgical pathology laboratory. Healthy human liver was macroscopically excised from the tu-
mor (used for histopathologic studies) and used for the preparation of the membrane fractions according to the method described next.
Preparation of Membrane Fractions PM vesicles from each species were prepared according to the method of Loten and Redshaw-Loten14 (1986) by fractionation on a self-forming Percoll (Sigma, St Louis, MO) gradient. The livers were homogenized in ice-cold medium containing 0.25 mol/L of sucrose and 10 mmol/L of HEPES/ potassium hydroxide (KOH), pH 7.4 (homogenization buffer), and centrifuged at 1,500g for 10 minutes at 4°C. The resulting pellets were resuspended in 75.0 mL of the homogenization buffer, thoroughly mixed with Percoll and sucrose (75.0 mL of pellet suspension: 10.1 mL of Percoll, 1.45 mL of 2 mol/L of sucrose), and centrifuged at 35,000g for 30 minutes at 4°C to produce a crude PM fraction in the middle of the gradient. This fraction was further purified by repeating the Percoll separation in the presence of 1.3 mmol/L of CaCl2 to separate DNA from PM. For the preparation of crude microsomal (ER) fractions, a variation of the technique of Lievremont et al15 (1994) was used. Supernatant from the initial centrifugation described previously was centrifuged at 8,000g for 30 minutes at 4°C. After separation from the pellets, this supernatant was centrifuged at 100,000g for 1 hour at 4°C. The final pellets were washed and resuspended in an ice-cold washing medium containing 0.25 mol/L of sucrose and 25 mmol/L of HEPES/ KOH, pH 7.4. Purity of the PM and ER fractions was assessed by measuring the selective activity of PM and ER enzyme markers in the initial liver homogenates and in subsequent fractions obtained during separation. Enzyme markers included 5⬘-nucleotidase (PM characterization) and glucose-6-phosphatase (ER characterization). Both enzymes were measured spectrophotometrically using commercially available kits (Sigma Diagnostic, St Louis, MO).
Measurement of Ca2ⴙ Transport Measurement of Ca2⫹ uptake by liver PM and ER vesicles was performed as described previously.16 The incubation mixture (4 mL) was composed of 30 mmol/L of imidazole-histidine (pH 6.8), 200 mmol/L of KCl, 5 mmol/L of MgCl2, 5 mmol/L of ATP, 5 mmol/L of sodium azide, 5 mmol/L of ammonium oxalate, and 20 mol/L of CaCl2 containing (final concentra-
134
Janicki et al
tion) 0.1:Ci/mL of 45CaCl2 (NEM Products, Boston, MA; specific activity, 30.7 mCi/mg). The reaction was started by adding aliquots of membrane proteins to each tube (final concentration, 0.07 mg/mL) at 37°C. Aliquots of 0.5 mL were removed after 5 minutes and collected on 25-mm cellulose nitrate filters (0.45 m pore size; Gelman Sci, Ann Arbor, MI) that had been prewashed with 2 mL of 0.25 mol/L of KCl and 10 mL of deionized water. After vesicle collection, the filters were washed with 2 mL of 0.25 mol/L of sucrose. After overnight drying, the filters were placed in vials containing CytoScint (ICN, Costa Mesa, CA), and 45Ca2⫹ activity was assessed in a Beckman LS3801 beta counter (Beckman, Fullerton, CA). Results are expressed as nanomoles of Ca2⫹ accumulated per milligram of membrane protein per minute of incubation time. Blank results were obtained in the presence of 0.2 mmol/L of orthovanadate, a universal inhibitor of all P-type pumps, in the incubation mixture. The specificity of the measured activity for the PM and ER membrane fractions was confirmed by the inhibition of Ca2⫹ uptake by eosin (2 mol/L), a specific inhibitor of PMCA,17 and thapsigargin (300 nmol/L), a specific inhibitor of SERCA.18 Protein content in the PM and ER fractions was measured by the method of Bradford,19 using bovine serum albumin as a standard. The focus of these experiments was to evaluate the activity of Ca2⫹-ATPases, defined by the uptake of 45Ca2⫹ by the membrane vesicles. This activity is believed to be determined by the function of ATPases (defined as the rate at which they convert ATP to adenosine diphosphate and inorganic phosphate (Pi) while moving Ca2⫹ across their respective cellular membranes), as well as the overall numbers of these proteins present in the membranes. Determination of the function and number of PMCAs and SERCAs present in the membrane fractions was not the focus of these experiments.
Statistics Results are expressed as mean ⫾ SEM. Statistical significance of differences was assessed using one-way ANOVA and multiple comparison of the means with the least significant difference test. Statistical significance was inferred for P less than .05.
Results Preparation and Characterization of PM and ER Fractions From Pig Liver Purification of porcine liver subcellular membrane fractions into PM and ER fractions is indicated by the specific activity of the 2 enzyme markers tested. The 5⬘-nucleotidase showed a significant 12-fold PM enrichment in the PM fraction, and the activity of glucose-6-phosphatase showed a significant 7-fold ER enrichment in the ER fraction compared with the crude liver homogenate (Table 1). Similar enrichment was obtained for rat and human liver samples (results not shown). There was limited contamination of the PM fraction with ER and vice versa based on the enzyme marker tests compared with the initial crude membrane homogenate. PM vesicles prepared from homogenized pig liver showed PMCA pumping activity of 0.47 ⫾ 0.03 nmol of Ca2⫹/mg of protein/min. ER vesicles prepared from pig liver cells showed comparatively greater SERCA pumping activity (1.57 ⫾ 0.09 nmol of Ca2⫹/mg of protein/min) in relation to PMCA (Table 2). We used several criteria based on the addition of specific inhibitors (orthovanadate, eosin, and thapsigargin) to show that the active Ca2⫹ transport in our PM and ER preparations shares properties consistent with those known for PMCA and SERCA from other sources. Radioactive calcium accumulation in vesicles was substantially inhibited by orthovanadate, with inhibitory constant (Ki) ⫽ 45 mol/L for PMCA and Ki ⫽ 50 mol/L for SERCA, in the range expected for these P-type pumps.20 Ca2⫹ uptake values of membrane vesicles in the presence of orthovanadate were therefore considered blanks. No Ca2⫹ uptake (100% inhibition) was shown in PM vesicles incubated in the presence of specific PMCA inhibitor (eosin), whereas only slight SERCA inhibi-
Table 1. Activity of PM Marker 5⬘-Nucleotidase and ER Marker Glucose-6-Phosphatase in Subcellular Fractions Prepared From Porcine Liver
Fraction Liver homogenate PMS ER membranes
5⬘-Nucleotidase (nmol/mg protein/min)
Glucose-6-Phosphatase (mol/mg protein/min)
0.10 ⫾ 0.04 1.2 ⫾ 0.1* 0.10 ⫾ 0.07
0.25 ⫾ 0.03 0.31 ⫾ 0.06 1.8 ⫾ 0.3*
NOTE. Values expressed as mean ⫾ SD. Three separate livers were used in each experiment (run in triplicate). * P ⬍ .05 by ANOVA and t-test compared with liver homogenate.
2⫹
Cold Ischemia Inhibits Hepatic Ca
135
-ATPases
Table 2. Interspecies Differences of PMCA and SERCA Activities
Species
PMCA Activity (nmoles Ca2⫹/mg protein/min)
SERCA Activity (nmoles Ca2⫹/mg protein/min)
Human Pig Rat
0.53 ⫾ 0.1 (n ⫽ 5) 0.47 ⫾ 0.03 (n ⫽ 14) 2.6 ⫾ 0.4 (n ⫽ 20)*
1.4 ⫾ 0.1 (n ⫽ 4) 1.57 ⫾ 0.1 (n ⫽ 10) 9.2 ⫾ 0.9* (n ⫽ 15)
NOTE. Values expressed as mean ⫾ SD. N indicates the number of livers used in each experiment (each run in triplicate). * P ⬍ .05 by ANOVA and t-test (rat versus human and pig).
tion (15% to 20%) was observed in the presence of eosin in the microsomal fraction. The specific SERCA inhibitor (thapsigargin) produced a 20% inhibition of Ca2⫹ pumping in PM vesicles and greater than 50% to 60% inhibition of Ca2⫹ pumping in the microsomal preparation (Fig. 2). The cross-inhibition is explained by the minimal contamination of each purified fraction by the other membrane type, seen in the enzyme marker results previously described (Table 1). Interspecies Differences in Hepatic Ca2ⴙ-ATPase Activity Results of the interspecies comparison of the intracellular Ca2⫹-ATPase activity in PM and ER are listed in Table 2. PM vesicles prepared from pig liver homogenates showed PMCA pumping activity of 0.47 ⫾ 0.03 nmol of Ca2⫹/mg of protein/min. Similar PMCA pumping activity was observed in the human liver homogenates (0.53 ⫾ 0.1 nmol of Ca2⫹/mg of protein/
min). Conversely, rat liver PMs were characterized by 5 times greater PMCA pumping activity (2.6 ⫾ 0.4 nmol of Ca2⫹/mg of protein/min; P ⬍ .05 by ANOVA followed by t-test). A similar difference was observed for the SERCA activities. SERCA activity in the rat liver homogenate ER fraction was 9.2 ⫾ 0.9 nmol of Ca2⫹/mg of protein/min, which is significantly greater (P ⬍ .05) than the SERCA activity in pig and human liver homogenates (1.4 ⫾ 0.1 and 1.57 ⫾ 0.10 nmol of Ca2⫹/mg of protein/min, respectively). Effect of Cold Preservation on PMCA Activity Results of these experiments are shown in Figure 3A. There was no statistically significant difference between the PMCA pumping activity in PM vesicles prepared immediately after procurement and after 2 hours of cold preservation in UW solution. Storage of the liver on ice for 4 and 6 hours produced partial time-dependent inhibition of PMCA activity (P ⬍ .05). Cold storage for 12 and 18 hours produced almost complete inhibition of PMCA activity compared with liver samples from the time of procurement and 2 hours after flushing with UW solution (P ⬍ .01). Effect of Cold Preservation on SERCA Activity
Figure 2. Activity of Ca2ⴙ-ATPase in PM and ER fractions prepared from whole-pig liver homogenates. Columns represent means for 3 separate livers. All measurements were performed in triplicate. Error bars indicate SEM. *P < .05, significant difference from controls by ANOVA followed by least significant difference test.
Results of these experiments are shown in Figure 3B. There was no statistically significant difference between the SERCA pumping activity in ER vesicles prepared immediately after procurement and after 6 hours of cold preservation in UW solution. However, there was a statistically significant partial inhibition of SERCA pumping activity in ER vesicles prepared after 12 and 18 hours of cold storage compared with the activity immediately after liver procurement (45% and 69% inhibition, respectively; both P ⬍ .05).
Discussion Previous work on the separation and characterization of hepatic PMCA and SERCA has primarily focused
136
Janicki et al
Figure 3. Effects of increasing length of cold preservation of pig liver on the activity of (A) PMCA and (B) SERCA prepared from whole-porcine liver homogenates. Columns represent means for 3 separate livers. All measurements were performed in triplicate. Blank values were obtained from PM and ER membranes incubated in the presence of 0.2 mmol/L of orthovanadate. Error bars indicate SEM. *P < .05, significant difference by ANOVA followed by least significant difference test versus 0-hour (no preservation) controls. **P < .01.
on the use of the rat model.8,21 Our laboratory uses the porcine liver model not only because of our familiarity with it,22,23 but it allows for greater tissue volumes for fresh membrane studies than the rat and more analogous operative situations to humans. Although other laboratories have separated and identified PM and ER from pig liver,24,25 this is the first time these membranes have been separated from pig liver for use in Ca2⫹-ATPase studies. Our data from the separation and characterization of the membrane fractions from the pig liver are equivalent to those from pig and rat liver membrane separations,8,21 suggesting that our separation methods are adequate. Based on the membrane marker enzyme studies, there is some contamination of our different porcine membrane fractions, but this is also consistent with the results of the other studies.8,21 The second aim of our study is to compare the activity of PMCA and SERCA in membrane preparations from rat, pig, and human livers by using 45Ca 2⫹ uptake and specific Ca 2⫹-ATPase inhibitors. Our results confirm that PM and ER fractions from the livers of these 3 species contain significant activity of both PMCA and SERCA. There also appears to be another Ca2⫹ entry mechanism in the microsomal membrane fraction that is not affected by the SERCA-selective inhibitor, thapsigargin, based on only 50% to 60% ER fraction Ca2⫹ uptake in the presence of this inhibitor. This uptake is most likely explained by contamination of the microsomal fraction with other membranes containing Ca2⫹ exchange mechanisms or pumps (likely from mitochondria). The use of mitochondrial Ca2⫹ exchange
inhibitors (e.g., ruthenium red) might clarify the source of this additional microsomal Ca2⫹ uptake. PMCA activity in our pig liver preparations was similar to that observed in membrane preparations from human liver and approximately 5-fold less than the activity observed in rat preparations. A similar trend was observed for the interspecies differences in SERCA activity. These data show for the first time the interspecies differences in the activity of the PMCA and SERCA molecules from liver preparations. This further supports the conclusion that the porcine hepatic model is likely better than the rat model as an analogy for the study of human liver Ca2⫹-ATPases. Although the pig is closer both immunologically and physiologically to humans than the rat,26 the substantial activity differences of PMCA and SERCA between rat and human and pig livers are somewhat unexpected given what is known about the different Ca2⫹-ATPase isoforms in these species. PMCA has been found in all mammalian cells and is encoded by 4 independent genes. Alternative splicing further increases the number of possible isoforms. In humans, PMCA1, PMCA2, and PMCA4 isoforms have been identified in the liver, whereas only PMCA1 was identified in rat liver by 1 group27 and both PMCA1 and PMCA2 were the only isoforms in the rat liver identified by another group.9 Pig liver PMCA isoforms have not been specifically sequenced, but the 1 peptide sequence of porcine PMCA1 from smooth muscle has shown 98% homology to the human PMCA1 and 93% homology to the analogous rat isoform.28 The SERCA family is composed of products from at least 3 genes, but they have not been as thoroughly studied. SERCA2B, an isoform
2⫹
Cold Ischemia Inhibits Hepatic Ca
of SERCA2, appears to be expressed universally in all cells, including liver cell membranes. One of the 2 SERCA isoforms sequenced in the 3 species we studied shows greater homology between humans and pigs than rats (Swiss Institute of Bioinformatics peptide sequence databases search using BLAST network service, August 2000). The homology between the identified isoforms of pig and human PMCA and SERCA likely explains their similar activities in our experiment. Conversely, the substantial differences in their activities compared with rat liver membranes are more likely explained by preferential expression of isoforms with greater function in rats or as yet undiscovered alternative-spliced rat Ca2⫹-ATPase isoforms. It is also possible that the difference in the method of obtaining the human liver from resection specimens might alter their Ca2⫹ATPase activities. The role of Ca2⫹-ATPases in the maintenance of normal hepatic [Ca2⫹]i levels after ischemia-reperfusion appears to be dependent on maintaining their function after transplantation. Experimental evidence supports the presence of several mechanisms in liver cells that can be responsible for increased [Ca2⫹]i levels after ischemia-reperfusion.3 We postulated that liver ischemia during cold preservation causes a moderate increase in [Ca2⫹]i level, in part caused by emptying [Ca2⫹]i stores and decreased calcium clearance by PMCA and SERCA. Similar changes have also been observed in each of the major hepatic cell types during ischemia.29 When the liver graft is subsequently transplanted into a Ca2⫹-rich environment, there is a massive inflow
Figure 4. Hypothetical series of events producing intracellular Ca2ⴙ-overload and liver dysfunction and/or PGNF after liver preservation and reperfusion. *CCE, capacitative entry.
-ATPases
137
of calcium into the liver cells through the Ca2⫹ release– activated Ca2⫹ channels, a process known as capacitative calcium entry (CCE).30 If Ca2⫹-ATPases are significantly inhibited by cold preservation of the organ, they may be unable to compensate for the postreperfusion cellular calcium inflow. Subsequent cellular dysfunction and death can occur because of increased [Ca2⫹]i levels, causing uncontrolled activation of proteolytic activity31 and mitochondrial calcium overload, which can lead to mitochondrial damage and derangement of cellular energy production, resultant cell death, and tissue necrosis.32 There is evidence that ischemiareperfusion injury of the liver leads not only to cellular necrosis, but also apoptosis.33,34 If cellular injury is significant enough, graft dysfunction or PGNF may occur after transplantation (Fig. 4). If this hypothetical chain of events is correct, the major triggering points for Ca2⫹ overload during the reperfusion phase of transplantation are emptying of the [Ca2⫹]i stores during cold preservation, causing activation of CCE mechanisms, and the inhibition of cellular Ca2⫹ pumping. Therefore, approaches to maintain [Ca2⫹]i homeostasis during storage might improve organ preservation and decrease the intensity of reperfusion injury in liver transplantation. As the third aim of our study, we therefore wished to establish whether PMCA and SERCA activities are affected by cold preservation. Neither pump activity seemed to be significantly decreased after short-term cold preservation in UW solution (⬍12 hours). However, prolonged cold storage (ⱖ12 hours) of the porcine liver produced almost complete inhibition of the
138
Janicki et al
PMCA pumping mechanism. Similarly, SERCA activity showed a significant decrease after 12 hours of cold storage, but this inhibition was less prominent compared with the decrease in PMCA activity after the same period. This suggests that the PMCA pump is more sensitive to cold ischemia than the SERCA pump. That these pumps show significant inhibition after 12 hours correlates with the increasing rate of PGNF and graft dysfunction after 12 hours or more of cold ischemia in human liver transplantation.35 It is unclear at the present time whether the inhibition of the Ca2⫹ATPases observed in the cold-preserved liver in our experiment is permanent or reversible by reimplantation of the liver graft. It is also unclear whether the activity is decreased because the numbers of functional proteins are decreased or the function of those proteins is diminished. Because these studies were conducted on whole-liver homogenates consisting of a mixture of cells, it is unclear whether PMCA and SERCA activity changes are similar for all hepatic cell types. Different hepatic cells have different sensitivities to cold and warm ischemia and reperfusion,29 and their ability to regulate [Ca2⫹]i after an ischemic insult also differs.36 Whether these same liver cells have different sensitivities of their Ca2⫹ATPases after ischemia-reperfusion has not been established. It will be necessary to confirm our studies in membranes from isolated and purified liver cells to associate the PMCA and/or SERCA activity with the specific types of liver cells. In conclusion, the present study shows that PM and ER membranes can be separated from porcine, human, and rat liver homogenates, and the activity of the PMCA and SERCA pumps can be measured from these livers. We also found that the Ca2⫹-ATPase activities in the porcine liver homogenate are closer to the human liver than is the rat, showing that the pig is a better model for the study of human liver Ca2⫹-ATPases in vitro. Finally, we showed that ischemia associated with the cold storage of the porcine liver produces inhibition of the cellular Ca2⫹-ATPase pumps, PMCA more than SERCA. This inhibition may contribute to the disturbances in the [Ca2⫹]i clearance mechanisms after reperfusion that lead to liver graft dysfunction and PGNF after liver transplantation.
References 1. Porte RJ, Ploeg RJ, Hansen B, van Bockel JH, Thorogood J, Persijn GG, et al. Long-term graft survival after liver transplantation in the UW era: Late effects of cold ischemia and primary dysfunction. The European Multicentre Study Group. Transpl Int 1998;11(supp 1):S164-S167.
2. Tredger JM. Ischaemia-reperfusion injury in the liver: Treatment in theory and in practice. Biofactors 1998;8:161-164. 3. Clavien PA, Harvey PRC, Strasberg SM. Preservation and reperfusion injuries in liver allografts: An overview and synthesis of current studies. Transplantation 1992;53:957-978. 4. Kuo TH, Liu B-F, Yu Y, Wuytack F, Raeymaekers L, Tsang W. Co-ordinated regulation of the plasma membrane calcium pump and the sarco(endo)plasmic reticular calcium pump gene expression by Ca2⫹. Cell Calcium 1997;21:399-408. 5. Lidofsky SD, Xie M-H, Scharschmidt BF. Na2⫹-Ca2⫹ exchange in cultured rat hepatocytes: Evidence against a role in cytosolic Ca2⫹ regulation or signaling. Am J Physiol 1990;259:G56-G61. 6. Kraus-Friedmann N. Calcium sequestration in the liver. Cell Calcium 1990;11:625-640. 7. Kessler F, Bennardini F, Bachs O, Serratosa J, James P, Caride AJ, et al. Partial purification and characterization of the Ca2⫹pumping ATPase of the liver plasma membrane. J Biol Chem 1990;265:16012-16019. 8. Dawson AP, Fulton DV. Some properties of the Ca 2⫹-stimulated ATPase of a rat liver microsomal fraction. Biochem J 1983; 210:405-410. 9. Howard A, Barley NF, Legon S, Walters JR. Plasma-membrane calcium-pump isoforms in human and rat liver. Biochem J 1994; 303:275-279. 10. Carafoli E, Brini M. Calcium pumps: Structural basis for and mechanism of calcium transmembrane transport. Curr Opin Cell Biol 2000;4:152-161. 11. Forestal DA, Haimovici J, Haddad P. Different effect of cold storage and rewarming on three pH regulating transporters in isolated rat hepatocytes. Am J Physiol 1997;272:G638-G645. 12. Gupta M, Dobashi K, Greene EL, Orak JK, Singh I. Studies on hepatic injury and antioxidant enzyme activities in the rat subcellular organelles following in vivo ischemia and reperfusion. Mol Cell Biochem 1997;176:337-347. 13. Isozaki H, Fujii K, Nomura E, Hara H. Calcium concentration in hepatocytes during liver ischaemia-reperfusion injury and the effects of diltiazem and citrate on perfused rat liver. Eur J Gastroenterol Hepatol 2000;12:291-297. 14. Loten EG, Redshaw-Loten JC. Preparation of rat plasma membranes in a high yield. Anal Biochem 1986;154:183-185. 15. Lievremont J-P, Hill A-M, Hilly M, Mauger J-P. The inositol 1,4,5-triphosphate receptor is localized on specialized sub-regions of the endoplasmic reticulum in rat liver. Biochem J 1994; 300:419-427. 16. Franks JJ, Horn J-L, Janicki PK, Singh G. Halothane, isoflurane, xenon and nitrous oxide inhibit calcium ATPase pump activity in rat brain synaptic membranes. Anesthesiology 1995;82:108117. 17. Gatto C, Hale CC, Xu W, Milanick MA. Eosin, a potent inhibitor of the plasma membrane Ca pump, does not inhibit the cardiac Na-Ca exchanger. Biochemistry 1995;34:965-972. 18. Treiman M, Caspersen C, Christensen SB. A tool coming of age: Thapsagargin as an inhibitor of sarcoendoplasmic reticulum Ca2⫹-ATPases. Trends Pharmacol Sci 1998;19:131-135. 19. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248-254. 20. Chan K-M, Junger KD. Calcium transport and phosphorylated intermediate of (Ca2⫹ ⫹ Mg2⫹)-ATPase in plasma membranes of rat liver. J Biol Chem 1983;258:4404-4410.
2⫹
Cold Ischemia Inhibits Hepatic Ca
21. Kraus-Friedmann N, Biber J, Murer H, Carafoli E. Calcium uptake in isolated hepatic plasma-membrane vesicles. Eur J Biochem 1982;129:7-12. 22. Drougas JG, Barnard SE, Wright JK, Sika M, Lopez RR, Stockes KA, et al. A model for the extended studies of hepatic hemodynamics and metabolism in swine. Lab Anim Sci 1996;46:648654. 23. Sika M, Drougas JG, Becker YT, Chapman WC, Wright JK, Donovan KL, et al. Porcine model of orthotopic liver transplantation for chronic studies. In: Tumbleson ME, Schook LB (eds). Advances in swine in biomedical research, vol I. New York: Plenum, 1996:171-187. 24. Bachorik PS, Kwiterovich PO, Cooke JC. Isolation of a porcine plasma membrane fraction that binds low density lipoproteins. Biochemistry 1978;17:5287-5299. 25. Lhoest GJJ, Gougnard TYR, Verbeeck RK, Maton N, Dehoux JP, Wallemacq P, et al. Isolation from pig liver microsomes, identification by tandem mass spectometry and in vitro immunosuppressive activity of an SDZ-RAD 17,18,19,20,21,22-trisepoxide. J Mass Spectrom 2000;35:454-460. 26. Puviani AC, Ottolenghi C, Tassinari B, Pazzi P, Morsiani E. An update on high-yield hepatocyte isolation methods and on the potential clinical use of isolated liver cells. Comp Biochem Physiol A Mol Integr Physiol 1998;121:99-109. 27. Stauffer TP, Guerini D, Carafoli E. Tissue distribution of the four gene products of the plasma membrane Ca2⫹ pump. J Biol Chem 1995;270:12184-12190. 28. DeJaegere S, Wuytack F, Eggermont JA, Verboomen H, Casteels R. Molecular cloning and sequencing of the plasma-membrane
29.
30.
31.
32.
33.
34.
35.
36.
-ATPases
139
Ca2⫹ pump of pig smooth muscle. Biochem J 1990;271:655660. Haddad P, Cabrillac J-C, Piche D, Musalla L, Huet P-M. Changes in intracellular calcium induced by acute hypothermia in parenchymal, endothelial and Kupffer cells of the rat liver. Cryobiology 1999:39:69-79. Kim J-S, Southard JH. Alteration in cellular calcium and mitochondrial functions in the rat liver during cold preservation. Transplantation 1998;65:369-375. Kohli V, Madden JF, Bentley RC, Clavien PA. Calpain mediates ischemic injury of the liver through modulation of apoptosis and necrosis. Gastroenterology 1999;116:168-178. Simbula G, Glascott PA, Akita S, Hoek JB, Farber JL. Two mechanisms by which ATP depletion potentiates induction of the mitochondrial permeability transition. Am J Physiol 1997; 273:C479-C488. Borghi-Scoazec G, Scoazec JY, Durnad F, Bernuau J, Belghiti J, Feldmann G, et al. Apoptosis after ischemia-reperfusion in human liver allografts. Liver Transpl Surg 1997;3:407-415. McConkey DJ, Orrenius S. The role of calcium in the regulation of apoptosis. Biochem Biophys Res Commun 1997;239:357366. Adam R, Bismuth H, Diamond T, Ducot B, Morino M, Astarcioglu I, et al. Effect of extended cold ischemia with UW solution on graft function after liver transplantation. Lancet 1992;340:1373-1376. Geeraerts MD, Ronveax-Dupal M-F, LeMasters JJ, Herman B. Cytosolic free Ca 2⫹ and proteolysis in lethal oxidative injury in endothelial cells. Am J Physiol 1991;261:C889-C896.
From the Departments of *Anesthesiology and †Surgery, Division of Hepatobiliary Surgery and Liver Transplantation, Vanderbilt University Medical Center, Nashville, TN. Supported in part by Individual National Research Service Award no. 1 F32 DK09959-01 from the National Institutes of Health (P.E.W.). Presented in part at the Digestive Disease Week Meeting, May 21-24, 2000, Orlando, FL; the American Association for the Study of Liver Disease 50th Annual Meeting, November 5-9, 1999, Dallas, TX; and the American Association for the Study of Liver Disease 51st Annual Meeting, October 27-31, 2000, Dallas, TX. Address reprint requests to Piotr K. Janicki, MD, PhD, Department of Anesthesiology, VUMC, 504 Oxford House, 1313 21st Ave S, Nashville, TN 37232-4125. Telephone: 615-936-2800; FAX: 615-9362801; E-mail: [email protected] Copyright © 2001 by the American Association for the Study of Liver Diseases 1527-6465/01/0702-0018$35.00/0 doi:10.1053/jlts.2001.21459
132
than ER Ca2ⴙ-ATPase activity in pig liver homogenates. (Liver Transpl 2001;7:132-139.)
P
rimary graft nonfunction (PGNF) is an important and persistent problem in liver transplantation, occurring at the rate of up to 8%.1 Its cause remains unclear, but it appears to be related to ischemia-reperfusion injury, which is likely mediated at least in part by intracellular calcium ([Ca2⫹]i) accumulation in hepatic cells.2,3 These cells contain 2 families of Ca2⫹-adenosinetriphosphatases (ATPases; pumps) that maintain the optimal [Ca2⫹]i concentration in physiological conditions by transporting Ca2⫹ out of the cytosol externally (plasma membrane Ca2⫹-ATPase [PMCA]) or internally into the endoplasmic reticulum (sarcoendoplasmic Ca2⫹-ATPase [SERCA]).4 These 2 pumps are the primary means by which hepatic cells decrease cytosolic Ca2⫹ levels given that the Na⫹-Ca2⫹ exchanger does not have a significant role in the regulation of Ca2⫹ in the liver.5,6 PMCA and SERCA activity have been shown previously in rodent liver preparations,7,8 but the isoforms of PMCA differ somewhat between humans and rats.9 Conversely, peptide sequencing of some porcine PMCA and SERCA isoforms show that they have greater homology to human isoforms than the rat (Swiss Institute of Bioinformatics peptide sequence databases search using BLAST network service, August 2000). For this reason, we chose to test whether the porcine liver model is a closer equivalent to human liver than the rat model for the in vitro study of Ca2⫹ATPase activities. The first aim of this study is to show that our preparative techniques were sufficient to adequately separate plasma membrane (PM) and microsomal or endoplasmic reticulum (ER) fractions from porcine liver, as well as to show PMCA and SERCA activity in these fractions. The second aim is to compare the activity of PMCA and SERCA in membrane preparations from rat, pig, and human liver samples. After substantiating that the porcine model is closer to human liver than the rat, we used the porcine model to study the effects of cold preservation on the activity of hepatic PMCA and SERCA. The activity of all Ca2⫹-ATPases is temperature- and adenosine triphos-
Liver Transplantation, Vol 7, No 2 (February), 2001: pp 132-139
2⫹
Cold Ischemia Inhibits Hepatic Ca
-ATPases
133
Figure 1. List of putative factors involved in the modulation of activity of the Ca2ⴙ-ATPase pumps.
phate (ATP)-dependent10 and susceptible to modulation by a number of factors (Fig. 1). It has been shown that hypothermic storage produces disturbances in several cellular systems,11,12 including Ca2⫹-regulating systems.3,13 Therefore, the final aim of this study is to investigate whether cold preservation of clinically significant duration could produce changes in PMCA and SERCA activity in pig liver membrane fractions.
Materials and Methods Liver Procurement and Storage The Vanderbilt University Institutional Animal Care and Use Committee approved all protocols for animal use in this study. Experiments using the porcine model were performed on healthy cross-bred farm pigs (20 to 25 kg) aged between 9 and 12 weeks. Animals were housed in pens, kept on a 12:12hour light-dark cycle, fed once daily with pig chow, and given water ad libitum. The pigs were acclimatized at our facility for 7 to 10 days before the operation. After a 24-hour fast, the animals were anesthetized with ketamine (30 mg/kg intramuscularly [IM]), atropine (0.05 mg/kg IM), and acetylpromazine (0.1 mg/kg IM); the trachea was intubated; and the endotracheal tube was connected to the ventilator. Anesthesia was maintained with isoflurane, 0.5% to 1.25% (endtidal concentrations) in oxygen. The liver was harvested after midline celiotomy and used immediately for the preparation of membranes (see Preparation of Membrane Fractions). For the cold ischemia studies, aortic and portal cannulas were placed in situ, and the liver was perfused with ice-cold (4°C) University of Wisconsin (UW) lactobionate preservation solution (2 L). The livers were then stored in a plastic bag placed in ice slush for 2 to 18 hours before preparation of the membranes. For the rat model experiments, male Sprague-Dawley rats (250 to 300 g) were allowed food and water ad libitum until the morning of the experiment. Animals were anesthetized with isoflurane in oxygen and killed by decapitation. The liver was quickly harvested after midline celiotomy and processed as described (see Preparation of Membrane Fractions). Samples of fresh human liver from patients under isoflurane anesthesia undergoing partial liver resection for tumor were quickly obtained from the surgical pathology laboratory. Healthy human liver was macroscopically excised from the tu-
mor (used for histopathologic studies) and used for the preparation of the membrane fractions according to the method described next.
Preparation of Membrane Fractions PM vesicles from each species were prepared according to the method of Loten and Redshaw-Loten14 (1986) by fractionation on a self-forming Percoll (Sigma, St Louis, MO) gradient. The livers were homogenized in ice-cold medium containing 0.25 mol/L of sucrose and 10 mmol/L of HEPES/ potassium hydroxide (KOH), pH 7.4 (homogenization buffer), and centrifuged at 1,500g for 10 minutes at 4°C. The resulting pellets were resuspended in 75.0 mL of the homogenization buffer, thoroughly mixed with Percoll and sucrose (75.0 mL of pellet suspension: 10.1 mL of Percoll, 1.45 mL of 2 mol/L of sucrose), and centrifuged at 35,000g for 30 minutes at 4°C to produce a crude PM fraction in the middle of the gradient. This fraction was further purified by repeating the Percoll separation in the presence of 1.3 mmol/L of CaCl2 to separate DNA from PM. For the preparation of crude microsomal (ER) fractions, a variation of the technique of Lievremont et al15 (1994) was used. Supernatant from the initial centrifugation described previously was centrifuged at 8,000g for 30 minutes at 4°C. After separation from the pellets, this supernatant was centrifuged at 100,000g for 1 hour at 4°C. The final pellets were washed and resuspended in an ice-cold washing medium containing 0.25 mol/L of sucrose and 25 mmol/L of HEPES/ KOH, pH 7.4. Purity of the PM and ER fractions was assessed by measuring the selective activity of PM and ER enzyme markers in the initial liver homogenates and in subsequent fractions obtained during separation. Enzyme markers included 5⬘-nucleotidase (PM characterization) and glucose-6-phosphatase (ER characterization). Both enzymes were measured spectrophotometrically using commercially available kits (Sigma Diagnostic, St Louis, MO).
Measurement of Ca2ⴙ Transport Measurement of Ca2⫹ uptake by liver PM and ER vesicles was performed as described previously.16 The incubation mixture (4 mL) was composed of 30 mmol/L of imidazole-histidine (pH 6.8), 200 mmol/L of KCl, 5 mmol/L of MgCl2, 5 mmol/L of ATP, 5 mmol/L of sodium azide, 5 mmol/L of ammonium oxalate, and 20 mol/L of CaCl2 containing (final concentra-
134
Janicki et al
tion) 0.1:Ci/mL of 45CaCl2 (NEM Products, Boston, MA; specific activity, 30.7 mCi/mg). The reaction was started by adding aliquots of membrane proteins to each tube (final concentration, 0.07 mg/mL) at 37°C. Aliquots of 0.5 mL were removed after 5 minutes and collected on 25-mm cellulose nitrate filters (0.45 m pore size; Gelman Sci, Ann Arbor, MI) that had been prewashed with 2 mL of 0.25 mol/L of KCl and 10 mL of deionized water. After vesicle collection, the filters were washed with 2 mL of 0.25 mol/L of sucrose. After overnight drying, the filters were placed in vials containing CytoScint (ICN, Costa Mesa, CA), and 45Ca2⫹ activity was assessed in a Beckman LS3801 beta counter (Beckman, Fullerton, CA). Results are expressed as nanomoles of Ca2⫹ accumulated per milligram of membrane protein per minute of incubation time. Blank results were obtained in the presence of 0.2 mmol/L of orthovanadate, a universal inhibitor of all P-type pumps, in the incubation mixture. The specificity of the measured activity for the PM and ER membrane fractions was confirmed by the inhibition of Ca2⫹ uptake by eosin (2 mol/L), a specific inhibitor of PMCA,17 and thapsigargin (300 nmol/L), a specific inhibitor of SERCA.18 Protein content in the PM and ER fractions was measured by the method of Bradford,19 using bovine serum albumin as a standard. The focus of these experiments was to evaluate the activity of Ca2⫹-ATPases, defined by the uptake of 45Ca2⫹ by the membrane vesicles. This activity is believed to be determined by the function of ATPases (defined as the rate at which they convert ATP to adenosine diphosphate and inorganic phosphate (Pi) while moving Ca2⫹ across their respective cellular membranes), as well as the overall numbers of these proteins present in the membranes. Determination of the function and number of PMCAs and SERCAs present in the membrane fractions was not the focus of these experiments.
Statistics Results are expressed as mean ⫾ SEM. Statistical significance of differences was assessed using one-way ANOVA and multiple comparison of the means with the least significant difference test. Statistical significance was inferred for P less than .05.
Results Preparation and Characterization of PM and ER Fractions From Pig Liver Purification of porcine liver subcellular membrane fractions into PM and ER fractions is indicated by the specific activity of the 2 enzyme markers tested. The 5⬘-nucleotidase showed a significant 12-fold PM enrichment in the PM fraction, and the activity of glucose-6-phosphatase showed a significant 7-fold ER enrichment in the ER fraction compared with the crude liver homogenate (Table 1). Similar enrichment was obtained for rat and human liver samples (results not shown). There was limited contamination of the PM fraction with ER and vice versa based on the enzyme marker tests compared with the initial crude membrane homogenate. PM vesicles prepared from homogenized pig liver showed PMCA pumping activity of 0.47 ⫾ 0.03 nmol of Ca2⫹/mg of protein/min. ER vesicles prepared from pig liver cells showed comparatively greater SERCA pumping activity (1.57 ⫾ 0.09 nmol of Ca2⫹/mg of protein/min) in relation to PMCA (Table 2). We used several criteria based on the addition of specific inhibitors (orthovanadate, eosin, and thapsigargin) to show that the active Ca2⫹ transport in our PM and ER preparations shares properties consistent with those known for PMCA and SERCA from other sources. Radioactive calcium accumulation in vesicles was substantially inhibited by orthovanadate, with inhibitory constant (Ki) ⫽ 45 mol/L for PMCA and Ki ⫽ 50 mol/L for SERCA, in the range expected for these P-type pumps.20 Ca2⫹ uptake values of membrane vesicles in the presence of orthovanadate were therefore considered blanks. No Ca2⫹ uptake (100% inhibition) was shown in PM vesicles incubated in the presence of specific PMCA inhibitor (eosin), whereas only slight SERCA inhibi-
Table 1. Activity of PM Marker 5⬘-Nucleotidase and ER Marker Glucose-6-Phosphatase in Subcellular Fractions Prepared From Porcine Liver
Fraction Liver homogenate PMS ER membranes
5⬘-Nucleotidase (nmol/mg protein/min)
Glucose-6-Phosphatase (mol/mg protein/min)
0.10 ⫾ 0.04 1.2 ⫾ 0.1* 0.10 ⫾ 0.07
0.25 ⫾ 0.03 0.31 ⫾ 0.06 1.8 ⫾ 0.3*
NOTE. Values expressed as mean ⫾ SD. Three separate livers were used in each experiment (run in triplicate). * P ⬍ .05 by ANOVA and t-test compared with liver homogenate.
2⫹
Cold Ischemia Inhibits Hepatic Ca
135
-ATPases
Table 2. Interspecies Differences of PMCA and SERCA Activities
Species
PMCA Activity (nmoles Ca2⫹/mg protein/min)
SERCA Activity (nmoles Ca2⫹/mg protein/min)
Human Pig Rat
0.53 ⫾ 0.1 (n ⫽ 5) 0.47 ⫾ 0.03 (n ⫽ 14) 2.6 ⫾ 0.4 (n ⫽ 20)*
1.4 ⫾ 0.1 (n ⫽ 4) 1.57 ⫾ 0.1 (n ⫽ 10) 9.2 ⫾ 0.9* (n ⫽ 15)
NOTE. Values expressed as mean ⫾ SD. N indicates the number of livers used in each experiment (each run in triplicate). * P ⬍ .05 by ANOVA and t-test (rat versus human and pig).
tion (15% to 20%) was observed in the presence of eosin in the microsomal fraction. The specific SERCA inhibitor (thapsigargin) produced a 20% inhibition of Ca2⫹ pumping in PM vesicles and greater than 50% to 60% inhibition of Ca2⫹ pumping in the microsomal preparation (Fig. 2). The cross-inhibition is explained by the minimal contamination of each purified fraction by the other membrane type, seen in the enzyme marker results previously described (Table 1). Interspecies Differences in Hepatic Ca2ⴙ-ATPase Activity Results of the interspecies comparison of the intracellular Ca2⫹-ATPase activity in PM and ER are listed in Table 2. PM vesicles prepared from pig liver homogenates showed PMCA pumping activity of 0.47 ⫾ 0.03 nmol of Ca2⫹/mg of protein/min. Similar PMCA pumping activity was observed in the human liver homogenates (0.53 ⫾ 0.1 nmol of Ca2⫹/mg of protein/
min). Conversely, rat liver PMs were characterized by 5 times greater PMCA pumping activity (2.6 ⫾ 0.4 nmol of Ca2⫹/mg of protein/min; P ⬍ .05 by ANOVA followed by t-test). A similar difference was observed for the SERCA activities. SERCA activity in the rat liver homogenate ER fraction was 9.2 ⫾ 0.9 nmol of Ca2⫹/mg of protein/min, which is significantly greater (P ⬍ .05) than the SERCA activity in pig and human liver homogenates (1.4 ⫾ 0.1 and 1.57 ⫾ 0.10 nmol of Ca2⫹/mg of protein/min, respectively). Effect of Cold Preservation on PMCA Activity Results of these experiments are shown in Figure 3A. There was no statistically significant difference between the PMCA pumping activity in PM vesicles prepared immediately after procurement and after 2 hours of cold preservation in UW solution. Storage of the liver on ice for 4 and 6 hours produced partial time-dependent inhibition of PMCA activity (P ⬍ .05). Cold storage for 12 and 18 hours produced almost complete inhibition of PMCA activity compared with liver samples from the time of procurement and 2 hours after flushing with UW solution (P ⬍ .01). Effect of Cold Preservation on SERCA Activity
Figure 2. Activity of Ca2ⴙ-ATPase in PM and ER fractions prepared from whole-pig liver homogenates. Columns represent means for 3 separate livers. All measurements were performed in triplicate. Error bars indicate SEM. *P < .05, significant difference from controls by ANOVA followed by least significant difference test.
Results of these experiments are shown in Figure 3B. There was no statistically significant difference between the SERCA pumping activity in ER vesicles prepared immediately after procurement and after 6 hours of cold preservation in UW solution. However, there was a statistically significant partial inhibition of SERCA pumping activity in ER vesicles prepared after 12 and 18 hours of cold storage compared with the activity immediately after liver procurement (45% and 69% inhibition, respectively; both P ⬍ .05).
Discussion Previous work on the separation and characterization of hepatic PMCA and SERCA has primarily focused
136
Janicki et al
Figure 3. Effects of increasing length of cold preservation of pig liver on the activity of (A) PMCA and (B) SERCA prepared from whole-porcine liver homogenates. Columns represent means for 3 separate livers. All measurements were performed in triplicate. Blank values were obtained from PM and ER membranes incubated in the presence of 0.2 mmol/L of orthovanadate. Error bars indicate SEM. *P < .05, significant difference by ANOVA followed by least significant difference test versus 0-hour (no preservation) controls. **P < .01.
on the use of the rat model.8,21 Our laboratory uses the porcine liver model not only because of our familiarity with it,22,23 but it allows for greater tissue volumes for fresh membrane studies than the rat and more analogous operative situations to humans. Although other laboratories have separated and identified PM and ER from pig liver,24,25 this is the first time these membranes have been separated from pig liver for use in Ca2⫹-ATPase studies. Our data from the separation and characterization of the membrane fractions from the pig liver are equivalent to those from pig and rat liver membrane separations,8,21 suggesting that our separation methods are adequate. Based on the membrane marker enzyme studies, there is some contamination of our different porcine membrane fractions, but this is also consistent with the results of the other studies.8,21 The second aim of our study is to compare the activity of PMCA and SERCA in membrane preparations from rat, pig, and human livers by using 45Ca 2⫹ uptake and specific Ca 2⫹-ATPase inhibitors. Our results confirm that PM and ER fractions from the livers of these 3 species contain significant activity of both PMCA and SERCA. There also appears to be another Ca2⫹ entry mechanism in the microsomal membrane fraction that is not affected by the SERCA-selective inhibitor, thapsigargin, based on only 50% to 60% ER fraction Ca2⫹ uptake in the presence of this inhibitor. This uptake is most likely explained by contamination of the microsomal fraction with other membranes containing Ca2⫹ exchange mechanisms or pumps (likely from mitochondria). The use of mitochondrial Ca2⫹ exchange
inhibitors (e.g., ruthenium red) might clarify the source of this additional microsomal Ca2⫹ uptake. PMCA activity in our pig liver preparations was similar to that observed in membrane preparations from human liver and approximately 5-fold less than the activity observed in rat preparations. A similar trend was observed for the interspecies differences in SERCA activity. These data show for the first time the interspecies differences in the activity of the PMCA and SERCA molecules from liver preparations. This further supports the conclusion that the porcine hepatic model is likely better than the rat model as an analogy for the study of human liver Ca2⫹-ATPases. Although the pig is closer both immunologically and physiologically to humans than the rat,26 the substantial activity differences of PMCA and SERCA between rat and human and pig livers are somewhat unexpected given what is known about the different Ca2⫹-ATPase isoforms in these species. PMCA has been found in all mammalian cells and is encoded by 4 independent genes. Alternative splicing further increases the number of possible isoforms. In humans, PMCA1, PMCA2, and PMCA4 isoforms have been identified in the liver, whereas only PMCA1 was identified in rat liver by 1 group27 and both PMCA1 and PMCA2 were the only isoforms in the rat liver identified by another group.9 Pig liver PMCA isoforms have not been specifically sequenced, but the 1 peptide sequence of porcine PMCA1 from smooth muscle has shown 98% homology to the human PMCA1 and 93% homology to the analogous rat isoform.28 The SERCA family is composed of products from at least 3 genes, but they have not been as thoroughly studied. SERCA2B, an isoform
2⫹
Cold Ischemia Inhibits Hepatic Ca
of SERCA2, appears to be expressed universally in all cells, including liver cell membranes. One of the 2 SERCA isoforms sequenced in the 3 species we studied shows greater homology between humans and pigs than rats (Swiss Institute of Bioinformatics peptide sequence databases search using BLAST network service, August 2000). The homology between the identified isoforms of pig and human PMCA and SERCA likely explains their similar activities in our experiment. Conversely, the substantial differences in their activities compared with rat liver membranes are more likely explained by preferential expression of isoforms with greater function in rats or as yet undiscovered alternative-spliced rat Ca2⫹-ATPase isoforms. It is also possible that the difference in the method of obtaining the human liver from resection specimens might alter their Ca2⫹ATPase activities. The role of Ca2⫹-ATPases in the maintenance of normal hepatic [Ca2⫹]i levels after ischemia-reperfusion appears to be dependent on maintaining their function after transplantation. Experimental evidence supports the presence of several mechanisms in liver cells that can be responsible for increased [Ca2⫹]i levels after ischemia-reperfusion.3 We postulated that liver ischemia during cold preservation causes a moderate increase in [Ca2⫹]i level, in part caused by emptying [Ca2⫹]i stores and decreased calcium clearance by PMCA and SERCA. Similar changes have also been observed in each of the major hepatic cell types during ischemia.29 When the liver graft is subsequently transplanted into a Ca2⫹-rich environment, there is a massive inflow
Figure 4. Hypothetical series of events producing intracellular Ca2ⴙ-overload and liver dysfunction and/or PGNF after liver preservation and reperfusion. *CCE, capacitative entry.
-ATPases
137
of calcium into the liver cells through the Ca2⫹ release– activated Ca2⫹ channels, a process known as capacitative calcium entry (CCE).30 If Ca2⫹-ATPases are significantly inhibited by cold preservation of the organ, they may be unable to compensate for the postreperfusion cellular calcium inflow. Subsequent cellular dysfunction and death can occur because of increased [Ca2⫹]i levels, causing uncontrolled activation of proteolytic activity31 and mitochondrial calcium overload, which can lead to mitochondrial damage and derangement of cellular energy production, resultant cell death, and tissue necrosis.32 There is evidence that ischemiareperfusion injury of the liver leads not only to cellular necrosis, but also apoptosis.33,34 If cellular injury is significant enough, graft dysfunction or PGNF may occur after transplantation (Fig. 4). If this hypothetical chain of events is correct, the major triggering points for Ca2⫹ overload during the reperfusion phase of transplantation are emptying of the [Ca2⫹]i stores during cold preservation, causing activation of CCE mechanisms, and the inhibition of cellular Ca2⫹ pumping. Therefore, approaches to maintain [Ca2⫹]i homeostasis during storage might improve organ preservation and decrease the intensity of reperfusion injury in liver transplantation. As the third aim of our study, we therefore wished to establish whether PMCA and SERCA activities are affected by cold preservation. Neither pump activity seemed to be significantly decreased after short-term cold preservation in UW solution (⬍12 hours). However, prolonged cold storage (ⱖ12 hours) of the porcine liver produced almost complete inhibition of the
138
Janicki et al
PMCA pumping mechanism. Similarly, SERCA activity showed a significant decrease after 12 hours of cold storage, but this inhibition was less prominent compared with the decrease in PMCA activity after the same period. This suggests that the PMCA pump is more sensitive to cold ischemia than the SERCA pump. That these pumps show significant inhibition after 12 hours correlates with the increasing rate of PGNF and graft dysfunction after 12 hours or more of cold ischemia in human liver transplantation.35 It is unclear at the present time whether the inhibition of the Ca2⫹ATPases observed in the cold-preserved liver in our experiment is permanent or reversible by reimplantation of the liver graft. It is also unclear whether the activity is decreased because the numbers of functional proteins are decreased or the function of those proteins is diminished. Because these studies were conducted on whole-liver homogenates consisting of a mixture of cells, it is unclear whether PMCA and SERCA activity changes are similar for all hepatic cell types. Different hepatic cells have different sensitivities to cold and warm ischemia and reperfusion,29 and their ability to regulate [Ca2⫹]i after an ischemic insult also differs.36 Whether these same liver cells have different sensitivities of their Ca2⫹ATPases after ischemia-reperfusion has not been established. It will be necessary to confirm our studies in membranes from isolated and purified liver cells to associate the PMCA and/or SERCA activity with the specific types of liver cells. In conclusion, the present study shows that PM and ER membranes can be separated from porcine, human, and rat liver homogenates, and the activity of the PMCA and SERCA pumps can be measured from these livers. We also found that the Ca2⫹-ATPase activities in the porcine liver homogenate are closer to the human liver than is the rat, showing that the pig is a better model for the study of human liver Ca2⫹-ATPases in vitro. Finally, we showed that ischemia associated with the cold storage of the porcine liver produces inhibition of the cellular Ca2⫹-ATPase pumps, PMCA more than SERCA. This inhibition may contribute to the disturbances in the [Ca2⫹]i clearance mechanisms after reperfusion that lead to liver graft dysfunction and PGNF after liver transplantation.
References 1. Porte RJ, Ploeg RJ, Hansen B, van Bockel JH, Thorogood J, Persijn GG, et al. Long-term graft survival after liver transplantation in the UW era: Late effects of cold ischemia and primary dysfunction. The European Multicentre Study Group. Transpl Int 1998;11(supp 1):S164-S167.
2. Tredger JM. Ischaemia-reperfusion injury in the liver: Treatment in theory and in practice. Biofactors 1998;8:161-164. 3. Clavien PA, Harvey PRC, Strasberg SM. Preservation and reperfusion injuries in liver allografts: An overview and synthesis of current studies. Transplantation 1992;53:957-978. 4. Kuo TH, Liu B-F, Yu Y, Wuytack F, Raeymaekers L, Tsang W. Co-ordinated regulation of the plasma membrane calcium pump and the sarco(endo)plasmic reticular calcium pump gene expression by Ca2⫹. Cell Calcium 1997;21:399-408. 5. Lidofsky SD, Xie M-H, Scharschmidt BF. Na2⫹-Ca2⫹ exchange in cultured rat hepatocytes: Evidence against a role in cytosolic Ca2⫹ regulation or signaling. Am J Physiol 1990;259:G56-G61. 6. Kraus-Friedmann N. Calcium sequestration in the liver. Cell Calcium 1990;11:625-640. 7. Kessler F, Bennardini F, Bachs O, Serratosa J, James P, Caride AJ, et al. Partial purification and characterization of the Ca2⫹pumping ATPase of the liver plasma membrane. J Biol Chem 1990;265:16012-16019. 8. Dawson AP, Fulton DV. Some properties of the Ca 2⫹-stimulated ATPase of a rat liver microsomal fraction. Biochem J 1983; 210:405-410. 9. Howard A, Barley NF, Legon S, Walters JR. Plasma-membrane calcium-pump isoforms in human and rat liver. Biochem J 1994; 303:275-279. 10. Carafoli E, Brini M. Calcium pumps: Structural basis for and mechanism of calcium transmembrane transport. Curr Opin Cell Biol 2000;4:152-161. 11. Forestal DA, Haimovici J, Haddad P. Different effect of cold storage and rewarming on three pH regulating transporters in isolated rat hepatocytes. Am J Physiol 1997;272:G638-G645. 12. Gupta M, Dobashi K, Greene EL, Orak JK, Singh I. Studies on hepatic injury and antioxidant enzyme activities in the rat subcellular organelles following in vivo ischemia and reperfusion. Mol Cell Biochem 1997;176:337-347. 13. Isozaki H, Fujii K, Nomura E, Hara H. Calcium concentration in hepatocytes during liver ischaemia-reperfusion injury and the effects of diltiazem and citrate on perfused rat liver. Eur J Gastroenterol Hepatol 2000;12:291-297. 14. Loten EG, Redshaw-Loten JC. Preparation of rat plasma membranes in a high yield. Anal Biochem 1986;154:183-185. 15. Lievremont J-P, Hill A-M, Hilly M, Mauger J-P. The inositol 1,4,5-triphosphate receptor is localized on specialized sub-regions of the endoplasmic reticulum in rat liver. Biochem J 1994; 300:419-427. 16. Franks JJ, Horn J-L, Janicki PK, Singh G. Halothane, isoflurane, xenon and nitrous oxide inhibit calcium ATPase pump activity in rat brain synaptic membranes. Anesthesiology 1995;82:108117. 17. Gatto C, Hale CC, Xu W, Milanick MA. Eosin, a potent inhibitor of the plasma membrane Ca pump, does not inhibit the cardiac Na-Ca exchanger. Biochemistry 1995;34:965-972. 18. Treiman M, Caspersen C, Christensen SB. A tool coming of age: Thapsagargin as an inhibitor of sarcoendoplasmic reticulum Ca2⫹-ATPases. Trends Pharmacol Sci 1998;19:131-135. 19. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248-254. 20. Chan K-M, Junger KD. Calcium transport and phosphorylated intermediate of (Ca2⫹ ⫹ Mg2⫹)-ATPase in plasma membranes of rat liver. J Biol Chem 1983;258:4404-4410.
2⫹
Cold Ischemia Inhibits Hepatic Ca
21. Kraus-Friedmann N, Biber J, Murer H, Carafoli E. Calcium uptake in isolated hepatic plasma-membrane vesicles. Eur J Biochem 1982;129:7-12. 22. Drougas JG, Barnard SE, Wright JK, Sika M, Lopez RR, Stockes KA, et al. A model for the extended studies of hepatic hemodynamics and metabolism in swine. Lab Anim Sci 1996;46:648654. 23. Sika M, Drougas JG, Becker YT, Chapman WC, Wright JK, Donovan KL, et al. Porcine model of orthotopic liver transplantation for chronic studies. In: Tumbleson ME, Schook LB (eds). Advances in swine in biomedical research, vol I. New York: Plenum, 1996:171-187. 24. Bachorik PS, Kwiterovich PO, Cooke JC. Isolation of a porcine plasma membrane fraction that binds low density lipoproteins. Biochemistry 1978;17:5287-5299. 25. Lhoest GJJ, Gougnard TYR, Verbeeck RK, Maton N, Dehoux JP, Wallemacq P, et al. Isolation from pig liver microsomes, identification by tandem mass spectometry and in vitro immunosuppressive activity of an SDZ-RAD 17,18,19,20,21,22-trisepoxide. J Mass Spectrom 2000;35:454-460. 26. Puviani AC, Ottolenghi C, Tassinari B, Pazzi P, Morsiani E. An update on high-yield hepatocyte isolation methods and on the potential clinical use of isolated liver cells. Comp Biochem Physiol A Mol Integr Physiol 1998;121:99-109. 27. Stauffer TP, Guerini D, Carafoli E. Tissue distribution of the four gene products of the plasma membrane Ca2⫹ pump. J Biol Chem 1995;270:12184-12190. 28. DeJaegere S, Wuytack F, Eggermont JA, Verboomen H, Casteels R. Molecular cloning and sequencing of the plasma-membrane
29.
30.
31.
32.
33.
34.
35.
36.
-ATPases
139
Ca2⫹ pump of pig smooth muscle. Biochem J 1990;271:655660. Haddad P, Cabrillac J-C, Piche D, Musalla L, Huet P-M. Changes in intracellular calcium induced by acute hypothermia in parenchymal, endothelial and Kupffer cells of the rat liver. Cryobiology 1999:39:69-79. Kim J-S, Southard JH. Alteration in cellular calcium and mitochondrial functions in the rat liver during cold preservation. Transplantation 1998;65:369-375. Kohli V, Madden JF, Bentley RC, Clavien PA. Calpain mediates ischemic injury of the liver through modulation of apoptosis and necrosis. Gastroenterology 1999;116:168-178. Simbula G, Glascott PA, Akita S, Hoek JB, Farber JL. Two mechanisms by which ATP depletion potentiates induction of the mitochondrial permeability transition. Am J Physiol 1997; 273:C479-C488. Borghi-Scoazec G, Scoazec JY, Durnad F, Bernuau J, Belghiti J, Feldmann G, et al. Apoptosis after ischemia-reperfusion in human liver allografts. Liver Transpl Surg 1997;3:407-415. McConkey DJ, Orrenius S. The role of calcium in the regulation of apoptosis. Biochem Biophys Res Commun 1997;239:357366. Adam R, Bismuth H, Diamond T, Ducot B, Morino M, Astarcioglu I, et al. Effect of extended cold ischemia with UW solution on graft function after liver transplantation. Lancet 1992;340:1373-1376. Geeraerts MD, Ronveax-Dupal M-F, LeMasters JJ, Herman B. Cytosolic free Ca 2⫹ and proteolysis in lethal oxidative injury in endothelial cells. Am J Physiol 1991;261:C889-C896.