Orthotopic liver transplantation has emerged as the

Transcription

1 Cold Preservation Warm Reperfusion Perturbs Cytosolic Calcium Ion Homeostasis in Rat Liver Sinusoidal Endothelial Cells Stéphanie Auger, Diane Vallerand, and Pierre Selim Haddad Increases in intracellular calcium ion (Ca 2 ) levels of sinusoidal endothelial cell (SEC) may have a crucial role in mediating the expression of adhesion molecules and thus contribute to the microcirculatory disturbances observed in primary graft dysfunction. The effect of changes in the composition and/or temperature of the reperfusion solution on cytosolic Ca 2 was studied in isolated rat SECs. Cells were preserved in cold University of Wisconsin (UW) solution for 0, 12, or 24 hours and loaded with Fura-2AM dye (Cedarlane, Eugene, OR) at 20 C in N-2-hydroxyethylpiperazine-propanesulfonic acid (HEPES)-buffered physiological solution (HEPES 20 C) or UW solution (UW 20 C). SEC Ca 2 levels were measured by cytofluorimetry. Basal steady-state Ca 2 levels were much lower when SECs were loaded in UW 20 C (37 2 nmol/l) than in HEPES 20 C ( nmol/ L). In unstored controls (0 hour), going from UW 20 C to HEPES 37 C induced a large transient increase ( nmol/l) in SEC Ca 2 levels, which was greatly inhibited (43 13 nmol/l) in Ca 2 -free HEPES 37 C. A similar large transient increase was observed going from UW 20 C to HEPES 20 C ( nmol/l). Changing temperature only (20 C to 37 C) in UW or HEPES solution had a much smaller effect on SEC Ca 2 levels (14 2 and nmol/l, respectively). These changes were similar in cold-preserved cells. In unstored controls, solution changes greatly attenuated the intensity of subsequent Ca 2 responses to the purinergic agonist adenosine triphosphate (ATP). Cold preservation (CP) greatly attenuated both the frequency of appearance and intensity of ATP-induced Ca 2 responses. Hence, changing reperfusion solution composition has a greater impact on SEC steady-state Ca 2 levels than changing temperature. Cold From the Department of Pharmacology and Membrane Transport Research Group, Université de Montréal, Quebec, Canada. Supported in part by grant no. MOP14579 from the Canadian Institutes of Health Research; studentships from the Groupe de Recherche en Transport Membranaire, Montréal (S.A.) and the Faculty of Graduate Studies of the Université de Montréal (S.A.); and P.S.H. is a National Research Scholar of the Fonds de la Recherche en Santé du Québec. Presented in part at the 52 nd Annual Meeting of the American Association for the Study of Liver Diseases, November 9-13, 2001, Dallas, TX. Published in abstract form in Hepatology 34:194A, Address reprint requests to Pierre S. Haddad, PhD, Department of Pharmacology, Université de Montréal, PO Box 6128, Downtown Station, Montreal, QC, H3C 3J7, Canada. Telephone: ; FAX: ; pierre.haddad@umontreal.ca Copyright 2003 by the American Association for the Study of Liver Diseases /03/ $35.00/0 doi: /jlts preservation does not significantly affect changes in SEC steady-state Ca 2 levels, but greatly impairs the capacity of SECs to subsequently respond to Ca 2 -mobilizing agonists. (Liver Transpl 2003;9: ) Orthotopic liver transplantation has emerged as the most effective means of treatment for end-stage liver disease during the last 15 to 20 years. Unfortunately, liver graft failure remains a significant problem because approximately 15% of grafts still show primary dysfunction or nonfunction, which cause serious morbidity and mortality. 1 It now is widely suggested that this dysfunction may be caused by organ injuries during the process of cold preservation (CP) and warm reperfusion (WR) necessitated by the transplantation procedure. 1-3 Precise mechanisms of these injuries are still unclear, but sinusoidal endothelial cells (SECs) appear to be particularly vulnerable to cold storage of the liver and could be implicated in local inflammatory reactions. 4-6 During CP-WR events, SECs undergo dramatic morphological changes 6-8 and express adhesion molecules that contribute to the microcirculatory occlusion observed during primary dysfunction. 1,7,9 Intracellular calcium ions (Ca 2 ) could have a key role in this disorder. In general, Ca 2 is usually maintained at very low levels in the cytoplasm (10 7 mol/l) compared with the extracellular medium (10 3 mol/l). 10 These low concentrations are maintained by compartmentalization of free Ca 2 into the mitochondria and endoplasmic reticulum, its active extrusion across the plasma membrane, and its chelation by cell proteins. 1 However, hypothermic preservation reduces the production of adenosine triphosphate (ATP) by the mitochondria, which limits the extrusion and/or storage of cell Ca 2 by ATP-dependent pumps. Hypothermia also induces an imbalance of the cell electrical potential that can lead to an uncontrolled influx of Ca 2 through plasma membrane Ca 2 channels. 1 Moreover, reactive oxygen intermediates released during reperfusion 1,11 are known to cause Ca 2 influx. 12 Such alterations in cytosolic Ca 2 levels in preserved SECs could activate multiple intracellular cascades 150 Liver Transplantation, Vol 9, No 2 (February), 2003: pp

2 Endothelial Cell Ca 2 in Cold Storge Injury 151 leading to translocation and expression of adhesive molecules on the surface of the cells, among other events. Increases in cytosolic Ca 2 levels trigger the fusion of Weibel-Palade bodies, the secretory organelles of the endothelium, with the plasma membrane. This results in the translocation of GMP-140 (P-selectins) proteins that they contain. 1,13 Conversely, an increase in cytoplasmic Ca 2 levels is known to activate phospholipase A 2, the enzyme that releases platelet-activating factor, a proinflammatory lipid. 9 The appearance of P-selectin and platelet-activating factor on the SEC surface can mediate the interaction between hepatic microvasculature and circulating blood cells and platelets For instance, such an inflammatory process causes an accumulation of white blood cells at the site of injury 14,16,18 and may induce the occlusion of microvessels, causing areas of no reflow during reperfusion, as observed in myocardial ischemia. 19,20 In the liver, it appears that platelets are more involved in this phenomenon at intrahepatic sites. 14 Finally, release of such cytotoxic mediators as reactive oxygen intermediates at the site of inflammation can be harmful for surrounding cells, 1 even leading to the detachment of SECs from the extracellular matrix. 1,4,7 Together, these effects may participate in graft dysfunction; however, the exact role of Ca 2 has not yet been elucidated. Different studies have begun to explore the role of Ca 2 in CP-WR injury. For instance, the use of inhibitors of L-type voltage-dependent Ca 2 channels (e.g., nisoldipine or verapamil) to improve preservation of rat livers was one of the first elements suggesting the important role of intracellular Ca 2 in liver preservation injury. 21,22 However, SECs do not seem to be the target of these drugs. 23,24 Studies by Kim and Southard 25 and our laboratory 24 have shown that acute hypothermia causes an increase in steady-state Ca 2 levels in isolated hepatocytes. We also observed the same process in isolated SECs. 24 In hepatocytes subjected to transplantation-like conditions (CP and WR), our laboratory recently found an increase in steady-state Ca 2 levels and an exaggeration of the response to a Ca 2 -mobilizing agonist. 26 However, studies of SEC Ca 2 homeostasis after CP-WR have never been presented. The aim of this study is to determine the effect of changes in the composition and temperature of reperfusion solutions on steady-state intracellular Ca 2 levels in cold-preserved liver SECs, as well as on their response to the purinergic Ca 2 -mobilizing agonist ATP. Experimental Procedures Cell Isolation, Culture, and Preservation Animals used in this study were treated in accordance with the Canadian Council on the Care of Animals guidelines, and all experimental protocols were approved by our university s ethics committee. SECs were isolated from livers of fed male Wistar rats weighing 250 to 300 g (Charles River, St-Constant, QC, Canada) by a variation of the method of SEC isolation proposed by Braet et al. 27 Briefly, rats were anesthetized with intraperitoneal injections of 50 mg/kg of sodium pentobarbital (Centre de Distribution de Médicaments Vétérinaires, Ste- Hyacinthe, QC, Canada) before surgery. Livers were digested with 0.035% collagenase-a (Roche Diagnostics, Laval, QC, Canada) in a 5% FCS-HI-GBSS solution (137 mmol/l of NaCl, 5 mmol/l of KCl, 0.84 mmol/l of Na 2 HPO 4, 0.22 mmol/l of KH 2 PO 4, 0.28 mmol/l of MgSO 4-7H 2 O, 1 mmol/l of MgCl 2-6H 2 O, 5.5 mmol/l of D-glucose, 2.7 mmol/l of NaHCO 3, and 1.6 mmol/l of CaCl 2-2H 2 O). SECs were purified on a discontinuous Percoll gradient (25% top to 50% bottom) and plated at a concentration of approximately 1 to cells/ml onto collagen S-coated round glass coverslips. The purity of SEC cell preparations was assessed by FACS Scan using fluorescently labeled low-density lipoprotein and exceeded 82.5% (result not shown). SECs then were incubated for 8 hours in RPMI 1640 medium (ph 7.4 at 37 C) supplemented with 2% FCS-HI (Sigma-Aldrich, Mississauga, ON, Canada), NaHCO 3 (24 mmol/l), penicillin (100 U/mL), and streptomycin (100 g/ml) in a humidified atmosphere of 95% oxygen and 5% carbon dioxide to allow the cells to recover from the isolation procedure. Culture medium was then changed for cold UW solution (Du Pont Pharma, Mississauga, ON, Canada), and cells were kept at 4 C for 0 (unstored controls), 12, and 24 hours. This step reproduced the period of cold storage of a liver transplantation. Intracellular Ca 2 Measurement Cells were loaded with 2.5 mol/l of the fluorochrome Fura-2AM (Cedarlane, Eugene, OR) for 30 minutes at 20 C in University of Wisconsin (UW) solution (UW 20 C) or an N-2-hydroxyethylpiperazine-propanesulfonic acid (HEPES)-buffered solution (HEPES 20 C; 10 mmol/l of HEPES, 137 mmol/l of NaCl, 3.8 mmol/l of KCl, 1.2 mmol/l of KH 2 PO 4, 1.2 mmol/l of MgSO 4-7H 2 O, 5 mmol/l of D-glucose, 1 mmol/l of Na-pyruvate, and 1.8 mmol/l of CaCl 2-2H 2 O, ph 7.4, at 20 C). Preliminary experiments showed that dye

3 152 Auger, Vallerand, and Haddad Table 1. Experimental Protocols for Measurement of Intracellular SEC Ca 2 After Changes in Composition and/or Temperature of Reperfusion Solutions Protocol A B C D E UW (20 C) 3 UW (20 C) 3 HEPES (37 ) 3 ATP response [Dye loading 30 ] [2 3 ] [10 ] [1 ] UW (20 C) 3 UW (20 C) 3 Ca 2 -free HEPES (37 C) 3 Ca 2 free ATP response [Dye loading 30 ] [2 3 ] [10 ] [1 ] HEPES (20 C) 3 HEPES (20 C) 3 HEPES (37 C) 3 ATP response [Dye loading 30 ] [2 3 ] [10 ] [1 ] UW (20 C) 3 UW (20 C) 3 UW (37 C) 3 HEPES (37 C) 3 ATP response [Dye loading 30 ] [2 3 ] [5 ] [10 ] [1 ] UW (20 C) 3 UW (20 C) 3 HEPES (20 C) 3 HEPES (37 C) 3 ATP response [Dye loading 30 ] [2 3 ] [10 ] [5 ] [1 ] loading was impossible at 4 C. Coverslips with adhered SECs then were placed in a specially designed 100- L plastic chamber (Groupe de Recherche en Transport Membranaire, Montréal, QC, Canada) on the stage of an inverted microscope (Olympus IMT-2; Carsen Medical, Markham, ON, Canada) coupled to a spectrofluorimeter (Deltascan RF-D4010; Photon Technology Int, London, ON, Canada). Excitation wavelengths were 350 and 380 nm, and fluorescence emission was measured at 505 nm. Intracellular dye calibration was performed in situ on each preparation by perfusion of ionomycin-containing solutions, as previously described. 28 The OSCAR software supplied by Photon Technology Int was used to convert fluorescence ratios (F350/F380, values corrected for autofluorescence) into intracellular Ca 2. After the dye-loading period, cells were perfused at a flow rate of 2 to 3 ml/min with solutions of different composition (UW, HEPES, HEPES Ca 2 -free [containing 4 mmol/l of ethylene glycol-bis( -aminoethyl ether)-n,n,n,n -tetraacetic acid (EGTA)] or HEPES with 100 mol/l of ATP) and different temperatures (20 C or37 C), as listed in Table 1. Changes of solution were made using an eight-way solenoid valve (General Valve, Fairfield, NJ) with zero dead space. When used, physiological temperature was maintained at 37 C 1 C by means of a custom-made heating tube (Groupe de Recherche en Transport Membranaire). This step reproduced the WR step of a liver transplantation. Cell Necrosis Measurement To assess cell necrosis, SECs were treated with the vital dye propidium iodide (PI), 1.5 mol/l. This polar fluorochrome is unable to penetrate intact cell membranes and stains only DNA from necrotic cells. SECs first were isolated, plated, and preserved as described in Cell Isolation, Culture, and Preservation. Cells were subjected to experimental protocols A, B, and C (Table 1), with the difference that WR was mimicked by changing rewarming solutions (UW or HEPES at 20 C) with the appropriate buffer (HEPES with or without Ca 2 at 37 C) into which PI was added. Petri dishes then were placed in an incubator for 30 to 45 minutes at 37 C under a humid 95% oxygen and 5% carbon dioxide atmosphere before cell counts were performed. Cell counts were performed using an inverted fluorescent microscope (Nikon Eclipse TE-200). The red-orange PI cell coloration was observed under a 530- to 545-nm excitation filter and a 610- to 675-nm emission filter (HQ R/D11 cube). Image acquisition was assessed by a Hamamatsu Orca-II CCD digital camera and an Inovision computer controlled by a Isee software (Inovision Corp, Raleigh, NC). Between 100 and 200 cells were observed in different regions of each glass coverslip for each experimental group and for several SEC preparations. Data Analysis and Statistics Each experiment was performed with at least 10 to 15 cells in the microscopic field per experimental protocol and per SEC preparation. Values are expressed as mean SEM for the indicated number of glass coverslips obtained from at least seven different cell preparations and thus encompass averages of a much larger number of cells. Results were analyzed by two-way analysis of variance, except for the frequency of SEC Ca 2 responses to ATP, which was analyzed by Chi-squared test. P less than.05 is considered statistically significant.

4 Endothelial Cell Ca 2 in Cold Storge Injury 153 Figure 1. Morphological changes induced by CP in vitro in isolated rat liver SECs. (A) Isolated SECs adhered to collagen-s coated glass coverslips and allowed to recover for 8 hours in RPMI medium show the typical elongated shape of cultured endothelial cells. (Original magnification 400.) (B) After 12 hours in UW solution at 4 C, isolated SECs take on a more rounded appearance, reproducing morphological changes induced by cold storage of the whole organ. 4 (Original magnification 400.) Results Isolated SECs show a typical elongated shape when put in culture for 8 hours after the isolation procedure (Fig. 1A). Preservation of cells in cold UW solution for 12 hours causes a great proportion of the cells to adopt a more rounded shape (Fig. 1B). Such rounding up was seen previously in the whole organ 7 and confirms the validity of our in vitro model. Steady-state intracellular Ca 2 concentration before reperfusion with solutions of different composition and temperature was greater in SECs loaded with Fura- 2AM in HEPES buffer 20 C ( nmol/l; n 10 preparations) than in counterparts loaded in UW 20 C (37 2 nmol/l; n 44 preparations; P.05). These values were not affected by previous CP time (Table 2). Unstored controls (0 hour) loaded in UW 20 C and reperfused with HEPES 37 C showed a large transient increase in SEC Ca 2 levels (increase of nmol/l; n 16 preparations; Table 3, protocol A). This increase was greatly inhibited when a Ca 2 -free HEPES 37 C solution (4 mmol/l of EGTA) was used to reperfuse the preparations (increase of only nmol/l; n 11 preparations; Table 3, protocol B; P.05). Similarly, a very weak increment in cytosolic Ca 2 levels was observed when SECs were perfused with UW at 37 C, a medium also nominally exempt of Ca 2 (increase of 14 2 nmol/l; n 9 preparations; Table 3, protocol D; P.05). Temporal changes in cytosolic Ca 2 levels in these conditions are shown in Figure 2A (protocols A, B, and D). Changing only the temperature (20 C to 37 C) of HEPES buffer induced a rapid, transient, but significantly smaller, increase in cytosolic Ca 2 concentrations (60 18 nmol/l; n 9 preparations; Table 3) than that caused by the change in both solution and temperature; UW 20 C to HEPES 37 C buffer (P.05, protocol C versus A). Absolute peak Ca 2 values reached also were greater after protocol A ( nmol/l) than protocol C ( nmol/l), although Table 2. Basal Steady-State Intracellular SEC Ca 2 Levels in the Various Experimental Protocols Cold Preservation Time (hr) Experimental Protocol A UW (20 C) 3 HEPES (37 C) 37 4(n 16) 28 6(n 13) 33 7(n 8) B UW (20 C) 3 Ca 2 -free HEPES (37 C) 35 4(n 11) 23 4(n 15) 31 8(n 9) C HEPES (20 C) 3 HEPES (37 C) * (n 10) * (n 8) 98 19* (n 7) D UW (20 C) 3 UW (37 C) 33 3(n 9) 27 5(n 14) 41 8(n 8) E UW (20 C) 3 HEPES (20 C) 42 4(n 8) 29 6(n 11) 37 9(n 7) NOTE. Ca 2 values in nanomolar. Values expressed as mean SEM of the specified number of coverslips (n) taken from a minimum of seven cell preparations. SEC intracellular Ca 2 levels were measured after a Fura-2AM dye loading of 30 minutes in UW or HEPES 20 C, followed by one of the different indicated protocols (A to E) after 0, 12 or 24 hours of CP (4 C) in UW. Values in this table refer to the steady-state Ca 2 concentration observed immediately before the indicated solution and/or temperature change. *Significantly different from all other groups (P.05).

5 154 Auger, Vallerand, and Haddad Table 3. Peak Increase in SEC Ca 2 Levels Induced by Changes in Solution Composition and Temperature After CP-WR Cold Preservation Time (hr) Experimental Protocol A UW (20 C) 3 HEPES (37 C) (n 16) (n 13) (n 8) B UW (20 C) 3 Ca 2 -free HEPES (37 C) 43 13* (n 11) 47 10* (n 15) 90 12* (n 7) C HEPES (20 C) 3 HEPES (37 C) 60 18* (n 9) 68 25* (n 8) 46 16* (n 7) D UW (20 C) 3 UW (37 C) 14 2 (n 9) 6 2 (n 11) 1 7 (n 8) E UW (20 C) 3 HEPES (20 C) (n 8) (n 11) (n 7) NOTE. Ca 2 in nanomolar. Values expressed as mean SEM of specified number of coverslips (n) taken from a minimum of seven cell preparations. SEC intracellular Ca 2 levels were measured after a Fura-2AM dye loading of 30 minutes in UW or HEPES 20 C, followed by one of the different indicated protocols (A to E) after 0, 12 or 24 hours of CP (4 C) in UW. Peak increases were calculated as the difference between peak Ca 2 concentration and basal steady-state levels observed immediately before the solution and/or temperature change. *Significantly different from groups A and E (P.05). Significantly different from all other groups (P.05). the difference did not reach statistical significance. Conversely, changing solution composition (UW to HEPES) while keeping the temperature constant at 20 C elevated SEC intracellular Ca 2 levels by nmol/l (n 8 preparations; Table 3, protocol E), which was not statistically significantly different from UW 20 C to HEPES 37 C group (protocol A). Temporal changes in cytosolic Ca 2 levels in these conditions are shown in Figure 2 (protocols A, C, and E). CP for 12 and 24 hours before all solution composition and/or temperature changes did not significantly affect basal steady-state Ca 2 levels in SECs (Table 2). Also, previous cold storage in UW solution did not modify the response of SEC cytosolic Ca 2 to subsequent changes in reperfusion solution composition and/or temperature (Table 3, all protocols). Once SEC Ca 2 levels stabilized to a new steady state after solution and/or temperature changes (and always eventually back to HEPES 37 C buffer for a period of 5 to 10 minutes; Table 1), cells were challenged with the purinergic agonist ATP. Response was quantified by analyzing the area under the Ca 2 versus time curve. Figure 3 shows a representative response obtained in unstored cells subjected to experimental protocol C, which represents the control condition and is similar to responses obtained in cells loaded and maintained in cell culture medium (results not shown). We observed that when SECs were loaded with Fura-2AM dye in HEPES 20 C buffer and then reperfused at 37 C in the same solution (protocol C, control condition), the response was significantly greater ( mol/l; n 8 preparations; P.05) than under the other protocols (ranging on average from 0.6 to 0.9 mol/l; Table 4). In EGTA-chelated Ca 2 -free buffer, SECs were particularly perturbed, evidenced by the weak frequency or absence of response to the Ca 2 - mobilizing agonist. In contrast to results on steady-state SEC Ca 2 presented, previous CP significantly attenuated the intensity and frequency of appearance of SEC Ca 2 responses to ATP (P.01, Table 5). This was most prominent for groups preserved 24 hours in UW solution and reperfused with Ca 2 -free buffers (EGTA-containing HEPES) or UW. We also verified the extent of cell necrosis in selected protocols A, B, and C. When control unstored (0 hour) cells were switched from UW 20 C to HEPES 37 C (protocol A), 3.7% 1.2% (n 6) were necrotic as opposed to only 1.5% 0.2% (n 6; P.05, protocol A versus C) when only temperature was changed (HEPES 20 C to 37 C, protocol C). WR of unstored cells in EGTA-chelated Ca 2 -free buffer was associated with the greatest extent of necrosis (5.7% 1.5%; P.05, protocol B versus C). CP time increased the number of necrotic cells in protocol C (3.6% 0.6% at 24 hours; n 6; P 0.05 versus 0 hour), whereas that of protocols A and B remained stable (3.6% 1.3% and 4.4% 0.7% at 24 hours, respectively; P not significant for time). Discussion In this study, we measured intracellular Ca 2 concentrations in SECs after CP-WR, steps that mimic in vitro conditions the grafted liver is subjected to during the in vivo transplantation procedure. Our model appropriately reproduced morphological changes (rounding up) induced to SECs by cold storage, as observed previously

6 Endothelial Cell Ca 2 in Cold Storge Injury 155 Figure 2. Effects of changes in solution composition and/or temperature on steady-state intracellular Ca 2 levels in SECs. (A) SEC intracellular Ca 2 levels were measured after a Fura-2AM dye-loading period of 30 minutes in UW 20 C, followed by HEPES 37 C (curve A), Ca 2 -free HEPES 37 C (curve B), or UW 37 C (curve D). *Peak increase in SEC Ca 2 level for protocol A is significantly greater than those in protocols B andd(p <.05). (B) SEC intracellular Ca 2 levels were measured after a Fura-2AM dye-loading period of 30 minutes in UW 20 C, followed by HEPES 20 C (curve E). In curve C, cells were loaded with Fura- 2AM in HEPES 20 C and reperfused with HEPES 37 C. *Peak increase in SEC Ca 2 level for protocol C is significantly less than that for protocol E (P <.05). in the isolated perfused rat liver. 7 Conversely, we show for the first time that Ca 2 homeostasis in SECs is significantly perturbed after CP-WR. Steady-state cytosolic Ca 2 concentrations in SECs are affected by changes in composition and temperature of the reperfusion solution, as is the subsequent response to a Ca 2 - mobilizing agonist. The greatest increase in basal steady-state SEC cytosolic Ca 2 levels in response to solution changes was observed when UW 20 C was changed to physiological HEPES 37 C (Table 3). This large transient increase was caused in major part by an influx of Ca 2 from the external milieu. It was greatly inhibited by chelation of extracellular Ca 2 with EGTA and was absent when warm UW (a nominally Ca 2 -free solution) was used. The small increase in SEC cytosolic Ca 2 levels that was nonetheless observed after WR in Ca 2 -free HEPES suggests that a small degree of release from intracellular stores may occur, but further studies are necessary to address this point. Moreover, after a 30-minute dye-loading period in UW solution, we found that basal SEC cytosolic Ca 2 levels were less than 40 nmol/l compared with more than 100 nmol/l when physiological HEPES buffer was used (the latter concentration corresponding to the normal baseline steady-state SEC Ca 2 concentration 24 ). This suggests that SECs can readily lose Ca 2 when kept in low-ca 2 solutions. This also may lead to exhaustion of intracellular Ca 2 stores, discussed further below. On the other hand, our results clearly show that changes in composition of the perfusing solution are more important than changes in temperature to medi-

7 156 Auger, Vallerand, and Haddad Figure 3. Response of SEC Ca 2 levels to purinergic agonist ATP after CP-WR. SEC intracellular Ca 2 responses to agonist ATP were measured once SEC cytosolic Ca 2 stabilized to a new steady state after a Fura-2AM dye loading of 30 minutes in UW or HEPES 20 C, followed by different established protocols (A to E; Table 1) in cells previously preserved for 0, 12, or 24 hours in cold (4 C) UW. The curve is representative of responses obtained in experimental protocol C in unstored cells, which represents the control condition. ate the initial transient increase in SEC cytosolic Ca 2 levels. Even when the solution change from UW to HEPES was performed at room temperature (20 C), SEC cytosolic Ca 2 levels increased by an extent very similar to that seen with HEPES at physiological temperature. Conversely, when perfusion temperature was increased from 20 C to 37 C in physiological HEPES buffer, a much smaller increase in SEC cytosolic Ca 2 levels was observed. Although smaller, the absolute peak in SEC cytosolic Ca 2 levels after this protocol (Table 1, protocol C) was not statistically significantly different from that observed after protocol A. As discussed next, our results indicate that the amplitude of initial Ca 2 transient is principally deleterious to SECs, rather than the absolute level of cytosolic Ca 2 reached. Our results are relevant to the clinical setting because UW solution has an intracellular-like composition (high potassium/low sodium) and thus must be flushed out of the liver before reestablishment of blood flow in the recipient. In our studies, the fluorochrome Fura- 2AM was loaded into cells at room temperature because loading at 4 C was found to be impossible. However, this obligatory step in our experimental procedure mimics the rewarming phase of liver transplantation, in which the liver graft warms up while surgical reanastomosis is underway. Table 4. Intensity of SEC Ca 2 Response to Purinergic Agonist ATP After CP-WR Cold Preservation Time (hr) Experimental Protocol A UW (20 C) 3 HEPES (37 C) (n 15) (n 13) (n 8) B UW (20 C) 3 Ca 2 -free HEPES (37 C) (n 11) NR (n 15) NR* (n 7) C HEPES (20 C) 3 HEPES (37 C) (n 8) (n 8) * (n 7) D UW (20 C) 3 UW (37 C) (n 9) (n 11) NR* (n 8) E UW (20 C) 3 HEPES (20 C) (n 8) (n 11) NR* (n 7) NOTE. Ca 2 in micromolar. Values expressed as mean SEM of the specified number of coverslips (n) taken from a minimum of seven cell preparations. When less than three coverslips responded, values obtained are listed separated by a slash. The SEC intensity of Ca 2 responses to purinergic agonist ATP were measured after a Fura-2AM dye loading of 30 minutes in UW or HEPES 20 C, followed by one of the different indicated protocols (A to E) after 0, 12, or 24 hours of CP (4 C) in UW. ATP stimulation (1 minute) was applied once SEC Ca 2 levels stabilized for 10 minutes in HEPES 37 C to a new steady state after solution and/or temperature changes. The intensity of Ca 2 response to ATP was measured by the area under the Ca 2 versus time curve. Abbreviation: NR, no response. *Significantly different from all other preservation times within an experimental protocol (P.05). Significantly different from all other groups within a given CP time (P.05).

8 Endothelial Cell Ca 2 in Cold Storge Injury 157 Table 5. Frequency of SEC Ca 2 Response to Agonist ATP After CP-WR Cold Preservation Time (hr) Experimental Protocol P Time A UW (20 C) 3 HEPES (37 C) 80 (n 15) 54 (n 13) 13 (n 8).01 B UW (20 C) 3 Ca 2 -free HEPES (37 C) 18 (n 11) 0 (n 15) 0 (n 7) NS C HEPES (20 C) 3 HEPES (37 C) 100 (n 8) 63 (n 8) 57 (n 7) NS D UW (20 C) 3 UW (37 C) 78 (n 9) 18 (n 11) 0 (n 8).01 E UW (20 C) 3 HEPES (20 C) 100 (n 8) 64 (n 11) 0 (n 7).01 P Protocol NOTE. Values represent the percent of responses observed in total number of coverslips (n) taken from a minimum of seven cell preparations for each experimental protocol and at each CP time. Statistical significance was evaluated by -squared analysis for the various experimental protocols (P Protocol ) and preservation times (P Time ). Abbreviation: NS, not significant. Our results indicate that rewarming (dye-loading step) in a physiological solution is preferable to that in UW solution. This again is relevant because there is still debate about whether it is better to rinse the graft with UW at room temperature or use solutions of physiological composition. 29 In particular, the group at the University of North Carolina has developed a Carolina rinse solution that appears to improve initial graft function Our physiological HEPES buffer approximates this condition. It is also in HEPES physiological buffer that responses to such Ca 2 -mobilizing agonists as ATP were best preserved. Our results clearly show that SECs are particularly sensitive to low or absent external Ca 2. Responses to ATP were dramatically reduced in frequency and intensity after reperfusion with Ca 2 -free HEPES. Previous CP greatly exacerbated this effect, and reperfusion with low-ca 2 solutions (UW, Ca 2 - free HEPES) abolished ATP responses in 24 hour preserved cells (Tables 4 and 5). This may be related to the exhaustion of intracellular Ca 2 pools mentioned. Gq-coupled agonist responses in SECs, as in other cells, rely on an initial mobilization of Ca 2 from intracellular stores. 33 Cells subjected to a change in solution composition in normal external Ca 2 conditions also were affected negatively by CP in a time-dependent manner. Conversely, cells loaded in HEPES 20 C and reperfused in HEPES 37 C (protocol C) maintained the best responses to the Ca 2 -mobilizing agonist ATP. Our results therefore strongly suggest that using a physiological buffer to flush cold UW and rewarm the liver may be the best way to maintain a normal steady-state Ca 2 level, minimize initial cytosolic Ca 2 transients on physiological reperfusion, and preserve appropriate Ca 2 signaling in SECs after such reperfusion. In our opinion, these conditions should help maintain better functionality of cells. In support of this notion, our experiments also showed that SEC necrosis (assessed by PI staining) under experimental protocol C was one half times less important than that observed under protocol A (UW 20 C to HEPES 37 C). Conversely, reperfusion with Ca 2 -free HEPES at 37 C yielded nearly four times more necrosis than control condition C. These data thus confirm the deleterious effects of extracellular Ca 2 removal for SEC integrity. They also confirm that rewarming and reperfusion in physiological buffer is best suited to minimize SEC necrosis. It must be noted that absolute values of necrosis obtained with PI staining were low. However, as with other vital dyes, such as Trypan blue, PI staining will not allow the measurement of detached cells and thus will underestimate the total number of necrotic cells. Nonetheless, we are confident that proportions of PI-stained cells obtained under the different protocols accurately reflect the overall extent of necrotic cell death in each case. Our results are consistent with the recent preliminary observations from Strasberg et al. 34,35 These investigators showed that inhibiting changes in SEC intracellular Ca 2 levels with BAPTA or dantrolene prevented an increase in calpain activity and the rearrangement of actin after CP. Calpains are nonlysosomal cytoplasmic Ca 2 -dependant cysteine proteases involved in proteolysis of such cytoskeletal elements as actin fibers, as well as membrane proteins. 36 Upadhya and Strasberg 37 previously had shown that cold-preserved SECs secrete matrix metalloprotease, and this phenomenon requires actin disassembly. This group recently confirmed that calpain activation in cold-pre-

9 158 Auger, Vallerand, and Haddad served SECs results in actin disassembly and matrix metalloprotease secretion. 35 These events also provoke the rounding up and detachment of SECs from the extracellular matrix and eventually their apoptosis, a situation that can be prevented by the use of calpain inhibitors Our observation that cytosolic Ca 2 levels increase after reperfusion also can explain the relatively frequent microcirculatory problems in transplantation injury. Such an increase in Ca 2 levels can induce the expression or translocation of such inflammatory/adhesive mediators as platelet-activating factor or selectins on the SEC surface. 1,9,13 These mediators probably participate in the inflammatory process observed at the level of sinusoids in the grafted organ. This can lead to occlusion of microvessels, areas of no-flow, and, ultimately, graft failure. 16,18-20 Upadhya and Strasberg 38 recently confirmed that CP and WR lead to P-selectin expression and platelet adhesion in an in vitro model of SECs very similar to ours. Hence, conditions that can minimize the elevation in SEC cytosolic Ca 2 levels may reduce these local inflammatory reactions. However, future studies need to ascertain the cause-effect relationship between cytosolic Ca 2 level changes and expression of adhesion molecules on the surface of SECs. In conclusion, our results clearly show that steadystate intracellular SEC Ca 2 levels are affected by changes in composition and temperature of reperfusion solutions, with changes in solution composition having the greatest impact. Previous CP does not affect steadystate cytosolic SEC Ca 2 level or its response to solution changes, but greatly perturbs response to such Ca 2 -mobilizing agonists as ATP. Such responses are best preserved when SECs are rewarmed using a physiological buffer and then reperfused at 37 C with the same. Our study therefore supports that it is better to rinse the liver graft with a physiological solution at room temperature before reestablishing oxygenated blood flow in the recipient. Our studies also pave the way to test other solutions and/or modifications to minimize initial fluctuations in SEC steady-state Ca 2 levels and maintain agonist responses after CP-WR to improve initial graft function. Acknowledgment The authors thank Filip Braet from the Free University of Brussels, Belgium, for precious help and counseling with the SEC isolation technique. References 1. Clavien P-A, Harvey PRC, Strasberg SM. Preservation and reperfusion injuries in liver allografts. Transplantation 1992;53: Serracino-Inglott F, Habib NA, Mathie RT. Hepatic ischemiareperfusion injury. Am J Surg 2001;181: Jaeschke H. Preservation injury: Mechanisms, prevention and consequences. J Hepatol 1996;25: McKeown CMB, Edwards V, Phillips MJ, Harvey RC, Petrunka CN, Strasberg SM. Sinusoïdal lining cell damage: The critical injury in cold preservation of liver allografts in the rat. Transplantation 1988;46: Clavien P-A. Sinusoidal endothelial cell injury during hepatic preservation and reperfusion. Hepatology 1998;28: Caldwell-Kenkel JC, Currin RT, Tanaka S, Thurman RG. Reperfusion injury to endothelial cells following cold ischemic storage of rat liver. Hepatology 1989;10: Caldwell-Kenkel JC, Currin RT, Tanaka Y, Thurman RG, Lemasters JJ. Kupffer cell activation and endothelial cell damage after storage of rat livers: Effects of reperfusion. Hepatology 1991;13: Eberl T, Salvenmoser W, Rieger G, Gorny I, Hei V, Kumpitsch B, et al. Ultrastructural analysis of human endothelial cells after hypothermic storage in organ preservation solutions. J Surg Res 1999;82: Whatley RE, Nelson P, Zimmerman GA, Stevens DL, Parker CJ, McIntyre TM, Prescott SM. The regulation of platelet-activating factor production in endothelial cells. J Biol Chem 1989; 264: Cheung J, Bonventre J, Malis C, Leaf A. Calcium and ischemic injury. N Engl J Med 1986;314: Ernster L. Biochemistry of reoxygenation injury. Crit Care Med 1988;16: Lewis MS, Whatley RE, Cain P, McIntyre TM, Prescott SM, Zimmerman GA. Hydrogen peroxide stimulates the synthesis of platelet-activating factor by endothelium and induces endothelial cell-dependent neutrophil adhesion. J Clin Invest 1988;82: McEver RP, Beckstead JH, Moore KL, Marshall-Carlson L, Bainton DF. GMP-140, a platelet alpha-granule membrane protein, is also synthesized by vascular endothelial cells and is localized in Weibel-Palade bodies. J Clin Invest 1989;84: Yadav SS, Howell DN, Steeber DA, Harland RC, Tedder TF, Clavien P-A. P-Selectin mediates reperfusion injury through neutrophil and platelet sequestration on the warm ischemic mouse liver. Hepatology 1999;29: Geng JG, Bevilacqua MP, Moore KL, McIntyre TM, Prescott SM, Kim JM, et al. Rapid neutrophil adhesion to activated endothelium mediated by GMP-140. Nature 1990;343: Vollmar B, Menger MD, Glasz J, Leiderer R, Messmer K. Impact of leukocyte-endothelial interaction in hepatic ischemiareperfusion injury. Am J Physiol 1994;267:G786-G Walzog B, Gaehtgens P. Adhesion molecules: The path to a new understanding of acute inflammation. News Physiol Sci 2000; 15: Vollmar B, Richter S, Menger MD. Leukocyte stasis in hepatic sinusoids. Am J Physiol 1996;270:G798-G Engler RL, Schmid-Schonbein GW, Pavelec RS. Leukocyte capillary plugging in myocardial ischemia and reperfusion in the dog. Am J Pathol 2002;111:

10 Endothelial Cell Ca 2 in Cold Storge Injury Simpson PJ, Todd RF3, Fantone JC, Mickelson JK, Griffin JD, Lucchesi BR. Reduction of experimental canine myocardial reperfusion injury by a monoclonal antibody (anti-mo1, anti- CD11b) that inhibits leukocyte adhesion. J Clin Invest 1988;81: Cheng S, Ragsdale JR, Sasaki AW, Lee RG, Deveney CW, Pinson CW. Verapamil improves rat hepatic preservation with UW solution. J Surg Res 1991;50: Takei Y, Marzi I, Kauffman FC, Currin RT, Lemasters JJ, Thurman RG. Increase in survival time of liver transplants by protease inhibitors and a calcium channel blocker, nisoldipine. Transplantation 1990;50: de Broin E, Urata K, Giroux L, Lepage R, Huet P-M. Effect of calcium antagonists on rat liver during extended cold preservation-reperfusion. Transplantation 1997;63: Haddad P, Cabrillac J-C, Piche D, Musallam 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: Kim JS, Southard JH. Alterations in cellular calcium and mitochondrial functions in the rat liver during cold preservation. Transplantation 1998;65: Elimadi A, Haddad PS. Cold preservation-warm reoxygenation increases hepatocyte steady-state Ca 2 and response to Ca 2 - mobilizing agonist. Am J Physiol Gastrointest Liver Physiol 2001;281:G809-G Braet F, De Zanger R, Sasaoki T, Baekeland M, Janssens P, Smedsrod B, Wisse E. Methods in laboratory investigation Assessment of a method of isolation, purification, and cultivation of rat liver sinusoidal endothelial cells. Lab Invest 1994;70: Benzeroual K, Van de Werve G, Meloche S, Mathé L, Romanelli A, Haddad P. Insulin induces Ca 2 influx into isolated rat hepatocytes couplets. Am J Physiol Gastrointest Liver Physiol 1997; 272:G1425-G Kukan M, Haddad PS. Role of hepatocytes and bile duct cells in preservation-reperfusion injury of liver grafts. Liver Transpl 2001;7: Currin RT, Toole JG, Thurman RG, Lemasters JJ. Evidence that Carolina rinse solution protects sinusoidal endothelial cells against reperfusion injury after cold ischemic storage of rat liver. Transplantation 1990;50: Sanchez-Urdazpal L, Gores GJ, Lemasters JJ, Thurman RG, Steers JL, Wahlstrom HE, et al. Carolina rinse solution decreases liver injury during clinical liver transplantation. Transplant Proc 1993;25: Post S, Palma P, Rentsch M, Gonzalez AP, Menger MD. Differential impact of Carolina rinse and University of Wisconsin solutions on microcirculation, leukocyte adhesion, Kuppfer cell activity and biliary excretion after liver transplantation. Hepatology 1993;18: Schilling WP, Elliott SJ. Ca 2 signaling mechanisms of vascular endothelial cells and their role in oxidant-induced endothelial cell dysfunction. Am J Physiol Heart Circ Physiol 1992;262: H1617-H Upadhya GA, Strasberg SM. BAPTA-AM and dantrolene reduce actin disassembly in cold preserved rat sinusoidal endothelial cells (SEC): A novel preservation strategy. Hepatology 2000;32: 248A. 35. Upadhya GA, Topp SA, and Strasberg SM. Stimulation of calpain activity: A mechanism by which increased intracellular calcium in sinusoidal endothelial cells stimulates actin disassembly and MMP secretion and subsequent SEC surface activation. Hepatology 2001;34:234A. 36. Sindram D, Kohli V, Madden JF, Clavien P-A. Calpain inhibition prevents sinusoidal endothelial cell apoptosis in the cold ischemic rat liver. Transplantation 1999;68: Upadhya GA, Strasberg SM. Evidence that actin disassembly is a requirement for matrix metalloproteinase secretion by sinusoidal endothelial cells during cold preservation in the rat. Hepatology 1999;30: Upadhya GA, Strasberg SM. Platelet adherence from isolated hepatic sinusoidal endothelial cells after cold preservation. Transplantation 2002;73: