Preservation of Pancreas Tissue During Cold Storage Assessed by X-Ray Microanalysis

Authors


* Corresponding author: Godfried M. Roomans,godfried.roomans@medcellbiol.uu.se

Abstract

Clinical transplantation requires cold storage of tissue for several hours. We have examined the elemental content in exocrine and endocrine cells in mouse pancreas after cold storage by X-ray microanalysis, and in parallel carried out morphological studies. Tissue was stored at 4 °C for 4–12 h in Normal Krebs-Ringer's (high Na+/K+ ratio), Modified Krebs-Ringer's (low Na+/K+ ratio), Euro-Collins, University of Wisconsin (UW) solution, and seven modified version of UW solution. Incubation in Normal Krebs-Ringer's solution caused significantly increased Na and decreased K concentrations in contrast to incubation in other solutions. The cellular concentration of Na and Cl followed the concentration in the storage solution. Changes in the endocrine cells were similar to, but less pronounced than those in exocrine cells. Calcium was retained best in UW and some variants of UW, and least in Euro-Collins. This may indicate differences in preservation of secretory granules. Also, morphological studies showed that endocrine cells were less affected than exocrine cells. In conclusion, the only factor determining the intracellular concentration of diffusible ions after cold tissue storage is the ionic composition of the extracellular medium. X-ray microanalysis provides an objective method to assess whether the intracellular ionic composition of tissue is maintained during storage.

Introduction

A persistent problem in organ transplantation is the tissue damage that occurs during organ preservation. University of Wisconsin solution (UW) is considered one of the most effective cold storage solutions for the pancreas (1,2). UW solution was first developed experimentally for pancreas transplantation (3,4) and only later applied to the preservation of liver (5–7), kidney (8), heart (9,10) and lung (11). The solution has had an enormous impact on the practice of organ preservation for transplantation and has in many countries become the standard preservation solution in clinical use for all intra-abdominal organs. Several attempts have been made to create alternative solutions for preservation of tissue for transplantation. For cold preservation of the porcine pancreas the histidine-tryptophan-ketoglutarate (HTK) solution has been used (12). Los Angeles preservation solution 1 (LAP-1) was developed as cold preservation solution for isolation of high-quality human pancreatic islets (13). Evaluation of the storage solutions has been carried out using morphological techniques (13), studies of cell or islet viability (14,15), functional studies such as glucose tolerance tests (12,16), tissue ATP levels (17), or adenylyl cyclase activity (18), as well as by clinical assessment (19). However, no studies determining changes in ion content during storage have been carried out. In previous studies on other organs we have shown that storage of tissue in experimental solutions may result in significant changes in ion content (20,21). In the present study we propose X-ray microanalysis as an objective method for evaluation of the effect of storage solutions on the elemental content of cells in tissue for transplantation. This method allows the determination of elemental concentrations at the cell level. The present study was designed to compare the UW solution with seven different modifications of the UW solution, Normal Krebs-Ringer's and Modified Krebs-Ringer's solution (20), or Euro-Collins solution (22), in order to assess which components were critical for maintenance of the in vivo ion content of the cells.

In these experiments it was tested whether: (a) hydroxyethyl starch could be replaced by dextran, or even omitted; (b) the free radical scavenger glutathione could be replaced by N-acetyl cysteine; (c) raffinose and adenosine as energy sources could be replaced by glucose; and (d) Na+ as chloride or bicarbonate salt should be added to the UW solution to increase its Na+ concentration. Modified Krebs-Ringer's solution, Euro-Collins and the seven modification solutions based on the commercial UW solution have a high K+/Na+ ratio (while the commercial UW solution does not contain any Na+ at all).

Materials and Methods

Tissue

Forty-four mice of both sexes, 4–5 weeks old, were used in the experiment. The animals were anaesthetized with pentobarbital. The pancreas was removed from all animals and dissected into 1.5–2-mm2-thick pieces. To determine the in situ elemental composition of the cells, some of the pieces were frozen immediately in liquid propane cooled by liquid nitrogen, and the remaining pieces were incubated in experimental solutions. The dissection of the tissue took less than 1 min (from removing the organ from the animals to the start of the incubation). The protocol was approved by the Regional Committee on Animal Experimentation for Uppsala County.

Tissue incubation

Dissected slices of pancreas were incubated in 11 different experimental solutions: commercial University of Wisconsin (UW) solution (ViaSpan®) (23) and seven modified versions of the UW solution, Normal Krebs-Ringer's solution and Modified Krebs-Ringer's solution (20), or Euro-Collins solution [prepared according to the composition given in (3)]. The composition of experimental solutions used is given in Table 1. The pH of all solutions was adjusted to 7.4 by addition of NaOH (at room temperature). The tissue was incubated for 4 h, 8 h and 12 h at 4 °C (by cooling with ice on a shaking board) and oxygenated with 95% O2 and 5% CO2. Tissue samples were removed from the solution at the end of the incubation and frozen in liquid propane cooled with liquid nitrogen.

Table 1.  : Composition (mmol/L) of experimental solutions for incubation of the pancreas (pH 7.4)
 NormalModified         
ComponentsKrebs-
Ringer's
Krebs-
Ringer's
Euro-CollinsUW (com)UW (A)UW (B)UW (C)UW (D)UW (E)UW (F)UW (G)
  1. HEStarch = hydroxyethyl starch. Modified Krebs-Ringer's solution = reversal of the Na/K ratio in Normal Krebs-Ringer's solution. UW (com) = University of Wisconsin solution; UW (A) = modified University of Wisconsin solution with 5% dextran; UW (B) = modified University of Wisconsin solution without dextran; UW (C) = modified University of Wisconsin solution with 5% dextran + N-acetylcysteine; UW (D) = modified University of Wisconsin solution with 5% dextran + glucose instead of raffinose and adenosine; UW (E) = modified University of Wisconsin solution without dextran + glucose instead of raffinose and adenosine; UW (F) = modified University of Wisconsin solution with 5% dextran + NaCl; UW (G) = modified University of Wisconsin solution with 5% dextran + NaHCO3.

Lactobionate   100       
KOH
Adenosine   555555
Allopurinol   1       
KH2PO4  152525252525252515
K2HPO4  43        
MgSO4   55555555
Raffinose   303030303030
Glutathione   3333333
HEStarch   50 g/L       
NaHCO3  10       10
CaCl21.51.5         
NaCl14026       10 
KCl1415        
Glucose55180    3535  
MgCl211         
K-gluconate5105  10010010010010090100
HEPES55         
N-acetylcysteine      3    
Dextran    5%5%5%5%5%
Insulin   40 U/L       
Dexamethasone   16 g/L       
Penicillin   200 000       

X-ray microanalysis

For X-ray microanalysis 16-μm-thick cryosections were cut on a conventional cryostat at − 30 °C, mounted on a carbon plate, freeze-dried inside the cryostat for 3 days and gradually brought to room temperature during a period of 3 h. The sections were coated with a thin conductive carbon layer to prevent charging in the electron microscope (24,25). The first 5–10 sections were cut from the dissected edge of the tissue block and discarded, because the edge might have been damaged. The sections were analyzed in a Philips 525 scanning electron microscope (Philips Electron Optics, Eindhoven, the Netherlands) at 20 kV with a LINK AN10000 energy-dispersive spectrometer system (Oxford Instruments, Oxford, UK). All analyses were carried out with a stationary beam (probe size 100 nm). Due to scattering of the electrons within the sample, the actual spatial resolution of analysis is in the order of several μm, i.e. at the cell level. In the sections of the pancreas both the exocrine part and the endocrine part were investigated (Figure 1). The following elements were determined: sodium (Na), magnesium (Mg), phosphorus (P), sulfur (S), chloride (Cl), potassium (K), and calcium (Ca). Twelve measurements of different cells were made on each sample. Quantitative analysis was carried out using the ratio of characteristic to continuum intensity and by comparing this ratio with that obtained by analysis of a standard with known elemental concentrations (26).

Figure 1.

Scanning electron micrograph of a freeze-dried 16-μm-thick cryosection of the pancreas used for X-ray microanalysis. Bar = 0.1 mm.

Morphological studies

Morphological studies were carried out on tissue fixed by immersion in fixative (2.5% glutaraldehyde in 0.1 m Na-cacodylate buffer pH 7.4). As equivalent of tissue that was frozen immediately, tissue was excised and fixed by immersion in fixative without prior incubation. Incubated tissue slices were fixed by immersion in the same fixative. For light microscopy, tissue was dehydrated in a graded ethanol series, and embedded in JB4 plastic (Polysciences, Warrington, PA, USA). Semi-thin sections (2 μm thick) were cut and stained with hematoxylin/eosin, and viewed in a light microscope at 400 × magnification. For electron microscopy, tissue was postfixed in 1% osmium tetroxide, dehydrated in a graded ethanol series and embedded in Agar 100 epoxy resin (Agar Scientific, Stansted, UK). Ultra-thin sections were placed on copper slot grids covered with a Formvar film, contrasted with uranyl acetate and lead citrate, and viewed at 75 kV in a Hitachi (Tokyo, Japan) 7100 transmission electron microscope (TEM).

Statistics

The statistical significance of differences between groups was determined by analysis of variance (anova), followed by Dunnett's test: comparing all experimental solutions to the control. In a similar way, the modifications of the UW solution were compared to the commercially available UW solution.

Results

Effect of the incubation medium on the ionic composition of the tissue

In the first series of experiments, tissue slices were incubated in experimental solutions with different ionic content (Table 1) at 4 °C for 4 h. In the exocrine pancreas, cellular Na was maintained close to the in situ level when the tissue was incubated in Modified Krebs-Ringer's solution, University of Wisconsin solution and five simplified versions of this solution, but increased significantly in Normal Krebs-Ringer's solution and two modifications of the University of Wisconsin solution. The Na content slightly decreased in Euro-Collins solution (Figure 2). The K concentration in exocrine pancreatic cells was retained most closely to the in situ content when the tissue was kept in Modified Krebs-Ringer's solution, Euro-Collins solution, UW solution and seven simplified versions of the UW commercial solution, but decreased significantly in Normal Krebs-Ringer's solution. Increase in Na and/or loss of K indicates damage to the cell. The concentration of Cl in the exocrine part of the pancreas was closer to the in situ value after incubation in Modified Krebs-Ringer's solution than after incubation in Euro-Collins solution, UW solution and all its modified versions. The calcium concentrations were increased in variants (A) and (B) of the UW solution. A comparison of the commercial UW solution with the seven modified versions of this solution showed no significant differences in P, S, and K concentrations (data for P and S not shown). The concentration of Na slightly increased in two simplified variants (F) and (G). The concentration of Cl significantly decreased in most of the variants of the UW solution. The Ca concentrations were significantly increased in variants (A), (B) and (D) of the UW solution, compared to the commercial UW solution.

Figure 2.

Effect of the ionic concentration in the incubation experimental solutions on the elemental contents of the exocrine part of the pancreas incubated at 4 °C for 4 h. Data in (mmol/kg dry weight) are given as mean and standard error (bars) of 4–8 animals in each group (10–12 measurements per animal). Significant differences between treated tissue and control are indicated by asterisks (* p < 0.05: *** p < < 0.001). Significant differences between variants (A)–(G) of the University of Wisconsin solution and the commercial UW solution are indicated by # p < 0.05; and ## p < 0.01.

In the endocrine part of the pancreas (different cell types could not be distinguished), we observed an increase in Na and a decrease in K in Normal Krebs-Ringer's solution (Figure 3). The Cl concentration was significantly lower than the in situ value in all experimental solutions, but it was closest to the in situ value in Normal and Modified Krebs-Ringer's solutions. In the endocrine pancreas cells, cellular Ca decreased significantly in Modified Krebs-Ringer's solution, Euro-Collins and five modified versions of UW incubation solutions.

Figure 3.

Effect of the ionic concentration in the incubation experimental solutions on the elemental contents of the endocrine part of the pancreas incubated at 4 °C for 4 h. Data in (mmol/kg dry weight) are given as mean and standard error (bars) of 4–8 animals in each group (10–12 measurements per animal). Significant differences between treated tissue and control are indicated by asterisks (* p < 0.05; *** p < 0.001). Significant differences between variants (A)–(G) of the University of Wisconsin solution and the commercial UW solution are indicated by # p < 0.05; and ## p < 0.01.

Cellular Mg, S, P concentrations were not affected in any of the solutions, except that values for S were lower in the Euro-Collins solution; this was valid for both the exocrine and the endocrine part (Figures 2 and 3) (data for P and S not shown). A comparison of the commercial UW solution with seven modified versions of this solution did not show any significant differences for the Na, P, S and K concentrations. The Cl concentrations were significantly decreased in all modified variants. The calcium concentrations slightly decreased in modified versions (C), (E), (F) and (G), but in variants (A), (B) and (D) the Ca values were close to that obtained for the commercially available UW solution.

In the second series of experiments, the incubation period at 4 °C was prolonged to 12 h. In the exocrine pancreas, P and K concentrations did not change significantly in commercial UW solution, Euro-Collins and Modified Krebs-Ringer's solutions (Figure 4). Na and Cl decreased in all solutions but stayed closest to the in situ value in Modified Krebs-Ringer's solution. In the endocrine pancreas, P and K concentrations did not change significantly, Cl decreased except in the Modified Krebs-Ringer's solution, and Na decreased in all three solutions (Figure 4).

Figure 4.

Effect of preservation time on elemental content of the exocrine(exo) part and endocrine(endo) part of the pancreas incubated at 4 °C for 4 h, 8 h and 12 h. Data in (mmol/kg dry weight) are given as mean and standard error (bars) of 4 animals in each group (10 measurements per animal). The 0 h-value is the value for control (in situ) tissue.

Morphological studies

Light microscopy showed in control tissue tightly packed exocrine acini and islets of Langerhans (Figure 5A). In tissue incubated for 12 h in Normal Krebs-Ringer's solution (Figure 5B) or in Euro-Collins solutions (Figure 5D) large spaces are present between the acini and between the exocrine and endocrine parts of the tissue. These spaces are also present in tissue incubated in Modified Krebs-Ringer's solution (Figure 5C) or in commercial University of Wisconsin solution (Figure 5E), but they are much smaller, indicating less damage to the tissue. The number of nuclei in a fixed area (130 × 170 μm2), counted in the light microscope, was independent of the solution in which the tissue had been stored, both for the exocrine and endocrine parts of the pancreas, indicating that little swelling of the tissue had occurred.

Figure 5.

Light microscopic images of mouse pancreatic tissue showing both the exocrine part(acini, ac) and the endocrine part(islet of Langerhans, L) after 12 h of cold storage. (A) Control tissue (fixed immediately without any incubation). (B) Tissue stored in Normal Krebs-Ringer's solution. (C) Tissue stored in Modified Krebs-Ringer's solution ( V = blood vessel). (D) Tissue stored in Euro-Collins solution. (E) Tissue stored in University of Wisconsin solution. Arrows indicate intercellular spaces. Bar = 50 μm (Magnification, × 330).

By transmission electron microscopy, morphological changes in the incubated tissue could be observed more clearly. In control tissue, cells are intact, endoplasmic reticulum (ER) and mitochondria are not swollen, and the cells are tightly packed (Figure 6A,B). In tissue incubated for 12 h in Normal Krebs-Ringer's solution, the ER is clearly swollen and the mitochondria are enlarged (Figure 6C); in the endocrine cells the granules are no longer intact (Figure 6D). These changes are already visible after 4 h incubation (not shown). After storage in Modified Krebs-Ringer's solution (Figure 6E,F), Euro-Collins solution (Figure 6G,H), UW solution (Figure 6J,K) and seven simplified versions of UW commercial solution (not shown), the swelling of the endoplasmic reticulum was less than after incubation in Normal Krebs-Ringer's solution. Damage to the mitochondria was particularly evident in tissue incubated in Euro-Collins solution (Figure 6G). The mitochondria had lost most of their cristae. In tissue incubated in Modified Krebs-Ringer's solution and in UW, the mitochondria were somewhat swollen, but better preserved than in Euro-Collins solution or in Normal Krebs-Ringer's buffer. In general, damage appeared less in the endocrine cells, compared to the exocrine cells. The swelling of the endoplasmic reticulum induced by incubation was less in the cells of the endocrine part than in the cells of the exocrine part of the pancreas. The mitochondria in the cells of the endocrine part were more abundant and less damaged than those in the exocrine part in all experimental solutions. Tissue incubated for 4 h or 8 h showed tissue damage at an intermediate stage (not shown).

Figure 6.

Transmission electron micrographs of mouse pancreatic tissue(left) exocrine part and(right) endocrine part after 12 h of cold storage. (A,B) Control tissue (fixed immediately without any incubation). (C,D) Tissue stored in Normal Krebs-Ringer's solution. (E,F) Tissue stored in Modified Krebs-Ringer's solution. (G,H) Tissue stored in Euro-Collins solution. (J,K) Tissue stored in University of Wisconsin solution. ER: endoplasmic reticulum, N: nucleus, z: zymogen granules, arrows: mitochondria; a: α-cell, b: β-cell, arrowhead: endocrine granules in β-cells. Bar = 2 μm (Magnification, × 5300).

Discussion

The original UW solution was developed primarily for pancreas and has been shown to effectively preserve the pancreas (1,2,4), kidney (8), liver (5,6), heart (9,10) and lung (11) in the dog and rat. Subsequently, it has been widely applied in clinical organ transplantation. The original experimental work demonstrated consistent adequate preservation during 72 h (4) and later findings reported successful 48 h preservation of the pancreas using UW and some simplified variants (27). In the present study, we have used in vitro assays to assess the effect of 11 cold storage solutions on pancreas tissue stored at 4 °C for 4 h, and evaluate the result of UW commercial, Modified Krebs-Ringer's and Euro-Collins cold preservation solutions on pancreas tissue stored at 4 °C for up to 12 h.

During 4 h of cold storage, the K+ concentration in pancreas cells was retained most closely to the in situ value when the tissue was kept in Modified Krebs-Ringer's solution, Euro-Collins solution, University of Wisconsin commercial solution, and all seven modified versions of the last solution, but decreased significantly in Normal Krebs-Ringer's solution. After incubation of tissue in UW, Modified and Euro-Collins solutions during 8 h and 12 h the K+ concentration in pancreas cells was slightly, but not significantly, increased. The Na+ concentration remained low in all experimental solutions, with the exception of the Normal Krebs-Ringer's solution, which has a high Na+/K+ ratio characteristic of the extracellular fluid. In exocrine pancreas, modified solutions (F) and (G) lead to higher intracellular sodium than commercial UW solution. This can be explained by the fact that solution (F) contains 10 mm NaCl and 90 mm K-gluconate instead of 100 mm K-gluconate, and solution (G) contains 10 mm NaHCO3 and 15 mm KH2PO4 instead of 25 mm KH2PO4.

Cell volume and condition is directly related to the movement of ions, with homeostasis being achieved by a balance of osmotic pressure across the plasma membrane. When the concentration of solute particles on each side of the membrane is equal, a net movement of water is inhibited, thus maintaining a constant cell size. Most cells achieve and maintain this osmotic balance through the continuous activity of the Na+/K+ ATPase pump, which creates and maintains an intracellular environment high in K+ and low in Na+. In contrast, the extracellular environment typically contains low levels of K+ and high levels of Na+. When cells are damaged, ion redistribution takes place, in which the intracellular concentrations of Na, Cl and Ca increase, and that of K decreases, if the cells are in an environment with high Na+ and low K+. However, by changing the composition of the extracellular fluid in an appropriate way, the intracellular ion concentrations can practically be set at any desired value (20,21). Incubation of the cells in a solution with a high K+/Na+ ratio does not necessarily preserve the cell membrane any better than incubation in a solution with a low K+/Na+ ratio, typical of the normal extracellular environment, but masks the effect of the damage to the membrane, since the extracellular fluid has about the same ionic composition as the cytoplasm. However, the effect is not purely cosmetic. Maintaining the cytoplasm during storage under ionic conditions close to the in vivo situation may prevent the onset of degenerative processes during incubation. It has been shown that loss of K+ from cells may induce apoptosis, and that apoptosis can be inhibited by preventing loss of K+ (28,29).

Our findings from X-ray microanalysis and from TEM were basically in agreement. First, Normal Krebs-Ringer's solution was found by both methods to be inadequate for maintenance of cell viability during cold storage, even for periods as short as 4 h. Second, Modified Krebs-Ringer's solution and UW commercial solution were found to be capable of maintaining the ultrastructure and the elemental composition of pancreas cells for 4 h. The ultrastructural damage induced by storage in Normal Krebs-Ringer's solution included swelling of the endoplasmic reticulum and damage to the mitochondria. The mitochondria are extremely sensitive to adverse conditions and undergo progressive structural changes during necrosis of the cell. The swelling of the endoplasmic reticulum induced by incubation was less in the cells of the endocrine part than in the cells of the exocrine part. Also, the mitochondria in the cells of the endocrine part were more abundant and less damaged than those in the exocrine part in tissue. The endocrine cells are clearly more resistant to adverse conditions than the exocrine cells. This agrees with findings of Vollmar et al. (30), who found that exocrine tissue is more susceptible to microvascular ischemia/reperfusion injury than endocrine pancreatic tissue.

The data show that the concentrations of Cl in the exocrine part of the pancreas after incubation in Modified Krebs-Ringer's solution are closer to the in situ levels than in tissue incubated in Euro-Collins solution, UW solution and seven modified versions of the UW solution, which have very low values for Cl. This is not surprising in view of the fact that UW solution and all modified versions of this solution do not contain any chloride at all, whereas Euro-Collins solution contains 15 mm chloride and Modified Krebs-Ringer's buffer 41 mm chloride (20). Fluid secretion from exocrine pancreas cells depends on the concentration gradient for chloride between cell and extracellular component, and it may thus be advantageous to keep the intracellular chloride concentration close to the in situ value.

It can be concluded, that the only factor determining the intracellular content of diffusible ions (Na+, K+, Cl) is the ionic composition of the experimental solution. The choice of the nonpermeant polysaccharide (hydroxyethyl starch or dextran), the choice between glucose and raffinose, lactobionate or gluconate, or the choice between glutathione or N-acetylcysteine appear to have no significance. On the other hand, even minor changes in the ionic composition of the experimental solution, such as in variants (F) and (G), significantly affect the intracellular ion content.

The results from the time-course experiment show that the most pronounced changes take place during the first 4 h of incubation and that little change occurs between 8 and 12 h. This is to be expected, if indeed the intracellular ion concentration is determined only by the ion concentration in the storage fluid and not by other components of the storage fluid. It can be expected that the cells have largely equilibrated with the storage solutions after 4 h of incubation, and that further changes after that point in time are not significant.

Calcium cannot be considered an easily diffusible ion. It should be taken into account that X-ray microanalysis measures total calcium, not free Ca2+ ions. In exocrine cells most of the calcium is bound to proteins in the endoplasmic reticulum and the zymogen granules (31) and in endocrine cells, which have a higher calcium concentration than exocrine cells, calcium appears to be mainly present in secretory granules (32,33). Loss of calcium from the endocrine cells therefore is likely to reflect changes in the granules: either diffusion of calcium out of the granules or loss of granules. It is therefore plausible that calcium content close to the in situ content is a sign of good preservation. In this respect, commercial UW solution and variants (A) and (B) give the best results, whereas Euro-Collins differs most from the in situ values. The differences between the UW variants are, however, small (less than 20%), and comparison of the composition does not give a clear indication what type of compound could be responsible for these differences. The fact that the simplified variants (A) and (B) of UW give the same results as commercial UW indicates that, at least if retention of elemental content is used as a criterion, it would not be necessary to depend on commercial UW solution for preservation of islets. The increase in Ca in some variants of the UW solution is puzzling, but because most of the calcium in these cells is bound to the secretory granules (31), it might indicate a change in the calcium-binding properties of the secretory material rather than cell damage.

X-ray microanalysis provides an objective method to assess whether the intracellular ionic composition of the tissue is maintained during storage. The data from X-ray microanalysis agree with the morphological evaluation, which always is a more subjective method. There may be a need to assess the efficacy of different storage solutions, e.g. simpler and less expensive alternatives to the UW solutions, and X-ray microanalysis could be a useful technique in such investigations.

Acknowledgments

The expert technical assistance of Anders Ahlander and Leif Ljung is gratefully acknowledged. The study was financially supported by the Swedish Research Council (project 12X-07125).

Ancillary