The perfused porcine pancreas as a model for testing organ protective solutions

Development of pancreatic machine perfusion: translational steps from porcine to human models

BACKGROUND

Hypothermic machine perfusion (HMP) is increasingly being used for extended criteria kidney grafts. Pancreatic HMP is challenging because physiologically the pancreas is a low-flow organ susceptible to edema. We report the successful development of preclinical HMP models using porcine pancreases, as well as human pancreases unsuitable for clinical transplantation.

METHODS

Ten porcine pancreases were used in the development of these perfusion models. Pancreases underwent 24 h of static cold storage (SCS, n = 3) and then viability assessment on an isolated oxygenated normothermic reperfusion (NRP) circuit or 24-h SCS, 5 h of HMP, and then NRP (SCS-HMP, n = 3). Human pancreases (n = 3) were used in the development of a preclinical model.

RESULTS

Porcine HMP demonstrated stable perfusion indices at low pressures, with a weight gain of between 15.3% and 27.6%. During NRP, SCS-HMP pancreases demonstrated stable perfusion flow indices (PFIs) throughout reperfusion (area under the curve was in the range of 0.49-2.04 mL/min/100 g/mm Hg), whereas SCS-only pancreases had deteriorating PFI with a decline of between 19% and 46%. Human pancreas models demonstrated stable PFI between 0.18 and 0.69 mL/min/100 g/mm Hg during HMP with weight gain of between 3.9% and 14.7%. NRP perfusion in porcine and human models was stable, and functional assessment via insulin secretion demonstrated beta cell viability. Exocrine function was intact with production of pancreatic secretions only in human grafts.

CONCLUSIONS

Application of machine perfusion in preclinical porcine and human pancreas models is feasible and successful; the development of these translational models could be beneficial in improving pancreas preservation before transplantation and allowing organ viability assessment and optimization.

Pig pancreas oxygenation at 20 degrees C increases islet ATP generation but deteriorates islet function

Successful pancreas preservation during storage in oxygenated perfluorodecalin (PFD) is mainly related to oxidative ATP generation during storage. Increasing the storage temperature would accelerate this process essential for resuscitation of ischemically damaged pancreatic tissue. The present study aimed at comparing islet isolation outcome from adult pig pancreata preserved in oxygenated PFD by means of a one-layer method during storage on ice or at 20 degrees C. Resected pancreata were intraductally flushed with cold UW solution and promptly processed (n = 6) or stored for 3 h in continuously oxygenated PFD at 4 degrees C (n = 5) or 20 degrees C (n = 7). Prior to digestion-filtration pancreata were intraductally injected with UW supplemented with Serva collagenase NB8 and neutral protease. Islet quality assessment determined viability, glucose stimulation index, mitochondrial activity, intracellular ATP content, and transplant function in diabetic nude mice. Pancreata oxygenated for 3 h at 20 degrees C yielded islet numbers similar to organs oxygenated at 4 degrees C. Compared to a storage temperature of 20 degrees C, preservation at 4 degrees C reduced islet ATP content (p < 0.05) as well as islet viability (p < 0.05). Nevertheless, PFD storage at 20 degrees C decreased insulin response to glucose compared to unstored pancreata (p < 0.05). In contrast to unstored pancreata or cold-stored organs, transplantation of islets isolated after oxygenation at 20 degrees C was characterized by an early loss of transplant function in 50% of recipients (p < 0.05). The present study demonstrates that PFD storage at 20 degrees C enhances islet ATP synthesis within a short period of oxygenation but deteriorates islet function. We conclude that the present data reflect an equilibration between reduced depression of metabolic activity resulting in damage of islets and temperature-stimulated acceleration of ATP synthesis. Future studies are required to adjust the optimum storage temperature for pancreas oxygenation in different species.

Pancreas storage in oxygenated perfluorodecalin does not restore post-transplant function of isolated pig islets pre-damaged by warm ischemia

BACKGROUND

Cold storage in oxygenated perfluorodecalin (PFD) restores transplant function of ischemically damaged dog pancreata and reduces the impact of cold ischemia on recovery of isolated human islets. Whether PFD storage can improve islet isolation from pancreata exposed to significant warm ischemia (WI) is unclear yet. The present study aimed to clarify this question in adult pigs.

METHODS

After exsanguination, the intestine was removed immediately or left in the cavity for 30 min of WI. Resected pancreata were intraductally flushed with cold University of Wisconsin solution. Subsequently, pancreata were processed immediately by digestion-filtration (group I: 0 min WI, n=6; II: 30 min WI, n=6) or first stored for 3 h in oxygenated PFD (III: 0 min WI+PFD, n=5; IV: 30 min WI+PFD, n=6).

RESULTS

Pancreata subjected to 30 min of WI yielded significantly less islets compared with the corresponding non-ischemic organs (I vs. II, P<0.01; III vs. IV, P<0.05). Oxygenation did not ameliorate the loss in islet yield (II vs. IV, NS). Ischemic islets were characterized by depleted ATP stores (388+/-73 (I) vs. 133+/-22 ng/1000 IEQ (II), P<0.01) and diminished insulin response to glucose calculated as stimulation index (SI; 2.47+/-0.36 (I) vs. 0.25+/-0.17 (II), P<0.05). PFD storage of ischemic organs partially restored ATP content (217+/-23 ng/1000 IEQ, II vs. IV, P<0.05) and glucose SI (1.60+/-0.09, II vs. IV, P<0.05) to a significant extent that reached the level of corresponding PFD-stored, non-ischemic pancreata (III vs. IV, NS). Sustained normoglycemia was exclusively observed in diabetic nude mice transplanted with islets isolated from non-ischemic organs. The significantly reduced graft function of ischemic islets (I vs. II, III vs. IV, P<0.001) was not increased by pancreatic oxygenation (II vs. IV, NS).

CONCLUSIONS

The present study suggests that pancreas short-term storage in oxygenated PFD improves in vitro but not the in vivo function of ischemically damaged pig islets.

Successful pancreas preservation by a perfluorocarbon-based one-layer method for subsequent pig islet isolation

BACKGROUND

The oxygenation of human pancreas by the two-layer method (TLM) during cold storage was recently established for clinical islet transplantation. Simplification of TLM would facilitate the application of perfluorocarbon (PFC) as a regularly used preservation solution for subsequent islet transplantation. The present study examined whether PFC can be used in a one-layer method (OLM) for long-term pancreas preservation before isolation of adult pig islets.

METHODS

Resected pancreases were intraductally flushed with cold University of Wisconsin solution and immediately processed (n=6) or subjected to 7-hour storage by OLM (n=8) or TLM (n=10). Subsequently, pancreases were intraductally distended with collagenase NB-8 supplemented with neutral protease. Isolation and purification were performed as previously described.

RESULTS

Compared with unstored pancreases (3,670+/-740 islet equivalents [IEQ]) purified islet yield in TLM-stored organs (2,080+/-290 IEQ, P<0.05) was significantly decreased in contrast with OLM-preserved pancreases (3,110+/-520 IEQ, NS). No differences were observed between groups regarding purity (>90%), trypan-blue exclusion (>95%), adenosine triphosphate content, and mitochondrial viability of islets. Stimulation index during static glucose incubation (20 vs. 2.8 mm) was decreased after storage by TLM (1.81+/-0.20, P<0.05) but not by OLM (2.27+/-0.57) if compared with unstored pancreases (2.47+/-0.36). However, transplantation into diabetic nude mice resulted in sustained normoglycemia of recipients of either group until nephrectomy of graft-bearing kidneys was performed.

CONCLUSIONS

This study demonstrates that PFC alone can be used in a one-layer procedure for successful pig-pancreas preservation. This simplification can facilitate the broad application of PFC as pancreas preservation solution without reducing its benefits demonstrated by TLM.

Influence of Collagenase Loading on Long-Term Preservation of Pig Pancreas by the Two-Layer Method for Subsequent Islet Isolation

BACKGROUND

The introduction of the two-layer method (TLM) for long-term human pancreas preservation revealed the enormous potential of TLM to improve graft function of isolated islets. It is still unclear whether pig islets can be successfully isolated from pancreases after prolonged cold ischemia. To clarify this question, pig pancreases were subjected to 7-hour preservation by University of Wisconsin solution (UWS) storage or TLM. Another aim was to verify whether TLM can be synergistically combined with intraductal collagenase injection before cold storage.

METHODS

After intraductal flush with UWS, organs were distended with 4.4 PZ-U/g of UWS-dissolved collagenase NB-8 and neutral protease adjusted to respectively 1.1, 0.2, 0.5, or 0.8 DMC-U/g for pancreases freshly procured (n=6) or distended with enzymes before (TLM preloaded, n=7) or after cold storage (UWS storage, n=4; TLM postloaded, n=10).

RESULTS

Purified islet yield decreased from 429,200+/-86,700 islet equivalents (IEQ) in unstored pancreases to respectively 37,670+/-19620, 210,400+/-22900 and 238,000+/-26600 IEQ in UWS-stored (P<0.01), TLM-preloaded, or postloaded organs (P<0.05). Purity (>90%), viability (>95%), and insulin content were not different between groups. Islets from UWS-stored pancreases fragmented extensively, preventing further assessment of in vivo function. Compared with other experimental groups, islets from TLM-preloaded organs were characterized by enhanced basal and stimulated insulin release. Sustained normoglycemia was observed in diabetic nude mice transplanted with islets from TLM-postloaded or unstored pancreases in contrast with transient function in TLM-preloaded islets.

CONCLUSIONS

This study demonstrates that significant amounts of intact pig islets can be isolated after prolonged pancreas preservation by TLM. Enzyme administration before TLM preservation decreases islet graft function.

Hydroxyethyl starch does not improve pancreas preservation with HTK

The effect of hydroxyethyl starch on pancreas preservation with cardioplegic histidine-tryptophan-ketoglutarate solution (HTK) was investigated. The study was performed using an in vitro reperfusion system of the porcine pancreas. During organ preservation pancreatic weight, arterial pressure, volume flow, and washout of amylase and lactate were quantified. Addition of hydroxyethyl starch did not affect arteriovenous volume flow or washout of amylase and lactate during protective perfusion after pancreas preparation. However, hydroxyethyl starch in HTK prevented an increase in pancreatic weight at the end of the protective perfusion (102.2 +/- 4.55% vs 127.8 +/- 4.62% in controls; p < 0.005) and after 24 h cold ischemia (72.9 +/- 3.91% vs. 83.5 +/- 3.49% in controls; p < 0.05). Hydroxyethyl starch did not affect postischemic organ quality assessed during reperfusion in a perfusion chamber by pancreatic vascular resistance, amylase and lactate release, insulin secretion, and oxygen consumption. We conclude that hydroxyethyl starch does not bring about any further improvement in immediate postischemic organ quality assessed in an in vitro reperfusion system.

Histamine inhibits 5-hydroxytryptamine release from the porcine small intestine: involvement of H3 receptors

Strips of the porcine small intestine were incubated in vitro, and the release of 5-hydroxytryptamine (5-HT) was determined by high-pressure liquid chromatography with electrochemical detection. Removal of the mucosa resulted in a large reduction (95%) of tissue 5-HT, suggesting that enterochromaffin cells are the main source of 5-HT. The release of 5-HT was reduced by 70% after omission of calcium. Tetrodotoxin and hexamethonium reduced the release of 5-HT by 30%-40% in a nonadditive manner, indicating a spontaneous neuronal (nicotinic) excitatory input to the enterochromaffin cells. Histamine inhibited the release of 5-HT by about 50%. This effect was not affected by mepyramine or cimetidine but was effectively blocked by thioperamide, indicating the involvement of H3 receptors. The selective H3-receptor agonist R-alpha-methyl-histamine also inhibited 5-HT release. Because the effect of R-alpha-methyl-histamine was also observed in the presence of tetrodotoxin, an indirect, neuronally mediated action could be excluded. Therefore, the inhibitory H3 receptors may be localized directly at the enterochromaffin cells.

Nicotinic and muscarinic modulation of 5-hydroxytryptamine (5-HT) release from porcine and canine small intestine

Strips of porcine and canine small intestine were incubated in vitro and the release of 5-hydroxytryptamine (5-HT) was determined by HPLC with electrochemical detection. The spontaneous outflow of 5-HT from the porcine and canine small intestine largely reflects calcium-dependent 5-HT secretion from enterochromaffin cells which are under a spontaneous neuronal, excitatory input as indicated by the inhibitory effect (30-40%) of tetrodotoxin. In both species, nicotine enhanced the release of 5-HT in a concentration-dependent manner by a maximum of about 50% at 100 microM. This effect was blocked by the nicotine receptor antagonist hexamethonium, but not by the subtype-selective nicotine receptor antagonist alpha-bungarotoxin. The effect of nicotine was rapidly desensitized. The presence of tetrodotoxin abolished the effect of nicotine on 5-HT release in canine tissue but not in porcine tissue. The presence of the muscarine receptor antagonist scopolamine prevented the effect of nicotine on 5-HT release from canine tissue, but significantly enhanced 5-HT release from porcine tissue. The muscarine receptor agonist oxotremorine inhibited 5-HT release from porcine tissue, but increased 5-HT release from canine tissue. However, in the presence of tetrodotoxin, oxotremorine enhanced 5-HT release in tissue from both species. In conclusion, activation of nicotine receptors induce the release of 5-HT from porcine and canine small intestine. In the dog, the effect of nicotine is mediated via the release of acetylcholine which then stimulates 5-HT release via muscarine receptors on the enterochromaffin cells. In the pig, the stimulatory effect of nicotine appears to be located directly on the enterochromaffin cells. In addition, activation of neuronal muscarine receptors in the porcine small intestine induced the release of a previously unidentified neurotransmitter which inhibited 5-HT release. Nicotine, via cholinergic interneurons, also appears to induce the release of this inhibitory neurotransmitter which opposes the direct stimulatory action of nicotine on 5-HT release.

Calcitonin gene-related peptide (CGRP) modulates cholinergic neurotransmission in the small intestine of man, pig and guinea-pig via presynaptic CGRP receptors

The effect of calcitonin gene-related peptide (CGRP) on the cholinergically mediated twitch contraction in longitudinal muscle strips of the small intestine (duodenum, jejunum, ileum) of guinea-pig, pig and man was investigated. Independently of the anatomical region, CGRP inhibited the twitch response in the different specimens of all three species by about 40% with similar IC50 values (1.5-2.4 nmol/l). Only in the guinea-pig small intestine CGRP induced a contraction of the smooth muscle which was sensitive to scopolamine and tetrodotoxin. The electrically evoked [3H]acetylcholine release from jejunal longitudinal muscle strips with myenteric plexus attached of the guinea-pig, which were incubated with [3H]choline, was concentration-dependently inhibited by CGRP. A direct relaxant effect of CGRP on smooth muscle tone of carbachol precontracted preparations was only observed in specimens of the guinea-pig. In conclusion, presynaptic inhibitory CGRP receptors on cholinergic neurones modulate the release of acetylcholine in different parts of the small intestine.