A Method for Mass Harvesting Islets (Brockmann Bodies) from Teleost Fish

Cloning and molecular characterization of the glucose transporter 1 in tilapia (Oreochromis niloticus)

Facilitative glucose transporters (GLUTs) are responsible for passively transporting monosaccharides across the plasma membrane. We sequenced and characterized the Nile tilapia (Oreochromis niloticus) GLUT-1 (tGLUT-1) cDNA and genomic DNA. Using rapid amplification of the cDNA ends (RACE), two tGLUT-1 transcripts were detected differing in the length of the 3' untranslated region, 2851 and 4577 bp. Translated tGLUT-1 is a 490 amino acid product, which shares 74% homology with that of humans. Computer analysis of the amino acid sequence predicted 12 transmembrane domains, which are conserved in the GLUT-1 of various species. The tGLUT-1 gene spans more than 11 kb, and similar to the mammalian GLUT-1 genes has a 10 exon, 9 intron organization. Potential promoter regulatory elements have some similarity to those recorded for human, mouse, and rat GLUT-1 genes. Tissue expression studies revealed both GLUT-1 transcripts in liver, Brockmann bodies (BB), heart, small intestine, adipose tissue, white and red muscle, gill, spleen, pituitary gland, and brain. The highest level of expression was detected in tilapia heart, followed by BB, brain, and muscle. Protein based food and glucose had minor or no effects on the level of tGLUT-1 expression in most tissues. The tGLUT-1 mRNA level was significantly induced by glucose and food only in white muscle. Current results suggest that tGLUT-1 is similar to the GLUT-1 of other teleost species and mammals at the genomic, mRNA, and amino acid levels, supporting the concept that tGLUT-1 functions as a ubiquitous basal level glucose transporter.

Pancreatic islet transplantation into the bone marrow of the rat

BACKGROUND

The liver is the current site for pancreatic islet transplantation, but presents important technical complications and limitations. We asked whether pancreatic islets could be engrafted in the bone marrow, an easily accessible and widely distributed transplant site that may lack the limitations seen in the liver.

METHODS

We implanted pancreatic islet isografts (Lewis islets to Lewis rats), allografts (Wistar Furth islets to Sprague Dawley rats), and xenografts (Tilapia islets to Sprague Dawley rats) into the bone marrow of nondiabetic recipients and assessed survival by histology and immunocytochemistry. No immunosuppression was used.

RESULTS

Isografts and allografts showed positive staining for insulin and glucagon and no evidence of allograft rejection up to 21 days posttransplant. Xenografts were acutely rejected.

CONCLUSIONS

The bone marrow may be an attractive alternative site for pancreatic islet transplantation. The acceptance of allografts and isografts but rejection of xenografts suggests a selective phenomenon for the inflammatory process.

Rapid destruction of encapsulated islet xenografts by NOD mice is CD4-dependent and facilitated by B-cells: innate immunity and autoimmunity do not play significant roles

BACKGROUND

Spontaneously diabetic NOD mice rapidly reject microencapsulated islet xenografts via an intense pericapsular inflammatory response.

METHODS

Tilapia (fish) islets were encapsulated in 1.5% alginate gel microspheres. Recipients in series 1 were spontaneously diabetic NOD mice and streptozotocin-diabetic nude, euthymic Balb/c, prediabetic NOD, and NOR (a recombinant congenic strain not prone to autoimmune diabetes) mice. Recipients in Series 2 were STZ-diabetic NOD, NOD-scid, NOD CD4 T-cell KO, NOD CD8 T-cell KO, and NOD B-cell KO mice.

RESULTS

In Series 1, encapsulated fish islet grafts uniformly survived long-term in nude mice but were rejected in Balb/c and, at a markedly accelerated rate, in spontaneously diabetic NOD, streptozotocin-diabetic NOD and NOR recipients. Histologically, intense inflammation (macrophages and eosinophils) surrounding the microcapsules was seen only in NOD and NOR recipients. In Series 2, encapsulated fish islets uniformly survived long-term in NOD-scid and NOD CD4 KO mice; graft survival was markedly prolonged in B-cell KO (P<0.001) but not CD8 KO mice.

CONCLUSIONS

The rapid rejection of alginate encapsulated islet xenografts by NOD mice is not solely a consequence of beta-cell directed autoimmunity nor is it merely a vigorous innate immune response. Graft rejection requires CD4 T-cells, is facilitated by B-cells, and does not require CD8 T-cells.

Development of the islets, exocrine pancreas, and related ducts in the Nile tilapia, Oreochromis niloticus (Pisces: Cichlidae)

Pancreatic development and the relationship of the islets with the pancreatic, hepatic, and bile ducts were studied in the Nile tilapia, Oreochromis niloticus, from hatching to the onset of maturity at 7 months. The number of islets formed during development was counted, using either serial sections or dithizone staining of isolated islets. There was a general increase in islet number with both age and size. Tilapia housed in individual tanks grew more quickly and had more islets than siblings of the same age left in crowded conditions. The pancreas is a compact organ in early development, and at 1 day posthatch (dph) a single principal islet, positive for all hormones tested (insulin, SST-14, SST-28, glucagon, and PYY), is partially surrounded by exocrine pancreas. However, the exocrine pancreas becomes more disseminated in older fish, following blood vessels along the mesenteries and entering the liver to form a hepatopancreas. The epithelium of the pancreatic duct system from the intercalated ducts to the main duct entering the duodenum was positive for glucagon and SST-14 in 8 and 16 dph tilapia. Individual insulin-immunopositive cells were found in one specimen. At this early stage in development, therefore, the pancreatic duct epithelial cells appear to be pluripotent and may give rise to the small islets found near the pancreatic ducts in 16-37 dph tilapia. Glucagon, SST-14, and some PPY-positive enteroendocrine cells were present in the intestine of the 8 dph larva and in the first part of the intestine of the 16 dph juvenile. Glucagon and SST-14-positive inclusions were found in the apical cytoplasm of the mid-gut epithelium of the 16 dph tilapia. These hormones may have been absorbed from the gut lumen, since they are produced in both the pancreatic ducts and the enteroendocrine cells. At least three hepatic ducts join the cystic duct to form the bile duct, which runs alongside the pancreatic duct to the duodenum.

Islet transplantation in the discordant tilapia-to-mouse model: a novel application of alginate microencapsulation in the study of xenograft rejection

BACKGROUND

Tilapia islet xenograft rejection is characterized by infiltration with macrophages (Mphis), eosinophils (Ephis), and T lymphocytes. The presence of these cells indicates they contribute to rejection; therefore, an attempt was made to assess their role through host immunomodulation.

METHODS

Tilapia islet cells were transplanted under the kidney capsule of streptozotocin diabetic Balb/c mice, which were then treated with one of several immunomodulatory regimes targeting Mphis, Ephis, or T cells. Mphis were depleted using either silica or liposome-entrapped Cl2MDP. Ephi migration was blocked using monoclonal antibodies (mAbs) targeting interleukin (IL)-4 or IL-5. T-cell function was altered with mAbs targeting CD3, CD4, or CD8. Finally, T helper (Th)1 and Th2 activity was altered by depleting essential Th1 or Th2 cytokines with mAbs or by promoting a Th1 response with the injection of exogenous IL-12. The effects of antibody-mediated immunomodulation on graft survival were initially screened by cotransplanting alginate-encapsulated, mAb-secreting hybridoma cells into the peritoneal cavity at the time of islet transplantation. Significant prolongation was then confirmed using purified antibodies injected at the time of islet transplantation. Rejected grafts were examined histologically, and immunohistochemistry was used to assess the cellular infiltrates for each of the treatment groups.

RESULTS

Modulation of Mphis and Ephis alone did not significantly delay functional rejection of tilapia islet grafts (maximal mean graft survival time [mGST]=7.1+/-1.7 and 9.4+/-3.4, respectively) compared with untreated controls (mGST=8.2+/-1.0). Treatment of transplanted animals with antibodies against CD3 or CD4 significantly promoted graft survival (maximal mGST=16.3+/-5.8 and 34.0+/-11.6, respectively), whereas targeting CD8 and Th1 and Th2 cytokines showed no prolonging effect (maximal mGST=7.8+/-2.9 and 9.5+/-4.3, respectively).

CONCLUSION

Our results indicate that rejection in the tilapia-to-mouse model follows a pattern similar to other models of discordant islet cell xenotransplantation.

Xenogeneic milieu markedly remodels endocrine cell populations after transplantation of fish islets into streptozotocin-diabetic nude mice

It is unknown whether irrelevant foreign endocrine products secreted by xenografts would be biologically active and potentially harmful to recipients; even if entirely inert, continuous production might result in harmful circulating antigen-antibody complexes. We examined the fate of such a product using a fish (tilapia)-to-mouse islet xenograft model. Teleost fish islets, like mammalian islets, are composed primarily of cells producing insulin, glucagon or somatostatin; however, teleost fish have two different populations of somatostatin (SST) producing delta cells, one producing SST-14, a 14 amino acid SST identical to mammalian SST, which is derived from the pre-proSST-I gene which is present in all vertebrates, and the other a "large" (i.e. 22 to 28 amino acid) SST derived from a pre-proSST-II gene, which is not found in mammals. In contrast to 'large' SST, which has no mammalian homolog, teleost fish insulins, glucagons and SST-14 exhibit significant biological activity in mammals. Tilapia islets were transplanted under the kidney capsules of streptozotocin-diabetic nude mice, and mice with functioning grafts were killed at various times after transplantation. Serial sections of graft-bearing kidneys were stained by immunoperoxidase for insulin, SST-14, SST-25 and glucagon positive cells, and the areas of each cell type in the graft were measured using image analysis. Sections of untransplanted tilapia islets (both in situ and after harvest/culture) were also immunostained and measured as controls. Xenotransplantation of fish islets into diabetic nude mice resulted in the rapid degeneration and near total loss of SST-25+ cells, as well as a marked redistribution of the proportions of the remaining endocrine cell types. The proportions of cell types in the grafts gradually changed from a piscine pattern to that of mammalian islets.

Human beta cells are exceedingly resistant to streptozotocin in vivo

Streptozotocin (STZ) causes beta cell death in rodents via the mechanism of DNA damage precipitating poly(ADP-ribose) synthetase activation followed by lethal nicotinamide adenine dinucleotide depletion. It is unclear whether humans are susceptible to this mechanism. Islets were isolated from STZ-sensitive (CD1 mice and Lewis rats) and resistant [fish (tilapia)] species and from man and then were transplanted into diabetic nude mice under the kidney capsule. Normoglycemic recipients with normal glucose tolerance tests on d 30 were injected with increasing iv doses of STZ and their plasma glucose levels followed for 5 d; glucose tolerance tests were repeated on nondiabetic mice. Mice were then killed; grafts and native pancreata were examined. Based upon three criteria (i.e. nonfasting plasma glucose levels, glucose tolerance tests, and islet histology), the following observations were made: 1) Recipients of rat islets were resistant to 25 mg/kg but were uniformly diabetic at doses of 50 or 75 mg/kg. 2) Recipients of mouse islets were resistant to 75 mg/kg but were uniformly diabetic at 150 or 200 mg/kg. 3) Recipients of the fish islets were resistant to 300, 400, and 450 mg/kg. 4) Recipients of human islets were resistant to 100, 200, 300, 400, and 450 mg/kg. The results in recipient mice bearing long-term rat, mouse, or fish islet grafts were the same as previously published dose-response data for each donor species. We extrapolate from our results based on human islet grafts in mice that human beta cells are exceedingly resistant to STZ.

Liposomal encapsulation significantly enchances the immunosuppressive effect of tacrolimus in a discordant islet xenotransplant model

BACKGROUND

Encapsulation of tacrolimus (TAC) in a lipid bilayer to form liposome-encapsulated tacrolimus (LTAC) alters the biodistribution profile, half-life, and efficacy in organ allotransplantation models. LTAC has not been applied to either cell transplantation or xenotransplantation.

METHODS

To test the efficacy of LTAC in a discordant islet xenograft model, tilapia (fish) islets were transplanted under the left kidney capsules of streptozotocin-diabetic Balb/c mice. Recipient mice (groups I-VI) were treated with: I, untreated; II, empty liposomes; III, TAC (2 mg/kg/day); IV, TAC (5 mg/kg/day); V, LTAC (2 mg/kg/day); or VI, LTAC (5 mg/kg/day); all treatments were for 35 days or until rejection (i.e., two glucose measurements >200 mg/dl). Graft-bearing kidneys were removed for histology after rejection.

RESULTS

Mean graft survival time (mGST) for control groups I and II were 6.7+/-1.4 (n=6) and 7.5+/-1.3 days (n=4), respectively. Daily TAC treatment at 2 mg/kg/d (III) did not prolong graft function (mGST=7.7+/-1.6; n=6) although 5 mg/kg/day (IV) produced minimal prolongation to 12.8+/-4.8 days (n=12). Treatment with LTAC at 2 mg/kg/day (V) significantly prolonged mGST to 26.6+/-4.9 (n=5); however, all recipients rejected during treatment (i.e.,<35 days). LTAC at 5 mg/kg/day (VI) further prolonged mGST to 39.9+/-11.8 days (n=12) with only one mouse rejecting before day 35. Histologically, at the time of functional rejection, grafts were generally either totally or partially effaced by mononuclear cells, eosinophils, and fibrosis. In groups VI, islet grafts removed from two mice that died while they were normoglycemic and from a mouse terminated while it was normoglycemic at day 36 were viable, well-granulated, and free from cellular infiltration. The group VI grafts examined at rejection (i.e., 1-2 weeks after discontinuing LTAC) were generally totally obliterated and were in two instances associated with nodular aggregates of atypical lymphocytes resembling posttransplant lymphoproliferative disorder.

CONCLUSIONS

LTAC is the most potent immunosuppressive compound we have tested in our discordant fish-to-mouse islet xenograft model; however, toxicity is an issue at high doses.

Heterotopic cardiac xenotransplantation: fish-to-rat

The purpose of this study was to determine whether a vascularized organ transplanted from a teleost fish to a rodent would be hyperacutely rejected. Previous reports describing the results of discordant xenotransplantation are almost exclusively across orders within the class Mammalia. We chose a species combination that crosses many phylogenetic barriers (i.e. species, genus, family, order, and class) as well as several hundred million years on an evolutionary timescale. Because no published methodology existed, we developed a microvascular surgical method for fish (tilapia)-to-rat heterotopic cardiac xenotransplantation. To minimize the blood pressure to which the graft would be exposed, the tilapia heart was placed on the venous side of the rat circulation between the left kidney and the inferior vena cava by end-to-side anastomoses of the donor aorta to the recipient inferior vena cava and by end-to-end anastomosis of the donor sinus venosus to the recipient left renal vein. Tilapia hearts were rejected hyperacutely, based on both routine histopathological examination and immunofluorescent staining for immunoglobulin and complement, but rejection required hours rather than minutes.

Co-encapsulation of Sertoli enriched testicular cell fractions further prolongs fish-to-mouse islet xenograft survival

BACKGROUND

We previously demonstrated that alginate microencapsulation can prolong fish (tilapia) islet xenograft survival in diabetic animals. However, at present, microencapsulation does not provide complete immune protection to discordant islet xenografts, and long-term graft survival requires supplemental low-dose systemic immunosuppression. In the present study, fish islets were co-encapsulated with Sertoli enriched testicular cell fractions to find out whether this would further prolong fish islet graft survival in diabetic mice.

METHODS

Sertoli enriched testicular cell fractions were enzymatically harvested from adult Balb/c or Wistar-Furth rats. They were cultured and co-encapsulated with fragmented tilapia islets in alginate microcapsules. Encapsulated islets alone or islets co-encapsulated with Sertoli cells were then intraperitoneally transplanted into streptozotocin-diabetic Balb/c mice, and graft survival times were compared. Encapsulated and co-encapsulated islet function was also confirmed in streptozotocin-diabetic athymic nude mice.

RESULTS

Co-encapsulation with Sertoli enriched testicular cell fractions further prolonged mean fish islet graft survival time from 21+/-6.7 days (encapsulated islet cells alone) to >46+/-6.3 days (co-encapsulated with syngeneic murine Sertoli cells), without additional systemic immunosuppression. Testicular cells harvested from xenogeneic Wistar-Furth rats produced similar protective results (>46+/-10.9 days).

CONCLUSIONS

Our results support the theory that Sertoli cells produce local immunosuppressive factors. These factors supplement the immune protective feature of alginate microcapsules in our model. Testicular cell fractions may be an important naturally occurring facilitator in the development of new microencapsulation systems for islet xenotransplantation.

TILAPIA—A Source of Hypoxia-resistant Islet Cells for Encapsulation

Encapsulation of pancreatic islets prevents graft revascularization after transplantation, resulting in graft hypoxia and attrition. Hypoxia-resistant islets would be ideal for encapsulation. Tilapia, a tropical teleost fish, have large, anatomically discrete islets that can be easily harvested without expensive, fickle islet isolation procedures and that provide mammalian-like glucose tolerance profiles when transplanted into diabetic recipients. Because tilapia can live in stagnant water, we speculated that tilapia islets might tolerate lower oxygen tensions than mammalian islets. Tilapia and rat islets (n = 30) were placed in paired 60-mm Petri dishes containing 10 mL of deoxygenated CMRL-1066 media and cultured together in sealed chambers gassed with 95% N2/5% CO2. Islet viability was determined by fluorscein diacetate/ethidium bromide staining at intervals varying from 2.5 h to 7 days; blood gas measurements were obtained on media samples at the end of selected incubation intervals. Rat islets underwent near-total necrosis and fragmentation in <24 h; occasional viable single cells could be identified until about 72 h. On the other hand, the fish islets showed no loss of viability until about 72 h when some showed mild central necrosis. Even at 7 days, all fish islets appeared roughly 50% viable. Fish islets cultured under hypoxic conditions for 72 h (media, pO2 = 27.8 mmHg) and then transplanted into streptozotocin-diabetic athymic nude mice were viable (6/6) but showed some diminished function (3/6) over a 25-day follow-up period. Our results suggest that tilapia islets will survive and function at lower oxygen tensions than mammalian islets.

Long-term function of fish islet xenografts in mice by alginate encapsulation

BACKGROUND

Large, anatomically discrete pancreatic islets, Brockmann bodies (BBs), exist in certain teleost fish. When transplanted under the renal capsules of streptozotocin-diabetic athymic nude mice, BB grafts produce uniform normoglycemia for 50 days and mammalian-like glucose tolerance profiles; however, these very discordant islets reject in 7-8 days when transplanted into euthymic BALB/c mice.

METHODS

In the present study, BBs were mass harvested, minced into <1-mm tissue fragments, and encapsulated in alginate-based macrospheres (5 mm diameter) or noodles (0.5x3 cm). Nonencapsulated and encapsulated BB fragments were transplanted intraperitoneally into streptozotocin-diabetic (nonfasting blood glucose >400 mg/dl) nu/nu and BALB/c mice. Glucose levels were monitored at least 3 times a week.

RESULTS

Encapsulated BB grafts uniformly survived >50 days (10/10) or >100 days (3/3) in nu/nu recipients. The mean graft survival time was 27+/-13 days in BALB/c recipients (n=7). Daily intraperitoneal administration of 2.5 mg/kg 15-deoxyspergualin, in combination with encapsulation, resulted in uniform long-term BB graft function in BALB/c recipients (n=5). Similarly, long-term function was achieved in four of six BALB/c recipients with daily intraperitoneal administration of 10 mg/kg cyclosporine (two grafts failed after 39 and 45 days). Nonencapsulated BB grafts transplanted intraperitoneally into BALB/c or nu/nu recipients functioned for <7 days; immunosuppression alone did not permit graft survival in BALB/c recipients. In all cases of graft survival of >50 days, grafts were surgically removed from the peritoneal cavity, and blood sugar levels returned to a diabetic state within a few days. Historical sections of grafts, stained with hematoxylin and eosin and immunoperoxidase for insulin, showed viable, well-granulated BB tissue.

CONCLUSIONS

This study demonstrates that tilapia BBs are suitable for encapsulation and that encapsulated BBs can be made to function long term in diabetic mice.