Changes in gene expression in beta cells after islet isolation and transplantation using laser-capture microdissection
The process of islet isolation can cause chemical and mechanical injury to beta cells. In addition, hyperglycaemia after islet transplantation can compromise beta cell function. The aim of this experiment was to evaluate changes in gene expression in endogenous islets using laser-capture microdissection (LCM).
Materials and methods
Islets from B6AF1 mice were studied in situ in the pancreas as well as those freshly isolated or cultured for 24 h. Fresh islets were transplanted under the kidney capsule of syngeneic diabetic (streptozocin-induced) and non-diabetic mice. Frozen sections from all the samples were prepared for LCM to obtain beta cell-enriched tissue; RNA was extracted and amplified using T7 polymerase. RT-PCR was used to assess expression of selected genes critical for beta cell function (Ins, Ipf1 [previously known as Pdx1], Slc2a2 [previously known as GLUT2] and Ldha) and the stress response (Hmox1 [previously known as HO-1], Gpx1, Tnfaip3 [previously known as A20] and Fas). Immunostaining was also performed.
In freshly isolated and cultured islets, insulin and Ipf1 mRNA levels were decreased by 40% (compared with islets in situ), while stress genes were upregulated. Comparison between in situ pancreatic islets and engrafted beta cells of cured mice showed declines in Ipf1 expression.
Our experiment, the first report to investigate changes in gene expression in endogenous islets using LCM, indicate that beta cells following islet isolation and residing in a foreign graft environment have decreased expression of genes involved in insulin production and increased expression of stress genes. Our data suggest that an islet graft, even in successful transplantation, may be different from endogenous islets in gene expression.
KeywordsIslet Laser-capture microdissection Transplantation
glutathione peroxidase 1
haematoxylin and eosin
haem oxygenase 1
insulin promoter factor 1
lactate dehydrogenase A
probable iron-regulated peptidyl-prolyl cis-trans isomerise A
solute carrier family 2 (facilitated glucose transporter) member 2
TATA box binding protein-related factor
TNF, alpha-induced protein 3 (previously known as A20)
Advances in human islet transplantation have increased the number of patients who achieve insulin-independence, yet there are still concerns about the survival and function of these islets in the liver [1, 2]. Even with islets being obtained from two to four donor pancreases, most patients continue to have impaired glucose tolerance . A substantial amount of beta cell death must occur in the immediate transplantation period due to anoxia  and inflammation . In addition, beta cell function may be impaired by even mild hyperglycaemia [5, 6] and hypoxia , and poorly understood factors related to the unnatural graft site. Over a period of 7–10 days islets are revascularised but have decreased oxygen tension compared with endogenous islets, which may in part be due to an abnormal distribution of the new vessels in the graft [8, 9, 10].
Laser-capture microdissection (LCM) makes it possible to obtain very small amounts of tissue from frozen sections, thus allowing collection of the beta cell-rich portions of islets from tissue sections. The RNA from these specimens can then be extracted and analysed for gene expression . The beauty of LCM is that tissue can be selectively obtained from the beta cell-rich core of islets, leaving behind the confounding non-beta cells. We expect that islets damaged by the isolation procedure and transplantation process will have gene profile patterns different from those of endogenous islets. The aim of the present study was to investigate the changes in gene expression in beta cells using LCM before and after islet isolation and transplantation.
Materials and methods
Animals and induction of diabetes
For islet donors and recipients, inbred male B6AF1 mice (Taconic Farms, Germantown, NY, USA), aged 6–8 weeks, were used. Islets were classified into four groups: (1) islets in the pancreas (n = 7); (2) freshly isolated islets (n = 6); (3) islets cultured for 24 h in RPMI 1640 medium (Sigma-Aldrich, St Louis, MO, USA) with a glucose concentration of 200 mg/dl (n = 6); and (4) islet grafts (normal; n = 6, diabetic; n = 8). Transplantation recipients were either normal or diabetic, having been treated with an i.p. injection of streptozocin (STZ) (180 mg/kg body weight, dissolved in citrate buffer, pH 4.5; Sigma, St Louis, MO, USA). Before transplantation, diabetes was confirmed by the presence of hyperglycaemia; only those mice with glucose levels ≥19.4 mmol/l for at least 1 week were used as recipients. At no point were diabetic mice treated with exogenous insulin. All animals were maintained in accordance with US Government and local institutional guidelines.
Islet isolation and culture
The islet isolation method has been described previously . Briefly, pancreases were distended with 3 ml cold M199 medium (Gibco BRL, Grand Island, NY, USA) containing 1.5 mg/dl of collagenase (type P; Boehringer Mannheim, Mannheim, Germany), excised, and incubated in a stationary bath at 37°C. Islets were separated by density gradients (Histopaque-1077; Sigma) and handpicked under a stereomicroscope. Islets were divided into samples of 250 and 500 islets for transplantation, whereas others, in groups of 500 islets, were cultured free-floating for 24 h in 5 ml culture medium consisting of RPMI 1640 supplemented with l-glutamine, benzyl penicillin (100 U/ml; Roche Diagnostics Scandinavia, Bormma, Sweden), and 10% FCS (Sigma-Aldrich).
Aliquots of islets were placed into 200 μl pipette tips (USA Scientific Plastic, Ocala, FL, USA) connected to a 1 ml Hamilton syringe (Hamilton Company, Reno, NV, USA) and were allowed to settle by gravity. After sedimentation, the tip was connected to polyethylene tubing (PE-50) and samples of 250 and 500 islets were transplanted under the kidney capsule of the normal and diabetic mice, respectively . After transplantation, blood glucose levels and body weights were monitored daily during the first week after transplantation and then every other day. Blood was obtained from snipped tails, and glucose was measured with a glucose meter (One Touch II; Johnson & Johnson, Milpitas, CA, USA). Grafted islets were retrieved 3 weeks after transplantation.
RNA extraction and amplification
cDNA synthesis and semiquantitative RT-PCR
To synthesise cDNA, 500 ng aRNA were reverse-transcribed in a 25-μl solution containing 5 μl first-strand buffer (Life Technologies), 10 mmol/l dithiothreitol, 1 mmol/l dNTP, 50 ng random hexamers, and 200 units of Superscript II RNase H− reverse transcriptase (Life Technologies). Reverse transcription reactions were incubated for 10 min at 25°C, 60 min at 42°C and 10 min at 95°C. cDNA products were diluted to a concentration corresponding to 20 ng starting RNA per 3 μl and stored at −80°C. PCRs were performed in 50-μl volumes consisting of 3 μl cDNA, 1.5–2.0 mmol/l MgCl2, 80–160 mmol/l dNTP, 40–500 nmol oligonucleotide primers, 92.5 kBq α-32P-labelled dCTP (111 TBq/mmol), 5 μl GeneAmp PCR buffer and 5 units AmpliTaq Gold DNA polymerase (Perkin-Elmer, Foster City, CA, USA). The Electronic supplementary material (ESM) Table 1 shows primers designed with Eugene version 2.2 software (Daniben Systems, Cincinnati, OH, USA). After two rounds of amplification, the produced aRNA is shorter and closer to the 3′ end than mRNA and thus requires primers designed within 300 bases from the poly-A tail with a product size less than 200 bases (Electronic Supplementary Material [ESM] Table 1). All reactions were performed in a 9700 Thermocycler (Perkin-Elmer), in which samples underwent a 10-min initial denaturing step to activate the DNA polymerase, followed by cycling of 1 min at 94°C, 1 min at the annealing temperature of each primer, and 1 min at 72°C. The final extension step was 10 min at 72°C. An internal control (housekeeping) gene (Ppia [probable iron-regulated peptidyl-prolyl cis-trans isomerise A, previously known as cyclophilin], Tbpl1 [TATA box binding protein-related factor] or Tuba2 [α-tubulin]) was used to normalise the input template cDNA to analyse relative expression of the gene of interest. Different internal control genes were used as required to match the amount of expression of the various tested genes. To be specific, Ppia was used for insulin (Ins) and glutathione peroxidase 1 (Gpx1); Tbpl1 was used for lactate dehydrogenase A (Ldha), haem oxygenase 1 (Hmox1, previously known as HO-1) and TNF receptor superfamily member 6 (Fas); and tuba2 was used for insulin promoter factor 1 (Ipf1, previously known as Pdx1), solute carrier family 2 (facilitated glucose transporter), member 2 (Slc2a2, previously known as GLUT2) and TNF receptor superfamily member 6 (Tnfaip3, previously known as A20). PCR products were separated on a 6% polyacrylamide gel in Tris-borate-EDTA buffer. Band intensity was measured with a PhosphoImager and quantified with ImageQuant software (Molecular Dynamics, Sunnyvale, CA, USA). Gene expression for control samples (see Fig. 4) was the ratio of the test gene expression to internal control expression and designated as 100%. Then the various experimental sample results were expressed as a percentage. Sequences of oligonucleotide primers and PCR conditions are available online (ESM Table 1).
Immunostaining of insulin, Tnfaip3 and Hmox1 protein was performed using the avidin–biotin complex (ABC) method (Vectastain ABC Kit; Vector Laboratories, Burlingame, CA, USA). Frozen tissue sections were fixed by immersion in acetone, incubated for 20 min and then incubated overnight at 4°C with rabbit anti-Tnfaip3 antibody (rabbit anti-human, a gift from S. Grey, Boston, MA, USA; 1:1,000 in PBS containing 1% BSA) and then for 30 min with biotinylated anti-rabbit IgG (Vector Laboratories; diluted 1:200). The same sections were stained for insulin (guinea-pig anti-human insulin, 1:200; Linco Research, St Charles, MO, USA). Staining for macrophages on frozen sections was done with anti-CD163 (mouse anti-rat 1:100 dilution; AbD Serotec Ltd, Oxford, UK). The secondary antibodies used for immunofluorescence were Texas Red-conjugated affinipure donkey anti-guinea-pig IgG (1:400; Jackson Immuno-Research, West Grove, PA, USA) for insulin and FITC-conjugated donkey anti-rabbit IgG and streptavidin-conjugated FITC (1:400; Jackson ImmunoResearch) for Tnfaip3. For Hmox1 immunostaining, tissues were fixed by immersion in 4% buffered formaldehyde, and then embedded in paraffin. Paraffin sections were incubated for 10 min with 3% H2O2 solution to block endogenous peroxidase and then overnight at 4°C with goat anti-Hmox1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) diluted 1:1,000 in PBS containing 1% BSA, and for 30 min with biotinylated anti-goat IgG (Vector Laboratories) diluted 1:200. These sections were then incubated with ABC reagent for 30 min, and positive reactions were visualised by incubation with the peroxidase substrate solution containing 3,3′-diaminobenzidine tetrahydrochloride.
Results are presented as the mean±SE. Analyses were performed using an unpaired Student’s t test, or a one-way ANOVA with the post-test of Scheffé. Differences were considered statistically significant at p < 0.05.
Blood glucose and body weight after transplantation
Changes in expression of genes concerned with beta cell function after islet isolation and transplantation
Changes in expression of stress genes after islet isolation and transplantation
Changes in stress gene mRNA expression in fresh, cultured and transplanted islets are shown in Fig. 4b. As previously noted , only minimal expression of Hmox1 mRNA was detected in endogenous islets, yet abundant expression was found in fresh and cultured islets, and even more in islet grafts. The mRNA levels for Gpx1 were higher in both fresh and cultured islets compared with endogenous islets (p < 0.01) and was further increased in transplanted islets (p < 0.05). Tnfaip3 was originally described as a TNF-α-induced anti-apoptotic gene in endothelial cells . The expression level of Tnfaip3 was markedly increased immediately after islet isolation (twofold, p < 0.01) and remained higher in both cultured islets and transplanted islets in cured diabetic animals. In contrast to genes that are potentially anti-apoptotic, expression of the pro-apoptotic gene Fas was not changed after islet isolation but increased twofold in transplanted islets.
Comparison of beta cell gene expression in transplanted islets between cured and normal mice
Comparison of islet gene expression in grafts of non-diabetic and diabetic cured mice
100 ± 7.6
77.1 ± 11.1
100 ± 12.5
64.7 ± 8.8
100 ± 10.1
105.1 ± 8.6
100 ± 17.8
113.9 ± 18.1
100 ± 21.3
118.7 ± 20.2
100 ± 19.2
121.8 ± 25.8
100 ± 20.4
91.0 ± 19.8
100 ± 17.79
105.1 ± 22.35
Comparison of stress gene expression in transplanted islets between cured and normal animals
Changes in stress gene mRNA levels in transplanted islets from normal and diabetic mice are shown in Table 1. The mRNA levels for Hmox1 and Gpx1 in transplanted islets of diabetic mice had the tendency to be increased compared with those of normal mice but were not statistically significant. Tnfaip3 mRNA expression in transplanted islets showed no difference between normal and diabetic mice. Similarly, Fas mRNA level was similar in both groups 3 weeks after transplantation.
In both freshly isolated and cultured islets there are marked changes in the expression of beta cell genes concerned with insulin secretion and the stress response that could influence beta cell survival. During the process of islet isolation, chemical and mechanical injury occurs and this interruption of the vascular supply compounds the problem by producing hypoxia. Thanks to LCM, this is the first time that gene expression in beta cell tissue from isolated islets could be rigorously compared with that of islets in situ in the pancreas.
Expression of genes important for insulin synthesis in beta cells, insulin and Ipf1, were decreased in both freshly isolated and cultured islets. These reductions in gene expression seen in freshly isolated islets cannot be explained by a fall in transcription because insulin, and to a lesser extent Ipf1, message levels are stable, with their half-lives being many hours to days . During the period of 3–4 h required for preparing fresh rodent islets, the fall of message levels must have been partially due to accelerated degradation of mRNA. Nevertheless, overnight-cultured islets are usually more compact because the extracellular spaces collapse in culture; the reduced intensity of staining for insulin from cultured islets can be explained by the degranulation that is seen after short-term culture.
It is of interest that the expressions of the glucose transporter Slc2a2 and of Ldha were not reduced. The practical importance of the reduction of insulin and Ipf1 message in islets prior to transplantation is that it may be one of various factors leading to reduced insulin secretion from grafts. It is of interest that in islet grafts retrieved 21 days after transplantation from mice with normalised glucose levels, there is still a reduction of Ipf1 message (p = 0.048) and a marginal change in insulin message (p = 0.115). These changes could be due to overwork of beta cells in this transplantation situation; as with humans , these mice might have had glucose intolerance. In our studies with minimal chronic hyperglycaemia, similar changes in beta cell gene expression were observed, including an increase in Ldha , as was found in the grafts of the present study. The Ldha increase may be part of a greater change in beta cell phenotype and may in itself contribute to altered function, since Ldha activity is usually very low in beta cells  and overexpression may have a deleterious influence . Many questions remain about how well islets function when transplanted, with a number of studies in rodents suggesting that insulin secretion is impaired [22, 23].
Data on abnormalities of islet grafts continue to accumulate and could help explain the poor outcomes now found clinically. The oxygen content of islets in a graft site is considerably lower than that found in islets in their normal pancreatic location ; also, there is lower pH and a shift to non-oxidative glucose metabolism [24, 25, 26]. In addition, islet grafts have been found to have lower insulin content than normal islets .
Isolated islets also manifest a stress response as shown by markers including TNF-α, IL-1, IL-6, monocyte chemoattractant protein and tissue factor [28, 29]. Activation of the islet stress response may cause problems during the peritransplantation period by promoting the instant blood-mediated inflammatory reaction, which may enhance immune destruction of transplanted islets . Some of this stress response may be due to the endotoxin content of collagenase preparations [28, 30], but mechanical insults and hypoxia could also be playing a role. Accumulation of intracellular reactive oxygen species after hypoxia can increase antioxidant enzymes in vitro, which can protect islets from oxidative injury [31, 32].
In the present study, evidence for a stress response was found with elevations in the expressions of Hmox1, Gpx1 and Tnfaip3. Interestingly, Fas expression was not elevated in isolated or cultured islets, but was in grafts 3 weeks after transplantation. The stress response is partially mediated by activation of nuclear factor κB, which leads to expression of both pro- and anti-apoptotic factors in islets [16, 33]. Therefore, even though the present experiments only show increased expression of anti-apoptotic genes in isolated and cultured islets, had a full array of stress genes been measured, activated pro-apoptotic genes might have been found. As for concerns that inflammatory cells in the graft might account for changes, we think this is unlikely because we find no evidence for such a reaction in these syngeneic grafts. On balance, it may be that Hmox1, Gpx1 and Tnfaip3 are playing important roles in allowing isolated islets to survive. Hmox1 overexpression has been found to protect islets from apoptosis [34, 35]. Expression of Tnfaip3 in islets, as in other cell types confers resistance to cytokine-induced apoptosis [36, 37]. The finding of increased Fas expression in the beta cell-rich tissue of grafts raises important questions about whether transplanted islets might be more susceptible to killing by the immune system.
In summary, using the novel approach of LCM we have been able to measure gene expression in beta cell-rich islet tissue obtained from pancreas, isolated islets and grafts and found important changes in gene expression that could have both adverse and protective effects upon islet function. Genetically modifying islets to overcome these adverse changes in gene expression could result in better transplantation outcomes.
This study was supported by grants from the National Institute of Health (DK 50657, U19DK6125), the Juvenile Diabetes Research Foundation and the Diabetes Research and Wellness Foundation. Help was also provided by the Joslin Diabetes and Endocrinology Research Center (DERC) supported by the National Institutes of Health (DK36836-16).
Duality of interest
The authors declare that they have no duality of interest.