Survival after solid organ transplantation has improved over time. However, the long-term quality and quantity of life of recipients is significantly impacted by complications such as posttransplant diabetes mellitus (PTDM).1 Obesity and associated metabolic alterations often develop after transplant.2 There is a 9-fold risk of diabetes in solid organ transplant recipients compared with age-matched controls.3,4
PTDM occurs in a large proportion of patients, with estimates ranging from 17% to 53%1,5 due to varying definitions of the condition.6 Criteria for PTDM have included elevated fasting blood glucose, increased glycated hemoglobin, abnormal oral glucose tolerance tests, or a need for hypoglycemic therapy following transplantation, in the absence of a preexisting diabetes diagnosis.6 More recently, the International Consensus Guidelines on PTDM recommended that this condition be diagnosed based on criteria for type 2 DM (T2DM) established by American Diabetes Association.7
PTDM was first described over 50 years ago8 and has gained recognition as a major complication with serious consequences.6 The presence of PTDM is associated with a 2- to 3-fold higher risk of fatal and nonfatal cardiovascular events than in nondiabetic patients.9 Additionally, patients with PTDM were reported to have a 63% higher risk of graft failure and 87% higher risk of overall mortality based on data from the United Renal Data System.10 A higher incidence of infectious complications in patients with PTDM has been documented, in addition to compromised graft and overall patient survival.11
The universal use of immunosuppressant drugs, such as corticosteroids, calcineurin inhibitors, and mammalian target of rapamycin inhibitors (mTORs), is known to significantly impact metabolic balance and is associated with the development of diabetes, hypertension, obesity, and dyslipidemia.12 Immunosuppressive therapy is therefore a key risk factor for PTDM, in addition to age, family history, and ethnicity.1 Tacrolimus is more diabetogenic than cyclosporine, and the combination of tacrolimus with sirolimus has been shown to be synergistic in its hyperglycemic effect.13 Additional risk factors include increased age, male gender, type of donor organ, donor age, history of rejection episodes, increased body mass index, and presence of other comorbidities, such as hypertension and dyslipidemia.1
Calcineurin inhibitors have been associated with decreased insulin signaling in conjunction with decreased synthesis and release as a manifestation of pancreatic β-cell dysfunction.14 Sirolimus is associated with peripheral insulin resistance and decreased pancreatic β-cell proliferation.15 These mechanisms have been postulated principally based on in vitro studies and histological findings. However, the molecular mechanisms underlying PTDM have not been clearly elucidated.
The aim of this study was to examine how immunosuppressants can increase susceptibility to development of diabetes posttransplant, using unique bioinformatics tools on publicly available data. We performed an integrative analysis of systematically curated, high-throughput gene expression data from solid organ transplant recipients with such data from patients with T2DM. In vitro validation was performed to confirm the key genes and pathways that increase risk of diabetes posttransplant, followed by investigation of the impact of hypoglycemic agents on gene expression.
MATERIALS AND METHODS
Data Collection, Analysis, and Database Compiling
All available high-throughput microarray gene expression datasets solid organ transplant recipients on immunosuppressants (tacrolimus, sirolimus, and cyclosporine) were retrieved from published datasets (PubMed, http://www.ncbi.nlm.nih.gov/PubMed) and the Gene Expression Omnibus (GEO), a public functional genomics data repository containing high-throughput array data (https://www.ncbi.nlm.nih.gov/geo) using the following Medical Subject Heading terms: transcriptome, gene expression profile, tacrolimus, transcriptome, sirolimus, cyclosporine, AND human. All entries on PubMed and GEO were considered for inclusion. The datasets selected from GEO were analyzed using GEO2R (https://www.ncbi.nlm.nih.gov/geo/info/geo2r.html), a web tool available on the portal. GEO2R compares original submitter-supplied processed data tables using the GEOquery and limma R packages from the Bioconductor project. Following instructions available online at (https://www.ncbi.nlm.nih.gov/geo/info/geo2r.html), all the dysregulated genes were retrieved and only those with an adjusted P value P < 0.05, and expression fold-change values ≤0.5 or ≥1.5 were collected for further analysis. Using this tool, we identified genes differentially expressed between samples from patients exposed to immunosuppressants versus control samples of individuals not exposed to immunosuppressants. Separately, we identified genes differentially expressed between samples from patients with T2DM when compared with healthy controls.
The study workflow for data from immunosuppressed solid organ transplant recipients is illustrated in Figure 1A. Studies were excluded if: (1) no original gene list was available, (2) they did not employ high-throughput methods (such as real-time amplification), (3) the data originated from cell lines or animal models, (4) the publications were case reports or review articles, (5) the data were not from patients exposed to tacrolimus cyclosporine or sirolimus, or (6) the data were not from solid organ transplant recipients (Supplemental Tables, SDC, http://links.lww.com/TP/B756).
Identification of Commonly Dysregulated Genes in Immunosuppression and T2DM
All dysregulated genes (up- or downregulated) that were overlapping and in common between the 2 groups of datasets (immunosuppression and T2DM) were identified using Venny 2.1, an online tool for Venn diagram design (http://bioinfogp.cnb.csic.es/tools/venny/index.html).
Pathway Enrichment Analysis
Dysregulated genes identified as being in common between immunosuppression and T2DM gene expression profiles were investigated for pathway enrichment using the Pathway Data Integration Portal, PathDIP v2.5 (http://ophid.utoronto.ca/pathDIP).16 We specifically selected 2 databases that focus on metabolic pathways (EHMN, HumanCyc) among the 20 databases available in the portal, while looking for literature-driven curated pathways.
Cell Culture Treatments Real-time Validation
THP-1 cells, a promonocytic cell line kindly provided by Dr Deepali Kumar, Toronto, and HepG2 cells were cultured in RPMI 1640 (Life Technologies, Grand Island, NY) supplemented with 10% fetal bovine serum, at a density of 0.75 × 106/mL. The media was refreshed every 3 days, and the cells were maintained in 5% CO2 at 37°C.
Cells were treated with tacrolimus (registered trademark LC Labs.com) for 6, 18, 24, and 48 hours. The drug was dissolved in dimethyl sulfoxide and resuspended in RPMI medium at a final concentration of 120 nmol/L.
Following treatment with tacrolimus, total RNA was extracted and quantified by spectrophotometry. RNA was reverse-transcribed to cDNA using QuantiTect Reverse Transcription Kit (Qiagen). Using the Step One Plus platform (Applied Biosystem), a real-time PCR (RT-PCR) was performed again to investigate the change in expression of PPARγ, known to be commonly modulated in the presence of immunosuppressants as well as in T2DM patients. The Step One Plus platform from Applied Biosystem was used for the RT-PCR. cDNA was amplified using SYBR Green Real-Time PCR Master Mix (Thermo Fisher) with specific oligonucleotide primers for target sequences or RPL13 (for normalization). H2O was the negative control. Specific oligonucleotide primers were PPARγ forward 5′-CGTGGCCGCAGATTTGAA-3′, reverse 5′-CTTCCATTACGGAGAGATCCAC-3′; RPL13 forward 5′-CCGGGTTGGCTGGAAGTACC-3′, reverse 5′-CTTCTCGGCCTGTTTCCGTAG-3′. Threshold cycles (Ct) were automatically calculated by the RT-PCR System. Each Ct value was normalized to the RPL13 Ct value and a control sample. The ΔΔCt methods were used to calculate the expressed fold-change of the genes of interest in the treated cells compared with control conditions.
Molecular Validation of the Commonly Dysregulated Genes
In order to perform an in vitro validation of the effect of tacrolimus on genes related to glucose metabolism and insulin signaling, a more comprehensive panel of focus genes in the glucose metabolism (PAHS-006Z) and insulin signaling pathway (PAHS-030Z) arrays was tested by a RT-PCR approach using the RT2 Profiler PCR Array from SABiosciences (Qiagen). Each array is a plate with prespotted primers a panel of 84 focused genes specifically involved in the chosen biological process. The whole list of genes is available online (http://www.SABiosciences.com/Metabolic_Diseases.php). The RNA extracted from THP-1 cells (1 μg) was retro transcribed using RT2 First Strand Kits (SABiosciences, Qiagen).
The RT-PCR reaction was prepared by adding the cDNA to the SYBR Green Master Mix (SABiosciences, Qiagen), and the protocol followed the manufacturer’s instructions for The Step One Plus platform from Applied Biosystems. Data were analyzed based on the ΔΔCt method using the software provided online by the manufacturer at http://pcrdataanalysis.SABiosciences.com/pcr/arrayanalysis.php. The list of dysregulated genes (either up- or downregulated) detected by SABiosciences arrays after exposure of the THP-1 cell line to Tacrolimus and their relative metabolic pathways of functional gene grouping (as reported at http://sabiosciences.com/rt_pcr_product/HTML/PAHS-006Z.html, http://sabiosciences.com/rt_pcr_product/HTML/PAHS-030Z.html) were obtained.
In order to investigate the interactions between genes of interest, the list of 146 dysregulated genes was uploaded into ingenuity pathway analysis (IPA Ingenuity Systems, www.ingenuity.com) software to obtain the networks. IPA identifies network and pathway interactions between genes based on an extensive manually curated database of published interactions. We uploaded the 146 dysregulated genes and the associated expression value from the RT2 Profiler PCR Arrays from SABiosciences (Qiagen) into IPA. These genes, called focus genes, were overlaid onto a global molecular network based on the Ingenuity knowledge database. IPA includes a large repository of gene-phenotype associations, molecular interactions, chemical knowledge, and regulatory events, manually curated from scientific publications. Networks based on these focus genes were algorithmically created based on their direct or indirect interactions. Scores, calculated on Fisher exact test by IPA, were obtained to rank networks based on their relevance to the dysregulated genes. The score evaluates the number of focus genes from our original dataset in the network, and the size of the network to approximate how relevant this network is to the original list of focus genes. The network is then shown as a graph representing the molecular relationships/interactions as an edge (line) between genes or gene products (nodes). The connectivity of these nodes representing the genes is based on the data collected in the IPA knowledge base. The node color indicates an upmodulation (red) or downmodulation (green). Nodes are displayed using various shapes to represent the functional role or class of gene product (ie, kinase, transcription regulator, enzyme). Edges are displayed with various colors or labels to better describe the nature of the relationship between nodes. The dysregulated genes were mapped onto the core networks to explore their connection to biological function or diseases affecting glucose metabolism. Canonical Pathways and chemical compounds significantly associated with our list of genes were obtained using IPA ingenuity’s knowledge base. The right-tailed Fisher exact test was used to calculate a P value determining the probability that the association or overlap between genes listed in the dataset and a given pathway’s neighborhood was due to chance alone.
Based on gene modulation, IPA is able to calculate a z score, defined as a statistical measure of correlation between relationship direction and a given set of modulated genes, as a value to measure “the non randomness” of directionality. A z score <−2 or >+2 is considered significant. A negative z score of <−2 predicts inhibition with high confidence, and a positive score of >+2 predicts activation with strong confidence. Outside these values, the prediction of activation or inhibition is less confident. In some cases, the z score cannot be calculated due to insufficient information stored in the IPA knowledge base, in which case there is no available prediction.
IPA protein-protein interaction network analysis core networks were merged and overlaid with related “functions and diseases” to determine genes associated with specific hepatic biological and pathological processes according to the IPA knowledge base. The molecular activity predictor tool was used to predict the crosstalk relationship among our genes and their interactors based on their modulation. The drug database was overlaid with Core Network #1 to identify the possible effect of calcineurin inhibitors in the context of glucose metabolism.
Identification of Optimal Hypoglycemic Agent for PTDM
The drug database and upstream regulators analysis tool from IPA were used to identify putative chemical compounds or therapeutic agents able to target the dysregulated genes in THP-1 cells after tacrolimus exposure. High-throughput gene expression datasets of humans and mice treated with the hypoglycemic agents metformin, rosiglitazone, and insulin were obtained from GEO. The list of dysregulated genes in treated versus nontreated conditions was obtained from the following datasets: GSE36714, GSE7193, GSE1458, GSE22309, and GSE7146. The genes obtained in animals were converted to their orthologues in humans using the web tool available online dbOrtho https://biodbnet-abcc.ncifcrf.gov/db/dbOrtho.php. We verified the overlap of these genes with those identified in treatment with tacrolimus, as described above. Additionally, we evaluated whether these genes were appropriately down- or upmodulated in the presence of the hypoglycemic agent.
In Vitro Molecular Validation of Optimal Hypoglycemic Agent in THP-1 Cells After Tacrolimus Exposure
Treatment of THP-1 and HepG2 cells with hypoglycemic agents was performed after exposing cells to tacrolimus at the final concentration of 120 nmol/L for 18 hours, followed by 24 hours exposure to 100 μM metformin (Sigma-Aldrich, Germany) or insulin 25 nM (Gibco, Thermo Fisher). RNA extraction was performed as described above. Real-time validation was performed as above, investigating the change in expression levels of IRS1, IRS2, IDH3A, and PKLR genes. Based on our bioinformatics analysis, these genes were predicted to be reverted by metformin from their dysregulation due to tacrolimus. Specific oligonucleotide primers were: IRS1 forward 5′-TGATGAACATCAGGCGCTGT-3′, reverse 5′-CCACCACAGAGTCATCCACC-3′, IRS2 forward 5′-GAAAAAGTGGCGGAGCAAGG-3′, reverse 5′- GGCGATCAGGTACTTGTGCT-3′, IDH3A forward 5′CGCGTGGATCTCTAAGGTCT-3′, reverse 5′-CAGCTGAAATTTCTGGGCCA-3′, PKLR forward 5′-CCCAATATTGTCCGGGTCGT-3′, reverse 5′-ACCACTAGGGAGATGAGCCC-3′, RPL13 forward 5′-CCGGGTTGGCTGGAAGTACC-3′, reverse 5′-CTTCTCGGCCTGTTTCCGTAG-3′. The ΔΔ Ct method was used to calculate the expressed fold-change of the genes of interest in the treated cells compared with control conditions.
Glucose Measurement in THP-1 Medium
The media collected from THP-1 and HepG2 cells exposed to tacrolimus (120 nM) for 18 hours and from THP-1 and HepG2 cells exposed to tacrolimus followed by insulin or metformin was quantified using Contour Next Blood Glucose Meters and Test Strips (Bayer, Germany).
Our search strategy for high-throughput datasets is depicted in Figure 1. There were 1113 publications in PubMed and 2446 high-throughput microarray gene expression datasets related to immunosuppressive drugs used in a solid organ transplant context identified on GEO. Three high-throughput gene expression datasets were eligible for inclusion in our study. The characteristics of these datasets, with the comparisons used and the type of tissue whose gene expression was elucidated, is documented in Table 1.
Dysregulated Genes and Pathways in Immunosuppression and Diabetes
We derived the list of dysregulated genes in the PBMCs, liver, and kidney of immunosuppressed patients. These genes were compared and contrasted with those significantly dysregulated in type 2 diabetes, with respect to overlap and direction of modulation as depicted in Figure 2. There were 571 genes in common between the immunosuppression and diabetes datasets, with 534 being upregulated and 37 being downregulated.
In Vitro Molecular Validation of Dysregulated Genes and IPA Analysis
Using the insulin and glucose metabolism gene arrays, we discovered 29 dysregulated genes with the same modulation (either up- or downregulated) in both the tacrolimus-treated and diabetes conditions in vitro. The pathways associated with these dysregulated genes are reported as from the SABiosciences webpage for each array and listed in Table 2. The upregulated genes and pathways were most related to Insulin signaling and insulin secretion by pancreatic β-cells. Insulin-like growth factor and receptor genes, as well as insulin, were significantly upregulated. Genes such as PCK1 related to gluconeogenesis and the TCA cycle were significantly downregulated. IPA network analysis listed 13 possible protein-protein interaction networks for the genes dysregulated by tacrolimus exposure, depicting involvement of the interactors in different biological functions and process (Table 3). Canonical pathways for the dysregulated genes were retrieved by IPA. Insulin and IGF-1 signaling were found not only to be significantly enriched but also activated (z score > +2) (Table 3). Type 2 diabetes was also predicted to be activated (Table 3), and PPAR signaling was predicted to be inhibited. Figure 3A shows the top network (pertaining to endocrine conditions, with score 45), illustrating the interactions between our dysregulated genes (red upregulated, green downregulated) and their protein interactors. The top network was merged and overlaid with related “functions and diseases” to determine genes associated with specific biological and pathological processes according to the IPA knowledge base. The molecular activity predictor tool predicted the crosstalk relationship among our genes and their interactors based on modulation. The Drug database was overlaid with the top network, and identified phenformin (sister compound of metformin) as a chemical compound able to target crucial genes in this network. Figure 3B shows insulin signaling pathway overlaid with the upregulated and downregulated genes in the arrays with diabetes. We then evaluated the overlap of all significantly dysregulated genes from the tacrolimus-treated THP-1 cells with genes dysregulated in diabetes (Figure 4A).
In Vitro Validation of Hypoglycemic Agents in Tacrolimus-treated Cells
IPA drug database and upstream analysis overlaid with our dysregulated genes identified 3 molecules targeting our genes: rosiglitazone, phenformin, and insulin. We then examined the gene expression changes in the context of treatment with the following hypoglycemic agents: metformin, rosiglitazone, and insulin (Table 4). High-throughput gene expression datasets from the GEO were interrogated for significantly dysregulated genes, and evaluated for overlap with the significantly dysregulated genes from our in vitro validation (Figure 4B). The in silico prediction showed metformin as the agent with the greatest overlap, and with the most appropriate reversion of gene expression of key diabetes genes, such as IRS1 and IRS2. Metformin was predicted to appropriately downregulate expression of these genes significantly upregulated in the context of tacrolimus treatment (Table 5), whereas it appropriately upregulated expression of down modulated genes such as IDH3A and PKLR (Table 6). In the same analysis, insulin did not optimally revert any of the dysregulated genes, and rosiglitazone suboptimally reverted gene expression. Treatment of THP-1 and HepG2 cells exposed to tacrolimus with metformin and insulin revealed that these drugs were able to revert IRS1, IRS2 IDH3A, and PKLR, but with differing efficiency. However, the crucial gene in the context of PTDM, IRS1, was most efficiently reverted in tacrolimus-exposed cells treated with metformin. This was reflected by 4.3-fold decrease in its expression compared with insulin, with a decrease of 1.2-fold only (Figure 4C). The hypoglycemic effect was also detected in the cell media of both THP-1 and HepG2 cells. Glycemia was more significantly increased with tacrolimus exposure in THP-1 cells (8.5 mmol/L, P = 0.0002, Student t test) and decreased significantly to 6 mmol/L in the same cells subsequently treated with metformin or insulin (Figure 4D, P = 0.0077, Student t test).
On the other hand, HepG2 cells were more responsive to insulin effect to decrease the glucose level from 20.6 mmol/L, after tacrolimus treatment, to 15.1 mmol/L (Figure 4D, P = 0.00009, Student t test).
Solid organ transplant recipients are commonly affected by PTDM, a condition known to have an adverse impact on long-term outcomes and patient survival. Through an integrative analysis approach, we elucidated the gene expression changes that occur in the setting of immunosuppression. These changes occur especially in genes related to insulin signaling and secretion and were best reverted with metformin. Therefore, the molecular effects of immunosuppressants that lead to increased risk of PTDM should be considered when deciding on the optimal hypoglycemic agent. This unique approach has previously been used to better understand the pathogenesis in other medical conditions from a more global viewpoint.17-21
We utilized high-throughput gene expression data from immunosuppressed solid organ transplant recipients and determined the key dysregulated genes in this context. These genes were then compared and contrasted with those genes expressed in patients with type 2 diabetes. We not only crosschecked the identity of the genes but also whether they were similarly and significantly up- or downregulated. KRAS, GRB2, PCK2, and BCL2L1 were highly upregulated and have been associated with pancreatic β-cell proliferation (affecting insulin production and secretion) and insulin signaling.22-24 We determined that pathways in insulin signaling and insulin secretion as a reflection of pancreatic β-cell function were highly represented in immunosuppressed conditions, namely tacrolimus, in our in vitro model. The identified genes were then validated in vitro. By using this approach, we were able to elucidate the metabolic effects of immunosuppression that likely contribute to development of PTDM. Genes related to insulin signaling were particularly important.
Previous literature on the molecular basis of PTDM has been limited, focusing on specific mechanisms rather than a global understanding. A seminal clinical study examined glucose metabolism in patients on immunosuppression, comparing the results of oral glucose tolerance tests and insulin levels between kidney transplant recipients and healthy controls. They discovered that impaired insulin secretion was the predominant pathophysiological feature after renal transplantation and suggested that interventions to preserve, maintain, or improve pancreatic β-cell function would be potentially beneficial.25 This is in keeping with our findings that genes involved in insulin secretion and pancreatic β-cell function are being significantly affected.
Genome-wide association studies have previously delved into the genetic susceptibility to PTDM. A Genome-wide association study of 26 kidney transplant recipients with PTDM and 230 controls identified 26 associated single nucleotide polymorphisms, 7 of which have been implicated in pancreatic β-cell apoptosis.26 This is concordant with our findings demonstrating the upregulation of genes involved in pancreatic β-cell function. Pathway analysis revealed that the P13K-AKT-mTOR signaling pathway, which is an insulin signaling pathway, had the highest enrichment score (P < 0.0001). A meta-analysis of 36 articles examining genetic associations in PTDM revealed genetic variants associated with PTDM at the 5% level of significance.27 CDKAL1 (rs10946398) was present in 696 individuals. The function of this gene is unknown, but single nucleotide polymorphisms in this gene have been correlated with type 2 diabetes.28 KCNQ1 (rs2237892), a potassium channel protein associated with hyperinsulinemia, was found in 1270 patients.29 IGF2BP2 (insulin-like growth factor 2 binding protein) and PPARγ were additional genes involved in the insulin signaling pathway and insulin sensitivity,27 also found to be dysregulated in our current integrative analysis.
We additionally verified whether the key affected genes were modulated in the opposite direction by hypoglycemic agents and discovered that metformin (compared with insulin and rosiglitazone) was best at appropriately reverting the expression of these genes. Metformin may be a preferred therapeutic agent for PTDM, given that it was the hypoglycemic agent that best reverted genes significantly dysregulated in the context of immunosuppression. This is in keeping with the known effect of metformin on the mTOR pathway, which is a known mediator of insulin signaling. Guidelines on management of PTDM have listed metformin among other hypoglycemic agents, although a glomerular filtration rate < 60 mL/min contraindicates its use.30 However, study of PTDM management has been mainly centered around the kidney transplant population.31 Clinical trials to assess the benefits of use of metformin have not been performed. Nonetheless, metformin has a host of beneficial effects on the various physiological aspects that affect transplant recipients: attenuation of abnormal glucose metabolism, weight neutrality, improved insulin resistance, subclinical inflammation, and endothelial dysfunction, lipid-lowering, cardiovascular protection, and cancer protection.32
In this study, we employed a novel approach to elucidate and validate the diabetogenic changes caused by immunosuppression. A similar bioinformatics approach has previously been used to investigate the genes involved in the development of insulinoma, and its overlap with diabetes.20 Additionally, bioinformatics and integrative analyses of genes in diabetes have been performed to identify the key genes and pathways in diabetes, along with drug repurposing.18,19 Drug development has been conducted by pharmaceutical companies for diabetes through successful use of the bioinformatics tools employed in our study.17 This approach provided us with insight into why immunosuppressants used in transplantation contribute to increased risk of PTDM, and what would be the optimal therapeutic agent for this condition based on reversion of global gene expression changes.
Our study admittedly has certain limitations, such as the small number of high-throughput datasets available to characterize the dysregulated genes in immunosuppressed conditions. These datasets were obtained from all transplant recipients on immunosuppression, and these patients had not been subcategorized into whether they had preexisting diabetes, PTDM, or no diabetes. Additionally, most were kidney transplant recipients, and racial differences could not be accounted for, given the lack of this associated information. A proportion of kidney transplant patients may have had preexisting or posttransplant diabetes, data that were not available to us. Nonetheless, our integrative analysis approach allowed us to identify overlapping genes in kidney and liver transplant patients that were common with diabetics, thereby obviating any interpatient heterogeneity. Subsequent in vitro validation confirmed the importance of these genes dysregulated by exposure to tacrolimus. Most of the patients in these datasets were on calcineurin inhibitors, and our results principally provide a window into PTDM engendered in this context. This is the reason for which we treated cells with tacrolimus to validate the genes involved in PTDM. Given that peripheral blood mononuclear cells and liver cells were used most commonly to assess gene expression in both transplant recipient and diabetic patient datasets, we used the THP-1 mononuclear cell line and HepG2 liver cell line for validation. We did not have access to a nonmalignant human pancreas cell line and therefore could not evaluate the effect of immunosuppressants directly on pancreas cells.
In conclusion, we have identified and validated key diabetogenic genes and pathways dysregulated in the context of immunosuppression using a novel bioinformatics approach. This offered a different, more global perspective of how immunosuppressants could lead to increased risk of PTDM in transplant recipients. Genes and pathways involved in insulin signaling and secretion as a reflection of pancreatic β-cell function were particularly altered. Metformin and insulin had differing effects on the crucially dysregulated genes, with metformin effecting the most appropriate reversion of these genes. These findings suggest that the effects of immunosuppressants on diabetes-associated genes should be considered when identifying the appropriate hypoglycemic agent for an individual patient. This will ultimately lead to a more precision medicine approach in the management of PTDM, thereby optimizing long-term outcomes in solid organ transplant recipients.
1. Kesiraju S, Paritala P, Rao Ch UM, et al. New onset of diabetes after transplantation - an overview of epidemiology, mechanism of development and diagnosis. Transpl Immunol. 2014; 30:52–58
2. Watt KD, Pedersen RA, Kremers WK, et al. Evolution of causes and risk factors for mortality post-liver transplant: results of the NIDDK long-term follow-up study. Am J Transplant. 2010; 10:1420–1427
3. Chanchlani R, Joseph Kim S, Kim ED, et al. Incidence of hyperglycemia and diabetes and association with electrolyte abnormalities in pediatric solid organ transplant recipients. Nephrol Dial Transplant. 2017; 32:1579–1586
4. Wyzgał J, Paczek L, Sańko-Resmer J, et al. Insulin resistance in kidney allograft recipients treated with calcineurin inhibitors. Ann Transplant. 2007; 12:26–29
5. Ali IH, Adberrahim E, Ben Abdelghani K, et al. Incidence and risk factors for post-renal transplant diabetes mellitus. Transplant Proc. 2011; 43:568–571
6. Davidson JA, Wilkinson A; International Expert Panel on New-Onset Diabetes after Transplantation. New-onset diabetes after transplantation 2003 international consensus guidelines: an endocrinologist’s view. Diabetes Care. 2004; 27:805–812
7. Sharif A, Hecking M, de Vries AP, et al. Proceedings from an international consensus meeting on posttransplantation diabetes mellitus: recommendations and future directions. Am J Transplant. 2014; 14:1992–2000
8. STARZL TE, MARCHIORO TL, RIFKIND D, et al. Factors in successful renal transplantation. Surgery. 1964; 56:296–318
9. Ojo AO. Cardiovascular complications after renal transplantation and their prevention. Transplantation. 2006; 82:603–611
10. Kasiske BL, Snyder JJ, Gilbertson D, et al. Diabetes mellitus after kidney transplantation in the United States. Am J Transplant. 2003; 3:178–185
11. Boudreaux JP, McHugh L, Canafax DM, et al. The impact of cyclosporine and combination immunosuppression on the incidence of posttransplant diabetes in renal allograft recipients. Transplantation. 1987; 44:376–381
12. Charco R, Cantarell C, Vargas V, et al. Serum cholesterol changes in long-term survivors of liver transplantation: a comparison between cyclosporine and tacrolimus therapy. Liver Transpl Surg. 1999; 5:204–208
13. Rodriguez-Rodriguez AE, Triñanes J, Velazquez-Garcia S, et al. The higher diabetogenic risk of tacrolimus depends on pre-existing insulin resistance. a study in obese and lean zucker rats. Am J Transplant. 2013; 13:1665–1675
14. Drachenberg CB, Klassen DK, Weir MR, et al. Islet cell damage associated with tacrolimus and cyclosporine: morphological features in pancreas allograft biopsies and clinical correlation. Transplantation. 1999; 68:396–402
15. Lamming DW, Ye L, Katajisto P, et al. Rapamycin-induced insulin resistance is mediated by mtorc2 loss and uncoupled from longevity. Science. 2012; 335:1638–1643
16. Rahmati S, Abovsky M, Pastrello C, et al. Pathdip: an annotated resource for known and predicted human gene-pathway associations and pathway enrichment analysis. Nucleic Acids Res. 2017; 45:D419–D426
17. Dangi-Garimella S. The role of bioinformatics in diabetes drug development--and precision medicine. Am J Manag Care. 2014; 20:E1
18. Wang Q, Zhao Z, Shang J, et al. Targets and candidate agents for type 2 diabetes treatment with computational bioinformatics approach. J Diabetes Res. 2014; 2014:763936
19. Zhang M, Luo H, Xi Z, et al. Drug repositioning for diabetes based on ‘omics’ data mining. Plos One. 2015; 10:e0126082
20. Wang H, Bender A, Wang P, et al. Insights into beta cell regeneration for diabetes via integration of molecular landscapes in human insulinomas. Nat Commun. 2017; 8:767
21. Zhou X, Yu S, Su J, et al. Computational study on new natural compound inhibitors of pyruvate dehydrogenase kinases. Int J Mol Sci. 2016; 17:340
22. Stewart AF, Hussain MA, García-Ocaña A, et al. Human β-cell proliferation and intracellular signaling: part 3. Diabetes. 2015; 64:1872–1885
23. Bateman AC, Turner SM, Thomas KS, et al. Apoptosis and proliferation of acinar and islet cells in chronic pancreatitis: evidence for differential cell loss mediating preservation of islet function. Gut. 2002; 50:542–548
24. Ackermann JA, Radtke D, Maurberger A, et al. Grb2 regulates B-cell maturation, B-cell memory responses and inhibits B-cell ca2+ signalling. Embo J. 2011; 30:1621–1633
25. Hecking M, Kainz A, Werzowa J, et al. Glucose metabolism after renal transplantation. Diabetes Care. 2013; 36:2763–2771
26. McCaughan JA, McKnight AJ, Maxwell AP. Genetics of new-onset diabetes after transplantation. J Am Soc Nephrol. 2014; 25:1037–1049
27. Benson KA, Maxwell AP, McKnight AJ. A huge review and meta-analyses of genetic associations in new onset diabetes after kidney transplantation. Plos One. 2016; 11:e0147323
28. Pascoe L, Tura A, Patel SK, et al; RISC Consortium; U.K. Type 2 Diabetes Genetics Consortium. Common variants of the novel type 2 diabetes genes CDKAL1 and HHEX/IDE are associated with decreased pancreatic beta-cell function. Diabetes. 2007; 56:3101–3104
29. Torekov SS, Iepsen E, Christiansen M, et al. KCNQ1 long QT syndrome patients have hyperinsulinemia and symptomatic hypoglycemia. Diabetes. 2014; 63:1315–1325
30. Ghisdal L, Van Laecke S, Abramowicz MJ, et al. New-onset diabetes after renal transplantation: risk assessment and management. Diabetes Care. 2012; 35:181–188
31. Jenssen T, Hartmann A. Emerging treatments for post-transplantation diabetes mellitus. Nat Rev Nephrol. 2015; 11:465–477
32. Sharif A. Should metformin be our antiglycemic agent of choice post-transplantation? Am J Transplant. 2011; 11:1376–1381