Small RNA-Seq to reveal islet-microRNA modulation under stress and transplantation conditions
PhD student: Bassam Aljani
Supervisor at TUD: Ezio Bonifacio, Anne Eugster
Supervisor at KCL: Pratik Choudhary, Stephanie Amiel, Peter Jones
Start date: 01.08.2017 Date of defense: 05.10.2021 Joint PhD
Replacement of β-cells via islet transplantation is a potential therapeutic approach for individuals with type 1 diabetes that is hard-to-control with conventional approaches. Post-transplantation, Islets are prone to stress such as hypoxia and inflammation. It is of utmost importance to diagnose β-cell insults before massive cell loss occur and hence rapid intervention can be taken. Currently utilized markers often fail to detect β-cell stress and death in timely manner and only indicate massive changes in β-cell function. Thus, biomarkers reflect the actual rate of β-cell stress and death are required.
MiRNAs are short (21–25 nucleotides in length) single-stranded RNA molecules, originated from a hairpin-like structure. miRNAs are considered as gene silencers post-transcriptionally. Studies have reported alteration in circulating miRNAs in type 1 diabetes and their implication in the pathogenesis of this disease. Such miRNAs could serve as circulating biomarkers for β-cell stress and death. After human islet transplantation, islets exposed to stresses release miRNAs potentially detectable in the circulation and reflecting the rate of islets stress or death.
I aimed to identify circulating miRNAs that could be reflective of β-cell stress and death. I analyzed potential miRNA biomarkers in the circulation of diabetic animal model as well as in the context in vitro islet stress model.
To establish a panel of potential islet specific miRNAs I reviewed profiling studies in the literature and validated them using RT-qPCR on human/mouse islets and serum. For miRNA profiling, I used small RNA sequencing in mouse islets and serum. Previously used circulating reference miRNAs were identified through the literature and tested in healthy human serum samples as well as in control- and STZ-treated mice. Reference miRNAs were analyzed using NormFinder and/or RefFinder algorithms to identify stably expressed miRNAs. For diabetes induction in vivo, BL6 mice were treated with STZ over 7 or 4 days experimental time course. MiRNAs were quantified in the circulation before and after STZ treatment using RT-qPCR. In the in vitro islet stress model, human or mouse islets were treated with a cocktail of recombinant IL-1β, TNF-α, and IFN-γ and/or hypoxia (1% O2). The supernatant of treated and untreated islets was collected, centrifuged, and filtered. MiRNAs were quantified in the supernatant after 3H, 6H, and 24H after cytokines and/or hypoxia treatment. Caspase-3 activity was measured in the islets to assess apoptosis in parallel with miRNA induction. Basal (2mM) and stimulated insulin secretion was quantified using ELISA in the supernatant of islets treated as described above. KEGG enrichment analysis was performed on significantly induced miRNAs in order to identify targeted pathways.
In mouse, a panel of 60 miRNAs was established through the validation of literature-selected miRNAs and NGS. Reference miRNAs for RT-qPCR data normalization of circulating miRNAs were analyzed and 2 miRNAs were selected as normalizers using NormFinder. In human 15 miRNAs were identified after validation of literature- selected miRNAs. 5 miRNAs were identified as reference miRNAs after confirmation of RefFinder and NormFinder analysis. The measurements of circulating miRNAs at day -14, 1, 3, and 4 resulted in the selection of 24 miRNAs that increased in the diabetic STZ-treated mice. The quantification of miRNAs in the circulation of STZ-treated animals at day 4 revealed 18 miRNAs increased significantly in the diabetic STZ-treated animals. Three miRNAs increased in both, diabetic- and non-diabetic STZ-treated mice after 4 days of STZ treatment. Jointly, a panel of 80 miRNAs was established after adding NGS-top 49 upregulated miRNAs in islets as compared to serum. The later panel was measured in the in vitro islet stress model. Of which, 64 miRNAs increased in the supernatant of islets, treated with high or low dose cytokines and/or hypoxia, as compared to untreated islets over 3H, 6H, and 24H. Those 64 miRNAs were the base to establish human miRNA panel for further validation in pre- and post-transplantation samples. Adding miRNAs identified in human study, human reference miRNAs, and others identified in the in vivo mouse model, a panel of 87 miRNAs was finally established. The pathway analysis revealed that cytokine-induced miRNAs were targeting genes primarily involved in MAPK signaling pathway. While combination treatment-induced miRNAs were targeting genes primarily involved in PI3K-Akt and MAPK signaling pathway.
MiRNAs are promising candidate biomarkers. I showed that miRNAs are detectable in the circulation of mice treated with β-cell toxin (STZ) up to 4 days after injection. I identified miRNAs released from islets in response to cytokines and/or hypoxia in vitro and examined their target pathways. The work in the in vivo and in vitro models allowed the identification of potential miRNA biomarkers of islet stress and death. Those miRNAs could be used to monitor islet graft post-transplantation. Further validation of those miRNAs is required in the context of islet transplantation in human samples.