I. Regulation of insulin production from transcription to translation

A. MicroRNAs

MicroRNAs (miRNAs) are short noncoding RNAs of 21 to 23 nucleotides that are created by the enzyme Dicer. MiRNAs mediate RNA silencing and posttranscriptional regulation of gene expression by interacting with their respective targets.4 Their implication in β-cell compensation (ie, increased insulin production and release in response to higher tissues requirements) has been demonstrated during pregnancy, obesity, and T2D development.5,6 As a proof of concept, Dicer knockout in β-cells leads to T2D in mice.7 Dr Eliasson’s lecture reviewed the essential role of miRNAs in defining β-cell identity through regulating genes involved in insulin biosynthesis, exocytic processes, glucose uptake and metabolism, electrical activity, and Ca2+ influx, as well as by downregulating “disallowed genes” (eg, genes involved in β-cell dedifferentiation) (Figure 1). The system is extremely complex as one miR has several targets, and one target can be regulated by several miRs. Nevertheless, based on Dr Eliasson’s work and the work of many other scientists, it is now well accepted that specific miRNAs are involved in the maintenance of α-cell and β-cell phenotypic identities.4

In particular, miR-375, one of the most abundant miRNAs, is involved in several cellular functions that maintain β-cell identity. Interestingly, its expression is unchanged in T2D donor islets. On the other hand, the less abundant miRNAs miR-335, miR-130a/b, and miR-152 are deregulated in T2D islets and play key roles in regulating β-cell metabolism.8,9

B. Long noncoding RNA

The noncoding part of RNA represents more than 98% of the total human genome. Tens of thousands of mammalian transcripts longer than 200 nucleotides with low protein-coding potential have been discovered by transcriptome surveys. Of these long noncoding RNAs (lncRNAs), a small fraction controls gene expression by modulating chromosomal structure, transcription, splicing, mRNA transport and stability, or translation.10

Most human lncRNAs are cell-type specific and evolutionarily conserved. Interestingly, in β-cells, several lncRNAs are regulated by extracellular glucose concentrations. These observations suggest that, like miRNAs, lncRNAs play a potential role in the functional adaptation of β-cells to increased insulin secretory demands.11 During his lecture, Anthony Beucher described how network analysis and transcript knockdown experiments have helped identify several lncRNAs involved in the regulation of insulin production and release. For example, the lncRNA HI-LNC71 affects local tri-dimensional chromatin structure and the transcription of PDX1, a gene encoding a key transcription factor for β-cell differentiation. For this reason, it has been renamed PLUTO, for PDX1 locus upstream transcript.10 Both PLUTO and PDX1 are downregulated in islets from T2D donors and

impaired glucose tolerance donors.10 Figure 2 recapitulates the β-cell genes that are regulated by islet lncRNAs (Figure 2A), and, as a proof of concept, knocking down three lncRNAs leads to impaired insulin secretion (Figure 2B). Recent studies demonstrated that islet transcription factors regulate transcription by targeting clusters of enhancers and that multiple transcription factors can bind one enhancer.12 Knocking down islets lncRNAs and transcription factors suggested that they regulate similar genes, which has been demonstrated by a gene-set enrichment analysis.10

C. Transfer RNAs

Protein synthesis requires the coordinated assembly of initiation factors, ribosomal RNAs, and mRNA, followed by the sequential recruitment of elongation factors and transfer RNAs (tRNA). tRNAs are indeed adaptor molecules composed of RNA, typically 76 to 90 nucleotides in length, that help decode a mRNA sequence into the amino acid sequence of proteins. Efficient and accurate protein translation is essential to produce insulin in pancreatic β-cells. In addition, several experiments from 1970 to 1980 showed that the initial phase of insulin synthesis in response to glucose stimulation exclusively depends on the translation of existing preproinsulin mRNA; whereas, the late phase requires newly transcribed mRNA.13,14
For these reasons, proper chemical modifications to tRNA using tRNA-modifying enzymes are crucial for ensuring proper insulin synthesis. Wide varieties of chemical modifications, which are posttranscriptionally catalyzed by tRNA-modifying enzymes, have been identified, and their aberrant modifications could lead to

impaired insulin secretion and T2D. Dr Tomizawa investigated the role of Cdk5 regulatory subunit associated protein 1-like 1 (CDKAL1), an enzyme important for lysine translation (please note that lysine within the proinsulin amino acid sequence is important for the correct cleavage into C-peptide and insulin) (Figure 3).15,16 CDKAL1 belongs to the prokaryotic tRNA methylthiotransferase family and contains multiple functional domains, CDKAL1-mediated modification has a profound impact on the codon-anticodon tRNALys(UUU) interaction.17

Thus, CDKAL1 variants could influence the insulin response and the risk of T2D, as it has been shown in the Icelandic population.18 Of note, insulin secretagogues commonly used in T2D patients, such as sulfonylureas and glinides, could be particularly harmful for people with deregulated tRNA modifications because of the increased production of aberrant insulin. Further studies are needed to propose precision medicine for such particular at-risk patients.

D. Alternative splicing and β-cell dysfunction

Alternative splicing, or differential splicing, is a posttranslational mechanism by which a single gene generates different mRNA and thereby different protein isoforms. In this process, specific exons of a gene may be included within or excluded from the final processed mRNA produced from that gene,19 thus enhancing protein diversity. Genome-wide studies have estimated that around 95% of human genes undergo alternative splicing, generating an average of four isoforms per gene.20 In β-cells, specific RNA-binding proteins have been described for their role in the regulation of insulin secretion, and a transcriptome analysis has suggested that splicing alterations contributed to β-cell failure in the etiology of both T1D and T2D.21,22 In addition, it is now accepted that inflammation or T2D susceptibility genes regulate alternative exon networks. Taking this data together, a better understanding of alternative splicing regulators in β-cells is of great interest and may provide new therapeutic targets.

Alternative splicing in the β-cell is very similar to the process in neuronal tissues (eg, the brain). Among the splicing factors already known to be preferentially expressed in the brain and β-cells, Eizirik et al found that NOVA1 and 2, RBFOX1, RBFOX2, and ELAVL4 affect β-cell function and/or survival.23 In addition, alternative splicing contributes to autoimmunity through the generation and presentation of specific antigens, thus probably contributing to β-cell death in T1D. Juan-Mateu et al highlighted the role of “neuron-specific” splicing factors in β-cells24 and suggested that the splicing regulator SRp55 is a key splicing factor in human pancreatic β-cells.25