Neuron and beta-cell evolution: Learning about neurons is learning about beta-cells


  • Daniel Eberhard

    Corresponding author
    1. Paul Langerhans Group for Beta Cell Biology, German Diabetes Center (DDZ), Düsseldorf, Germany
    2. German Center for Diabetes Research (DZD), Germany
    • Institute of Metabolic Physiology, Heinrich-Heine University, Düsseldorf, Germany
    Search for more papers by this author

Corresponding author:

Daniel Eberhard


Neurons and insulin-producing pancreatic beta-cells share multiple common features although they arise from different germ layers [1]. In analogy to neurons, beta-cells release insulin by membrane depolarization reminiscent of synaptic neurotransmitter release in neurons and produce the neurotransmitter gamma aminobutyric acid (GABA) in synaptic-like vesicles. Although expression of many individual genes has been shown in neurons and beta-cells, an estimation of the total amount of genes commonly expressed in these cell types has just been addressed recently [2, 3]. In a genome-wide mRNA expression scan, profiles of human and rodent beta-cells were compared to non-beta-cell tissues to find genes selectively expressed in beta-cells [2]. Interestingly, 15% of these conserved beta-cell markers were also expressed in neuronal tissue coding for proteins involved in neurotransmitter transport, synaptic vesicle formation and brain development [2].

To explain why ectodermal-derived neurons and endodermal-derived beta-cells share many common features, Arntfield and van der Kooy have recently speculated that beta-cells may have “borrowed from the brain”, i.e. their ancestors have deployed parts of the neuronal developmental program thereby creating an example of convergent evolution [1]. This idea is consistent with the fact that multiple transcription factors are expressed in the neuronal and the pancreatic endocrine lineage including Neurogenin 3 (Ngn3) and NeuroD, both essential for brain and pancreas development. Arntfield et al. have further suggested that the evolution of (modern) beta-cells may have resulted from stochastic effects, i.e. mutation, activating neuronal master transcription factors in gut cells [1]. However it remains unclarified which cell type may have first adopted the neuronal program, i.e. a more unspecified gut cell or a pre-differentiated endocrine cell. Recently it has been suggested that beta-cell progenitors co-opt the neural program rather late during their differentiation based on observations on the level of epigenetics [3]. Interestingly, neuronal genes are still repressed in multipotent pancreatic progenitors that can give rise to exo- or endocrine tissue suggesting that an activation of neuronal gene activity occured later in the endocrine lineage [3]. This would indicate that a pre-differentiated multipotent pancreatic progenitor adopted the neuronal program rather than a more unspecified cell.

The notion that neurons and beta-cells are similar has stimulated scientists to transfer principles from neurobiology to understand beta-cell function and to find drugs to support beta-cell function under stress conditions as observed in type 2 diabetes mellitus. For example, screening of a chemical library for neurogenic activators has recently led to the identification of a potent inducer of NeuroD, supporting insulin production in pancreatic beta-cells [4]. Moreover, known neuroprotective mechanisms may also be essential in beta-cells as observed in the case of the Parkinson-related antioxidant protein DJ-1 [5]. DJ-1 preserves mitochondrial integrity in dopaminergic neurons that are vulnerable to reactive oxygen species. In analogy to neurons, DJ-1 also protects the mitochondrial function of beta-cells suggesting that beta-cells and neurons share similar oxidative stress defence mechanisms. Thus, the study of common principles in neurons and beta-cells may provide a basis for prevention and treatment of diabetes mellitus and neurodegenerative diseases.