Studying human disease using human neurons

Abstract

Utilizing patient derived cells has enormous promise for discovering new drugs for diseases of the nervous system, a goal that has been historically quite challenging. In this review, we will outline the potential of human stem cell derived neuron models for assessing therapeutics and high-throughput screening and compare to more traditional drug discovery strategies. We summarize recent successes of the approach and discuss special considerations for developing human stem cell based assays. New technologies, such as genome editing, offer improvements to help overcome the challenges that remain. Finally, human neurons derived from patient cells have advantages for translational research beyond drug screening as they can also be used to identify individual efficacy and safety prior to clinical testing and for dissecting disease mechanisms. This article is part of a Special Issue entitled SI: Exploiting human neurons.

Keywords: Drug discovery; Human neurodegenerative disease; Stem cells.

Figure 1

hPSCs with Genotypes of interest can be generated in 2 principle ways. hIPSCs can be derived through ectopic expression of reprogramming factors in somatic cells from a patient known to carry the disease genotype. In this model control IPS lines are compared to patient derived iPS cell lines. Isogenic disease lines can be derived through the introduction of disease loci of interest through genome editing (knockout/homologous recombination). In this model donor lines are compared to the derived isogenic lines harboring the genetic locus. Alternatives might exist in miss-expression models. Independent of the source, the cells have to be differentiated into the affected cell types and a phenotype has to be established. This phenotype can just manifest itself in the derived neurons (a good example of this is the low expression of SMN in SMA-IPS cell lines) or specific treatments, like aging/oxidative stress/mitochondrial stress that have to be employed in order to establish a specific phenotype. It is of importance to note that the window of screening is established through the difference between the control lines and the disease lines. Several other controls can be employed in the pre-screen phase to ensure a specific phenotype. For example in our scheme the disease lines can be differentiated into a neuronal subtype that is not affected in the disease and should not show the specific phenotype found in the affected cell type. Once the disease model is validated, an assay has to be developed that can capture rescue/reversal or alleviation of the phenotype. A screen has to be conducted and analyzed. Potential hits have to be identified, confirmed and follow-up experiments have to be conducted to guide the discovery process towards the development of new therapeutics.

Figure 2

Screening in hPSC disease models relies on the established disease model and an assay that can be used to read-out the effect of treatments. High-Throughput assay read-outs are used to quantify the fluorescence of a whole well using a plate-reader (left). Reporter genes containing fluorophores, luciferase assays, fluorescent dyes or immunocytochemistry can be employed. High-Content-Screens employ the same fluorescent dyes but can also be used to capture bright-field morphology changes. Automated image capture and analysis are used to capture variables on a cell-by-cell basis or even changes in cell organelle structure (middle left). A novel approach is the use of calcium imaging or microelectrode arrays that can capture physiological parameters in response to small molecules (middle right), in the future sequencing of the transcriptome will be used to a.) capture the effect of single compound/gene function on the disease model and b.) whole genome functional genomics and phenotype/tag enrichment strategies (right).