Pharmacological rescue of mitochondrial and neuronal defects in SPG7 hereditary spastic paraplegia patient neurons using high throughput assays

Abstract

SPG7 is the most common form of autosomal recessive hereditary spastic paraplegia (HSP). There is a lack of HSP-SPG7 human neuronal models to understand the disease mechanism and identify new drug treatments. We generated a human neuronal model of HSP-SPG7 using induced pluripotent stem (iPS) cell technology. We first generated iPS cells from three HSP-SPG7 patients carrying different disease-causing variants and three healthy controls. The iPS cells were differentiated to form neural progenitor cells (NPCs) and then from NPCs to mature cortical neurons. Mitochondrial and neuronal defects were measured using a high throughout imaging and analysis-based assay in live cells. Our results show that compared to control NPCs, patient NPCs had aberrant mitochondrial morphology with increased mitochondrial size and reduced membrane potential. Patient NPCs develop to form mature cortical neurons with amplified mitochondrial morphology and functional defects along with defects in neuron morphology - reduced neurite complexity and length, reduced synaptic gene, protein expression and activity, reduced viability and increased axonal degeneration. Treatment of patient neurons with Bz-423, a mitochondria permeability pore regulator, restored the mitochondrial and neurite morphological defects and mitochondrial membrane potential back to control neuron levels and rescued the low viability and increased degeneration in patient neurons. This study establishes a direct link between mitochondrial and neuronal defects in HSP-SPG7 patient neurons. We present a strategy for testing mitochondrial targeting drugs to rescue neuronal defects in HSP-SPG7 patient neurons.

Keywords: cortical neurons; hereditary spastic paraplegia (HSP); high throughput imaging (HTI); induced pluripotent stem (iPS) cell; mitochondria.

Figure 1

Neuronal and mitochondrial phenotypes in live HSP-SPG7 patient neural progenitors and mature cortical neurons. High throughput imaging and analysis was used to evaluate neuron and mitochondrial morphology. (A) Shows the timeline to generate neural progenitors and mature cortical neurons. (B–G) Images of live control and patient neural progenitors and mature cortical neurons labelled with calcein – to identify viable cells, TMRM – to identify mitochondrial and hoechst – to identify nucleus. (H–M) Images of the cells presented in (B–G) with with TMRM and Hoechst label without the calcein label. (N–S) Neurites were segmented and identified using automated image analysis in control and patient neural progenitors and mature cortical neurons. (T–X) Multiple parameters of neurite morphology were analyzed in control and patient neural progenitors and mature cortical neurons. These parameters include neurite length (T), extremities (U), branching (V), segments (W) and roots (X). (Y–B1) Multiple parameters of mitochondrial morphology and membrane potential were analyzed in control and patient neural progenitors and mature cortical neurons. These parameters include (Y) mitochondrial area, (Z) perimeter, (A1) length, (B1) width and TMRM intensity (C1). (D1,E1) Cell viability (D1) and (E1) neurite degeneration in Day 10 mature control and patientneurons. Data is presented as Mean ± SEM. Scale bar: 100 μm.

Figure 2

Neuronal phenotypes confirmed in fixed and immunolabelled HSP-SPG7 patient neurons. (A–D) Control and patient mature cortical neurons express Tuj1 – neuronal marker and mature cortical neuronal markers - TBR1 (A,B) and CTIP2 (C,D). (E–G) The proportion of Tuj1, TBR1 and CTIP2 positive cells are comparable between controls and patients. (H–L) Multiple parameters of neurite morphology were analyzed in control and patient mature cortical neurons. These parameters include neurite (H) roots, (I) extremities, (J) branching, (K) length and (L) segments. (M) Axonal degeneration of control and patient neurites in mature cortical neurons. (N,O) Paraplegin expression was measured in control and patient mature cortical neurons. Data is presented as Mean ± SEM. Scale bar: 100 μm.

Figure 3

RNA-Seq analysis of HSP-SPG7 patient neurons. RNA-Seq analysis was performed using patient and healthy control neurons to identify gene expression pathways affected in HSP-SPG7 patient neurons. (A) Multidimensional scaling analysis was used to visualize the level of similarity in the gene expression of the patient and control neurons. (B) Smearplot shows a large number of differently expressed genes. (C) Pathway enrichment analysis was performed to understand which pathways/gene networks the differentially expressed genes are implicated in. Synaptic pathways was FIGURE 3 (Continued)downregulated in patient neurons. (D–F) List of genes that expressed lower in patient neurons compared to control neurons in the GABAergic synapse (D), Glutamergic synapse (E) and synaptic vesicle cycle pathway (F). (G,H) Western blot analysis and (I) RNA-Seq consistently showed reduced expression of synaptophysin expression. (J–M) RT-qPCR validated the findings of RNA-Seq based GAD1 and GAD2 gene expression. (N–Q) Whole cell patch clamping was used to measure neuronal synaptic activity in control and patient neurons. (R,S) List of genes that expressed higher in patient neurons compared to control neurons in the oxidative stress related p53 (R) and cellular senescence (S).

Figure 4

Pharmacological rescue of neuronal and mitochondrial phenotypes in HSP-SPG7 patient neurons. (A) Shows the Bz-423 drug treatment timeline. (B–D) Live untreated control, untreated patient and Bz-423 treated patient neurons labelled with calcein – to identify cells, TMRM – to identify mitochondrial and hoechst – to identify nucleus. (E–G) Neurites were segmented and identified using automated image analysis in untreated control, FIGURE 4 (Continued)untreated patient and Bz-423 treated patient neurons. (H–L) Multiple parameters of neurite morphology were analyzed in untreated control, untreated patient and Bz-423 treated patient neurons. These parameters included neurite (H) length, (I) roots, (J) branching, (K) segments and (L) extremities. (M–O) Multiple parameters of mitochondrial morphology and membrane potential were analyzed in untreated control, untreated patient and Bz-423 treated patient neurons. These parameters included mitochondrial (M) area, (N) perimeter and (O) length and (P) TMRM intensity. (Q–R) Viability and axonal degeneration of untreated control, untreated patient and Bz-423 treated patient neurons. Data is presented as Mean ± SEM. Scale bar: 100 μm.