Modeling amyotrophic lateral sclerosis in pure human iPSc-derived motor neurons isolated by a novel FACS double selection technique

Highlights

• We report the isolation of pure human iPSc-derived motor neurons by p75^NTR HB9::RFP double FACS selection.

• These motor neurons survive and grow in culture, and are electrically excitable.

• They contain disease-relevant subtypes and form functional connections with co-cultured myotubes.

• In response to mutant SOD1 astrocytes, they undergo massive cell death and axon disintegration.

• The approach allows modeling of the major phenotypic defects in amyotrophic lateral sclerosis (ALS).

Abstract

Amyotrophic lateral sclerosis (ALS) is a severe and incurable neurodegenerative disease. Human motor neurons generated from induced pluripotent stem cells (iPSc) offer new perspectives for disease modeling and drug testing in ALS. In standard iPSc-derived cultures, however, the two major phenotypic alterations of ALS—degeneration of motor neuron cell bodies and axons—are often obscured by cell body clustering, extensive axon criss-crossing and presence of unwanted cell types. Here, we succeeded in isolating 100% pure and standardized human motor neurons by a novel FACS double selection based on a p75^NTR surface epitope and an HB9::RFP lentivirus reporter. The p75^NTR/HB9::RFP motor neurons survive and grow well without forming clusters or entangled axons, are electrically excitable, contain ALS-relevant motor neuron subtypes and form functional connections with co-cultured myotubes. Importantly, they undergo rapid and massive cell death and axon degeneration in response to mutant SOD1 astrocytes. These data demonstrate the potential of FACS-isolated human iPSc-derived motor neurons for improved disease modeling and drug testing in ALS and related motor neuron diseases.

Introduction

Motor neuron diseases such as amyotrophic lateral sclerosis (ALS) are severe and most often rapidly fatal neurodegenerative disorders. ALS is characterized by progressive degeneration of motor neurons in the spinal cord, brainstem and cerebral cortex. ALS can be caused by mutations in more than 25 genes including C9ORF72, SOD1 and TARDBP or occurs sporadically (Robberecht and Philips, 2013). Animal models of ALS and corresponding cellular models have given important insights into disease mechanisms but also bear major limitations. Transgenic mutant SOD1 mice, for instance, develop motor neuron disease only when the mutant enzyme is expressed at high non-physiological levels (Gurney et al., 1994) and their response to numerous pharmacological compounds is not predictive for human ALS (Turner and Talbot, 2008, Mitsumoto et al., 2014).

Human motor neurons generated from induced pluripotent stem cells (iPSc) (Takahashi et al., 2007) offer a new alternative for disease modeling and drug testing in ALS. Recent studies have indeed provided evidence for molecular and functional alterations in motor neurons derived from ALS patients (Bilican et al., 2012, Chen et al., 2014, Egawa et al., 2012, Haeusler et al., 2014, Kiskinis et al., 2014, Sareen et al., 2013). Yet, modeling the two major phenotypic alterations of ALS, motor neuron cell body death and axon degeneration, faces important still unresolved problems in iPSc-derived cultures. First, the efficiency of motor neuron generation remains variable between protocols and from one iPSc clone to another (Amoroso et al., 2013, Qu et al., 2014). Second, motor neuron cell bodies tend to form large clusters which are inaccessible for analysis of cell survival (Chen et al., 2014, Dimos et al., 2008, Ebert et al., 2009, Hu and Zhang, 2009, Kiskinis et al., 2014, Maury et al., 2015, Qu et al., 2014, Sareen et al., 2013). Motor neuron axons often also display extensive criss-crossing and intermingling with axons from other neuronal types thus complicating their identification and analysis (Amoroso et al., 2013, Bilican et al., 2012, Boulting et al., 2011, Burkhardt et al., 2013, Chen et al., 2014, Egawa et al., 2012, Haeusler et al., 2014, Kiskinis et al., 2014, Maury et al., 2015, Qu et al., 2014, Sareen et al., 2013). Third, the proportion of disease-relevant motor neuron subtypes remains most often undetermined in human iPSc-derived cultures (Bilican et al., 2012, Boulting et al., 2011, Egawa et al., 2012, Hester et al., 2011, Karumbayaram et al., 2009, Mitne-Neto et al., 2011, Qu et al., 2014, Serio et al., 2013) but see Amoroso et al. (2013). Fourth, the presence of unwanted cell types such as neural precursors, various types of neurons (Amoroso et al., 2013, Boulting et al., 2011, Chen et al., 2014, Dimos et al., 2008, Ebert et al., 2009, Karumbayaram et al., 2009, Naujock et al., 2014, Qu et al., 2014, Sareen et al., 2012) or astrocytes (Haidet-Phillips et al., 2011, Re et al., 2014) may induce confounding effects on motor neuron survival. Likewise, pharmacological compounds may act indirectly through these cell types rendering difficult the interpretation of their effects. Generating pure and standardized human motor neuron cultures is therefore an important step towards robust ALS modeling and drug testing.

Fluorescence-activated cell sorting (FACS) represents a powerful means to isolate the desired neuronal populations. Yet, there are few well-characterized genetic reporters and surface markers for FACS isolation of motor neurons (Amoroso et al., 2013, Egawa et al., 2012, Kiskinis et al., 2014, Wichterle et al., 2002). In addition, FACS-isolated human motor neurons were reported to survive and grow poorly in culture (Egawa et al., 2012). We here developed a novel procedure for motor neuron isolation by combining a lentiviral reporter vector harboring the 3.6 kb minimal HB9 promoter (Marchetto et al., 2008), which drives RFP expression and a monoclonal antibody against the cell surface receptor p75^NTR (Ross et al., 1984). Using p75^NTR/HB9::RFP FACS double selection, we succeeded in isolating high numbers of 100% pure and functional human iPSc-derived motor neurons which comprise ALS-relevant motor neuron subtypes. The FACS-isolated motor neurons survive and grow well in low-density cultures without forming clusters or entangled axons. By contrast, they undergo increased cell death and form prominent axon blebs in response to mutant SOD1 astrocytes, indicating their potential for improved ALS modeling.