Reactive Astrocytes Derived From Human Induced Pluripotent Stem Cells Suppress Oligodendrocyte Precursor Cell Differentiation

Abstract

Astrocytes are instrumental in maintaining central nervous system (CNS) homeostasis and responding to injury. A major limitation of studying neurodegenerative diseases like multiple sclerosis (MS) is lack of human pathological specimens obtained during the acute stages, thereby relegating research to post-mortem specimens obtained years after the initiation of pathology. Rodent reactive astrocytes have been shown to be cytotoxic to neurons and oligodendrocytes but may differ from human cells, especially in diseases with genetic susceptibility. Herein, we purified human CD49f⁺ astrocytes from induced pluripotent stem cells derived from individual patient and control peripheral leukocytes. We compared TNF and IL1α stimulated human reactive astrocytes from seven persons with MS and six non-MS controls and show their transcriptomes are remarkably similar to those described in rodents. The functional effect of astrocyte conditioned media (ACM) was examined in a human oligodendrocyte precursor cell (OPC) line differentiation assay. ACM was not cytotoxic to the OPCs but robustly inhibited the myelin basic protein (MBP) reporter. No differences were seen between MS and control stimulated astrocytes at either the transcript level or in ACM mediated OPC suppression assays. We next used RNAseq to interrogate differentially expressed genes in the OPC lines that had suppressed differentiation from the human ACM. Remarkably, not only was OPC differentiation and myelin gene expression suppressed, but we observed induction of several immune pathways in OPCs exposed to the ACM. These data support the notion that reactive astrocytes can inhibit OPC differentiation thereby limiting their remyelination capacity, and that OPCs take on an immune profile in the context of inflammatory cues.

Keywords: astrocyte; induced pluripotent stem cell (iPSC); multiple sclerosis; neurotoxicity; oligodendrocyte.

FIGURE 1

Human induced pluripotent stem cell derived astrocytes from PwMS and NMSC respond similarly to TNF + IL1α stimulation. (A) Schematic depicting experimental design. (B) Representative images showing stages of astrocyte differentiation from a PwMS. (C) Principle component analysis of top 500 most variable genes. (D) Volcano plot depicting results of differential expression testing comparing +TI treated astrocytes with –TI controls with a selection of the most significant genes labeled. (E) Heatmap showing clustering and relative expression of genes previously reported as characteristic of distinct sub-types of reactive astrocytes. Genes are classified as being previously reported as pan-reactive (P), associated with neurotoxic (N), or associated with hypoxic conditions (H). Samples are of Type C (NMSC) or M (PwMS). (F) X–Y scatterplot showing fold change of genes when comparing +TI to –TI conditions. Fold change of NMSC are on X-axis and of PwMS are on Y-axis. Red line has intercept of 0 and slope of 1. Most discrepant genes are labeled.

FIGURE 2

Gene ontology enrichment analysis of astrocytes following TI stimulation suggests shifts in function. GO enrichment analysis was performed on genes with a Log₂ Fold Change value greater than 1 (up-regulated) or less than –1 (down-regulated) and adjusted p-value less than 0.05. A selection of enriched terms are shown in (A) (up-regulated pathways) and (B) (down regulated pathways), where the X-axis indicates the ratio of genes associated with that term to the total number of genes up or down regulated. BMP is bone morphogenic protein. Heatmaps showing relative expression values of genes associated with selected enriched GO terms in the up-regulated genes (C,D) and down-regulated genes (E,F). Unsupervised clustering of genes is shown along left side and of samples along top.

FIGURE 3

Conditioned media from TI stimulated astrocytes suppresses human OPC differentiation. (A) Schematic depicting experiment. (B) Barplot showing Nanoluciferase activity (normalized to media only control) on Y-axis. Each dot represents a well of hiPSC derived astrocytes whose media was placed onto five identical replicate wells of hOPC reporter cells. Columns represent mean of replicate astrocyte samples, dots represent mean of replicate OPC wells, and error bars represent standard error of the mean of replicate OPC wells. (C) Barplot showing nuclear counts following treatment with astrocyte CM, data presented in same format as in (C). (D) Principle component analysis of hOPC reporter RNA libraries comparing samples treated with media only, media only with TI, and CM from NMSC and PwMS derived astrocytes treated with or without TI. (E) Heatmap showing relative expression of genes previously associated with inflammatory OPCs in hOPC reporter cells treated with astrocyte CM (and media only controls). Dotplots showing selected GO terms found to be enriched in up-regulated (F) and down-regulated (G) genes from hOPC reporter cultures treated with astrocyte CM stimulated with or without TI. Genes were identified as differentially expressed if they had adjusted p-value less than 0.05 and Log₂ Fold Changes values greater than 0.5 or less than –0.5, respectively. Statistical analyses reported in (B,C) are result of linear mixed effect models.