Human cortical spheroids with a high diversity of innately developing brain cell types.

BACKGROUND: Three-dimensional (3D) human brain spheroids are instrumental to study central nervous system (CNS) development and (dys)function. Yet, in current brain spheroid models the limited variety of cell types hampers an integrated exploration of CNS (disease) mechanisms. METHODS: Here we report a 5-month culture protocol that reproducibly generates H9 embryonic stem cell-derived human cortical spheroids (hCSs) with a large cell-type variety. RESULTS: We established the presence of not only neuroectoderm-derived neural progenitor populations, mature excitatory and inhibitory neurons, astrocytes and oligodendrocyte (precursor) cells, but also mesoderm-derived microglia and endothelial cell populations in the hCSs via RNA-sequencing, qPCR, immunocytochemistry and transmission electron microscopy. Transcriptomic analysis revealed resemblance between the 5-months-old hCSs and dorsal frontal rather than inferior regions of human fetal brains of 19-26 weeks of gestational age. Pro-inflammatory stimulation of the generated hCSs induced a neuroinflammatory response, offering a proof-of-principle of the applicability of the spheroids. CONCLUSIONS: Our protocol provides a 3D human brain cell model containing a wide variety of innately developing neuroectoderm- as well as mesoderm-derived cell types, furnishing a versatile platform for comprehensive examination of intercellular CNS communication and neurological disease mechanisms.

Molecular Signature of Neuroinflammation Induced in Cytokine-Stimulated Human Cortical Spheroids.

Crucial in the pathogenesis of neurodegenerative diseases is the process of neuroinflammation that is often linked to the pro-inflammatory cytokines Tumor necrosis factor alpha (TNFα) and Interleukin-1beta (IL-1ß). Human cortical spheroids (hCSs) constitute a valuable tool to study the molecular mechanisms underlying neurological diseases in a complex three-dimensional context. We recently designed a protocol to generate hCSs comprising all major brain cell types. Here we stimulate these hCSs for three time periods with TNFα and with IL-1ß. Transcriptomic analysis reveals that the main process induced in the TNFα- as well as in the IL-1ß-stimulated hCSs is neuroinflammation. Central in the neuroinflammatory response are endothelial cells, microglia and astrocytes, and dysregulated genes encoding cytokines, chemokines and their receptors, and downstream NFκB- and STAT-pathway components. Furthermore, we observe sets of neuroinflammation-related genes that are specifically modulated in the TNFα-stimulated and in the IL-1ß-stimulated hCSs. Together, our results help to molecularly understand human neuroinflammation and thus a key mechanism of neurodegeneration.

Molecular Effects of FDA-Approved Multiple Sclerosis Drugs on Glial Cells and Neurons of the Central Nervous System.

Multiple sclerosis (MS) is characterized by peripheral and central inflammatory features, as well as demyelination and neurodegeneration. The available Food and Drug Administration (FDA)-approved drugs for MS have been designed to suppress the peripheral immune system. In addition, however, the effects of these drugs may be partially attributed to their influence on glial cells and neurons of the central nervous system (CNS). We here describe the molecular effects of the traditional and more recent FDA-approved MS drugs Fingolimod, Dimethyl Fumarate, Glatiramer Acetate, Interferon-ß, Teriflunomide, Laquinimod, Natalizumab, Alemtuzumab and Ocrelizumab on microglia, astrocytes, neurons and oligodendrocytes. Furthermore, we point to a possible common molecular effect of these drugs, namely a key role for NFκB signaling, causing a switch from pro-inflammatory microglia and astrocytes to anti-inflammatory phenotypes of these CNS cell types that recently emerged as central players in MS pathogenesis. This notion argues for the need to further explore the molecular mechanisms underlying MS drug action.

Reappraisal of Human HOG and MO3.13 Cell Lines as a Model to Study Oligodendrocyte Functioning.

Myelination of neuronal axons is essential for proper brain functioning and requires mature myelinating oligodendrocytes (myOLs). The human OL cell lines HOG and MO3.13 have been widely used as in vitro models to study OL (dys) functioning. Here we applied a number of protocols aimed at differentiating HOG and MO3.13 cells into myOLs. However, none of the differentiation protocols led to increased expression of terminal OL differentiation or myelin-sheath formation markers. Surprisingly, the applied protocols did cause changes in the expression of markers for early OLs, neurons, astrocytes and Schwann cells. Furthermore, we noticed that mRNA expression levels in HOG and MO3.13 cells may be affected by the density of the cultured cells. Finally, HOG and MO3.13 co-cultured with human neuronal SH-SY5Y cells did not show myelin formation under several pro-OL-differentiation and pro-myelinating conditions. Together, our results illustrate the difficulty of inducing maturation of HOG and MO3.13 cells into myOLs, implying that these oligodendrocytic cell lines may not represent an appropriate model to study the (dys)functioning of human (my)OLs and OL-linked disease mechanisms.

Evidence for a Priori Existence of Attentional Bias Subgroups in Emotional Processing of Aversive Stimuli.

Little is known regarding inter-individual differences in attentional biases for pain-related information; more knowledge is crucial, since these biases have been associated with differences in pain processing as well as in predicting the risk of postoperative pain. The present study investigated EEG correlates of attentional bias patterns for pain-related information, with specific focus on avoidance- and vigilance-like behavior. Forty-one participants performed a dot-probe task, where neutral and pain-related words were used to create neutral, congruent, incongruent, and double (two pain-related words) trials. EEG was recorded, which was used to generate ERP's of the word-processing phase and the post-dot phase. Participants were placed in two subgroups based on the direction of their attentional bias (either positive; toward the pain-related words, or negative; away from pain-related words). Using t-profiles, four latency windows were identified on which the two subgroups differed significantly. These latency windows yield areas which correspond with the P1-N1 domain and the P3b for the word-processing phase, while the post-dot phase latency windows cover the areas of the P200 and the P3b. The two subgroups show differences on congruent, incongruent, and the double trials, but interestingly also on the neutral trials. Most notably, the area in the word-phase associated with the P3b is diminished in the subgroup showing a negative bias. The deflections associated with both early and late attentional components, including the P3B, as well as a positive deflection in the timeframe of proposed response evaluation processes differ significantly between subgroups. In this study we demonstrated that different attentional biases exist in the healthy population, by showing differences in ERP's. We also show differences in processing neutral trials, which suggests there are fundamental differences between these groups in processing words in general.