Co-Culturing Microglia and Cortical Neurons Differentiated from Human Induced Pluripotent Stem Cells

In the human brain, microglia are the primary innate immune cells¹. Brain development is regulated by microglia via two routes: release of diffusible factors and phagocytosis¹. Microglia-derived diffusible factors help support myelination, neurogenesis, synaptic formation, maturation, cell death, and cell survival¹. Microglia also phagocytize various elements in brain synapses, axons and in both living and dead cells^{2,3,4,5,6,7,8}. Receptors on microglia recognize tags such as calreticulin, ATP, and sialic acid and regulate cellular phagocytosis^9,10. In the hippocampus, microglia maintain the homeostasis of neurogenesis through its phagocytic role¹¹.

Synaptic phagocytosis in the dorsolateral geniculate nucleus (dLGN) of the rodent brain has been shown to be regulated by microglia¹. In rodents, it has been shown that there are two periods during the development when intense microglial synaptic phagocytosis is observed. The first period occurs during initial synapse formation and the second period occurs when connections are being fine-tuned and pruned¹². Other factors that are involved in synaptic pruning are inflammatory proteins and the Class I major histocompatibility complex (MHC1, H2-K^b and D^b)^13,14. It has been suggested that C1q (complement component 1q) on the microglia colocalizes with MHC1, which triggers synaptic pruning¹⁵. Furthermore, mouse studies show that interleukin-33 (IL-33) secreted by astrocytes regulates synapse homeostasis in the thalamus and the spinal cord through its effects on microglia, though this has yet to be investigated in humans¹³. Microglia secrete a variety of cytokines that help maintain neuronal health, such as tumor necrosis factor α (TNFα), IL-1β, IL-6, IL-10 and interferon-γ (IFN-γ) and these cytokines can modulate dendritic spine and synapse formation^16,17,18. There are significant gaps in our knowledge of neuron-microglia interactions during human brain development. Most of our knowledge comes from studies from rodent models, while there is a paucity of information on the temporal and mechanistic aspects of synaptic pruning in the human cortex. Microglia support neuronal survival in the neo-cortex, and other cell types contribute as well¹. It is not clear how microglia contribute to this preservation and what the interplay between microglia and the other cell types are. Microglia release several cytokines that affect neuronal and synaptic development but the mechanistic basis of their effects of these cytokines in neurons are largely unknown^19,20. In order to develop a more complete understanding of the function of microglia in the human brain, it is critical to explore its interactions with different cell types found in the human brain. This report describes a method to co-culture human iPSC-derived neurons and microglia generated from the same individual. Establishing this methodology will enable well-defined investigations to interrogate the nature of microglia-neuronal interactions and to develop robust in vitro cellular models to study neuroimmune dysfunction in the context of different neurodevelopmental and neuropsychiatric disorders.

The role of microglia in schizophrenia
Synaptic pruning is a major neurodevelopmental process that takes place in the adolescent brain^21,22. Multiple lines of evidence suggest that synaptic pruning during this critical period is abnormal in schizophrenia (SCZ)^23,24,25,26. SCZ is a chronic, debilitating psychiatric disorder characterized by hallucinations, delusions, disordered thought processes and cognitive deficits^23,24. Microglia, the resident macrophages in the brain, play a central role in synaptic pruning^25,26. Postmortem and positron emission tomography (PET) studies show evidence for dysfunctional microglial activity in SCZ^{25,26,27,28,29,30,31,32}. Postmortem SCZ brains show well-replicated but subtle differences in the brain – pyramidal neurons in the cortical layer III show decreased dendritic spine density and fewer synapses^33,34,35. Synaptic pruning is a process by which superfluous excitatory synaptic connections are eliminated by microglia during adolescence, when SCZ patients usually have their first psychotic break^22,36. Postmortem studies show an association between SCZ and microglial activation, with increased density of microglia in SCZ brains, as well as increased expression of proinflammatory genes²⁷. In addition, PET studies of human brains using radioligands for microglial activation show increased levels of activated microglia in the cortex^25,26,27,28. Recent genome-wide association studies (GWAS) show that the strongest genetic association for SCZ resides in the major histocompatibility complex (MHC) locus, and this association results from alleles of the complement component 4 (C4) genes that are involved in mediating postnatal synaptic pruning in rodents³⁷. This association has provided additional support for the hypothesis that aberrant pruning by microglia may result in the decreased dendritic spine density seen in SCZ postmortem brains. Investigations of microglial involvement in synaptic pruning in SCZ have so far been limited to indirect studies with PET imaging or inferences from investigations of postmortem brains.

Generating human microglia in the laboratory
Cultured primary mouse microglia have been frequently used in studying microglia, though there are several indications that rodent microglia may not be representative of human microglial anatomy and gene expression (Table 1)³⁸. Several studies have also differentiated microglia directly from blood monocytes through transdifferentiation^39,40,41,42. Blood monocyte-derived microglia-like cells exhibit major differences from human microglia in gene and protein expression profile pro-inflammatory responses, and they appear to be more macrophage-like in their biology⁴³. Recent methodological advances now enable the generation of microglia from human iPSCs, which provide opportunities to study live microglia that more accurately resemble the biology of microglia found in the human brain (Table 2). These iPSC-derived microglial cells have been shown to recapitulate the phenotype, gene expression profiles, and functional properties of primary human microglia^{44,45,46,47,48}. This paper provides a method to co-culture human iPSC-derived neurons and microglia generated from the same individual in order to develop personalized in vitro models of neuron-microglia interactions. For this in vitro co-culture model, a microglial differentiation protocol from the Fossati group was adapted (Table 3) and combined with an adapted version of a cortical neuronal generation protocol from the Livesey group (Table 4)^49,50.