Functionalization of Brain Region-specific Spheroids with Isogenic Microglia-like Cells

Current brain spheroids or organoids derived from human induced pluripotent stem cells (hiPSCs) still lack a microglia component, the resident immune cells in the brain. The objective of this study is to engineer brain region-specific organoids from hiPSCs incorporated with isogenic microglia-like cells in order to enhance immune function. In this study, microglia-like cells were derived from hiPSCs using a simplified protocol with stage-wise growth factor induction, which expressed several phenotypic markers, including CD11b, IBA-1, CX3CR1, and P2RY12, and phagocytosed micron-size super-paramagnetic iron oxides. The derived cells were able to upregulate pro-inflammatory gene (TNF-α) and secrete anti-inflammatory cytokines (i.e., VEGF, TGF-β1, and PGE2) when stimulated with amyloid β42 oligomers, lipopolysaccharides, or dexamethasone. The derived isogenic dorsal cortical (higher expression of TBR1 and PAX6) and ventral (higher expression of NKX2.1 and PROX1) spheroids/organoids displayed action potentials and synaptic activities. Co-culturing the microglia-like cells (MG) with the dorsal (D) or ventral (V) organoids showed differential migration ability, intracellular Ca2+ signaling, and the response to pro-inflammatory stimuli (V-MG group had higher TNF-α and TREM2 expression). Transcriptome analysis exhibited 37 microglia-related genes that were differentially expressed in MG and D-MG groups. In addition, the hybrid D-MG spheroids exhibited higher levels of immunoreceptor genes in activating members, but the MG group contained higher levels for most of genes in inhibitory members (except SIGLEC5 and CD200). This study should advance our understanding of the microglia function in brain-like tissue and establish a transformative approach to modulate cellular microenvironment toward the goal of treating various neurological disorders.

The electrophysiological properties of the outgrowth cells of the derived organoids were examined via patch clamping. Both dorsal and ventral organoids displayed fast inward currents and long-lasting outward currents during voltage-clamp recording, suggesting the presence of functional voltage-gated Na + and K + channels, respectively (Fig. 5A,E). In addition, both spheroid groups fired rebound action potentials in response to hyperpolarizing current injection during current clamp recording (Fig. 5B,F). No action potential firing was observed during depolarizing current injection. Spontaneous postsynaptic currents were observed in the absence of stimulation during continuous voltage clamp recording (Fig. 5C,G). Cellular morphology of both spheroid types was stereotypically neuron-like, with small cell bodies and extensive projections (Fig. 5D,H). Together, these results suggest that the derived dorsal and ventral spheroids have the functional synaptic activities and the ability to fire action potentials. www.nature.com/scientificreports www.nature.com/scientificreports/

Co-culture of dorsal (D) and ventral (V) spheroids/organoids with microglia-like cells.
CellTracker Green or microdevice-labeled microglia-like cells migrated into the cortical spheroids after co-culture. MG incorporation increased over 2 days of culture ( Fig. 6A and Supplementary Fig. S6). The NPC to MG ratio at 4:1 showed a high portion of microglia incorporation into the spheroids. Microglia incorporation was faster in the dorsal spheroids than the ventral spheroids, while the dorsal/ventral (D-V) spheroids showed the intermediate microglia mobility. BrdU incorporation (showing cells in S phase of cell cycle) was determined for the co-culture (Fig. 6Bi). There were 17.6-19.1% of BrdU + cells for microglia only and D-V spheroids (Fig. 6Bii), while 38.3 ± 11.5%, 38.0 ± 11.6%, and 28.6 ± 1.9% of BrdU + cells were observed for dorsal-MG, ventral-MG, and D-V-MG groups respectively. Histological sections of D-MG and V-MG spheroids showed the cellular distribution throughout the spheroids without necrotic center ( Supplementary Fig. S7A). The microglia incorporation was confirmed by P2RY12 (red)/β-tubulin III (green) expression in the histological sections ( Supplementary  Fig. S7B).
Microglia-like cells respond to the ATP/ADP release through P2RY12/13 (purinergic receptors) at the damage area 33 , which would result in the intracellular Ca 2+ transients 35 . ADP-evoked intracellular Ca 2+ transients were observed in the derived MGs ( Fig. 6Ci and Supplementary Video 1), which is the distinctly different characteristics compared to macrophages (i.e., negative ADP-Ca 2+ transients for macrophages) 58 . The dorsal-MG (D-MG), ventral-MG (V-MG), and D-V-MG groups showed instant intracellular Ca 2+ transients after ADP challenge ( Fig. 6Ci and Supplementary Fig. S8). Cells from V-MG spheroids showed more ADP-responsive signals than the other groups (Fig. 6Cii).
The immune response of co-cultured organoids to pro-inflammatory stimuli Aβ42 oligomers was examined (Fig. 7). Dorsal spheroids and microglia control showed comparable levels of TNF-α gene expression (Fig. 7A). D-MG group had similar TNF-α gene expression to dorsal group, while Aβ42 oligomer stimulation upregulated TNF-α expression for the D-MG and V-MG groups (Fig. 7A). No significant difference was observed for IL-6 expression. Co-culturing microglia with dorsal or ventral spheroids also showed differential secretion of VEGF-A and PGE2 (Fig. 7B). D-V-MG spheroids had a higher level of VEGF-A and D-MG spheroids had a higher level of PGE2 after Aβ42 oligomer stimulation, while V-MG spheroids showed insignificant PEG2 secretion and less VEGF-A expression after Aβ42 oligomer stimulation. The reactive oxygen species (ROS) production increased for MG and V-MG groups after Aβ42 oligomer treatment ( Supplementary Fig. S9). But for D-MG and D-V-MG groups, the ROS increase was minimal. To further confirm the immune response, a NF-kB inhibitor BAY11-7082 was used to treat Aβ42 oligomer-stimulated MGs (Fig. 7Ci) and the co-cultured organoids ( Figure Cii). BAY11-7082 treatment reduced TNF-α expression for the MG and V-MG groups. Similarly, TREM2 expression was increased when the cells were treated with Aβ42 oligomers (Fig. 7D). BAY11-7082 treatment reduced TREM2 expression especially for the V-MG group.

Transcriptome analysis of MG and D-MG co-cultures.
The RNA sequencing analysis (RNA-Seq) on the MG and the D-MG groups (4:1 ratio at day 33) was performed to further characterize MG alone and MG in cortical spheroids. For each group, three independent biological replicates were analyzed. An average of 15,731 genes of the 26,364 genes in the reference genome were detected. A pairwise comparison of the two groups using DESeq2 found 11,301 differentially expressed genes ( Fig. 8A and Supplementary Spreadsheet 1). This is expected since the cultures represent different cell types and a majority of the cells in the D-MG group were neuronal. As evidence that the co-culture does contain microglia, 37 of microglia-specific genes, e.g., MERTK, GPR34, PROS1, GAS6, ITGAM (CD11b), CX3CR1, and TMEM119, were found in both groups (Table 1). ITGAM, GPR34, CX3CR1 etc. were enriched in D-MG group, while MERTK, TLR3, TMEM119, CD200R1, CD74 (HLA-DR), PROS1 etc. were highly expressed the MG group. Other immune functional genes such as CD163, CD14, RUNX1, AIF1 were similarly expressed in the two groups. The interferon pathway related genes were enriched in the MG group ( Table 2). The SIGLEC family genes were differentially expressed in the two groups. Microglia heterogeneity was known by the expression of immunoreceptors containing activating and inhibitory members 40 . Our data show that the D-MG group exhibited higher levels of immunoreceptor genes in activating members, but the MG group contained higher levels for most of genes in inhibitory members (except SIGLEC5 and CD200) ( Table 2).
To gain insight into the function of the differentially expressed genes, gene ontology (GO) and KEGG pathway enrichment were examined for the top 500 genes that were upregulated in each group for categories of Biological Process, Cellular Component, and Molecular Function (Supplementary Spreadsheet 2). The examples of "GO" analysis regarding neuron development, morphogenesis, and neural protection were shown in Fig. 8Bi. The expression of chemokines such as CXCL1, 2, 12, 16, and CCL2, 5 was also observed in both groups, but CCL7 and CXCL10 were much higher in the D-MG group (Fig. 8Bii). From the heatmap of top 100 upregulated or downregulated genes (Fig. 8Ci), the MG cells were enriched with collagens and laminin related genes (e.g., COL8A2, COL21A1, LAMB1, ITGA6, and RELN), while the cells of D-MG group were enriched with Glypican 2 (GPC2), CXCR4, Dll1 (a Notch ligand for cell-cell communication), and STC1 (a glycoprotein hormone involved in calcium/phosphate homeostasis). MG group also enriched FRZB, a Wnt-binding protein, and MYRF, the myelin regulatory factor. The D-MG group enriched the genes related to synaptic function, SYT1 and SYT11, and neuroprotective gene ENO2.
A list of 100 genes that share similar expression levels in both cultures was examined (Fig. 8Cii). From "GO" analysis, these genes are implicated in general cellular functions, such as basal transcriptional machinery binding, RNA polymerase II core binding, and several transmembrane transporter activities (Supplementary Table S3) The expression of genes associated with Alzheimer's disease such as APOE, PSEN1, PSEN2, PICALM, and APP were expressed in both groups (Supplementary Table S4). But MAPT, MEF2C, CASS4 were higher in the D-MG group while APOE and PTK2B were higher in the MG group. For genes-related to cortical neurons, they were all higher in the D-MG group as expected (Supplementary Table S5). In addition, oligodendrocyte-associated genes (e.g., Olig1, Olig2, ASCL1) were highly expressed in the D-MG groups (Supplementary Table S6), as well as astrocyte-associated genes (e.g. SLC1A2, S100B) (Supplementary Table S7) and genes related to brain-specific pericytes (e.g., ZIC1 and MCAM, the encoding gene for CD146) (Supplementary Table S8) 59 .

Discussions
Importance for studying hiPSC-microglia. Most of current microglia studies, in particular under disease condition, have to use mouse cells due to the limited access to human microglia 43,60 . iPSCs provide a promising platform to generate human microglia with the patients' genetic backgrounds. There are three microglia groups in human brain: homeostatic, intermediate, and disease-associated microglia (DAM) 43,61 . A shift in microglial phenotype, increased cytokine production, and reduced phagocytic capacity were observed during neural degeneration 62 . This DAM was found to be activated with a two-step mechanism from homeostatic microglia: (1) triggering receptor expressed on myeloid cells 2 (TREM2)-independent mechanism to intermediate microglia; www.nature.com/scientificreports www.nature.com/scientificreports/ the reduction in phenotypic markers CX3CR1 and P2RY12 and the upregulation of APOE were observed 43 . (2) TREM2-dependent activation, showing the upregulation of phagocytic and lipid metabolism genes 34,43 . TREM2 deficiency promotes microglia cell death and inhibits Wnt/β-catenin pathway 63 . During neural degeneration, microglia should clear the Aβ fibrils and secrete neuro-inflammatory cytokines 34,42 , but prolonged activation of microglia is detrimental to brain function. All these studies underscore the importance of investigating the phenotype and function of microglia in healthy or diseased brain environment.
Generation and characterization of hiPSC-microglia. Generation of hPSC-derived microglia-like cells can be differentiated through a CD235a + intermediate state or through a monocyte/macrophage intermediate 30 . Differentiation microglia-like cells from hPSCs was firstly reported in 2016 33 . Embryoid body formation in suspension using CSF1 and IL-34 was performed to obtain progenitors expressing VE-cadherin, c-kit, www.nature.com/scientificreports www.nature.com/scientificreports/ CD41 and CD235a, the markers of early yolk sac myelogenesis. Further differentiation generated CD11b + IBA-1 + semi-adherent cells with vacuolated and round morphology (about 8 weeks). In 2017, a two-step protocol was developed to generate microglia cells from hiPSCs in 5 weeks 34 . Primitive hematopoietic progenitor cells (CD41 + CD43 + CD235a + ) were obtained from iPSCs using Activin A, BMP-4, FGF2 and LiCl, followed by FGF2, VEGF, TPO, SCF, IL-3 and IL-6. Similarly, a monolayer-based differentiation used Activin A, BMP-4, SCF and VEGF for the generation of CD34 + CD43 + cells 37 , or with BMP-4 followed by FGF-2, SCF and VEGF 35 . Compared to previous protocols 37 , our study used the monolayer-based simplified protocol (i.e., remove hypoxia and feeder culture) through mesoderm intermediate to generate microglia-like cells with proper phenotype: 70 ± 6% CD11b and 80 ± 5% IBA-1 at Day 28; 68 ± 12% P2RY12 and 51 ± 10% CX3CR1 at Day 33. The role of VEGF during the development was elucidated: in the absence of VEGF, the KDR at day 5-7 and CX3CR1 at day 30-33 were lower. Functional analysis showed the phagocytosis ability, the increased TNF-α expression and cytokine secretion to pro-inflammatory Aβ42 oligomers, which were reported to modulate microglia responses by TREM2 binding 64 . In addition, ADP-evoked intracellular Ca 2+ transients were observed in iPSC-MGs, a characteristic function distinctly different from macrophages. www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ Co-culture of microglia with brain-region dependent spheroids. To investigate how central nervous system microenvironments affect microglia maturation and homeostasis, microglial cells were co-cultured with rat hippocampal neurons 34 or with isogenic cortical neurons to acquire microglia characteristics in 2D cultures 36 . www.nature.com/scientificreports www.nature.com/scientificreports/ The co-culture upregulated genes in neural protection and suppressed pro-inflammatory signaling, maintaining a homeostatic state 65 . Microglia cells were co-cultured with 3D whole brain organoids in only one study showing that microglia can mature, ramify, and respond to injury; another study derived microglia within cerebral organoids 34,51 . The novelty of our study is to investigate the neural-microglia interactions for brain "region-specific" organoids (dorsal or ventral). The dorsal or ventral identity was confirmed by the expression of TBR1/PAX6 or NKX2.1/PROX1 respectively, and displayed synaptic activities and action potentials.  Our co-culture results indicate that the isogenic microglia-like cells showed differential migration ability and immune response with dorsal or ventral cortical organoids. The immune responses of co-cultured organoids to pro-inflammatory stimuli showed higher TNF-α expression for ventral-MG co-culture. The co-culture also stimulated cell proliferation (higher BrdU + cells), reduced ROS expression, and exhibited the response to ADP treatment for intracellular Ca 2+ transients. During neural degeneration, knockdown of the low-density lipoprotein receptor-related protein (LRP1), a receptor of Aβ peptides, leads to activation of c-Jun N-terminal kinase and NF-kB pathways and enhanced the sensitivity to LPS response 60 . Similarly, in our study, NF-kB pathway is involved in Aβ42-induced pro-inflammatory response (through TREM2 binding). The BAY11-7082 treatment blocked the NF-κB signaling and reduced TNF-α expression in the activated microglia-like cells, especially when co-culturing with ventral organoids.
Our genomics analysis showed the differential expression of microglia-related genes ( Table 1) and genes related to neural degeneration (similar to Haenseler et al., 2017 with 2D co-culture 36 ). Distinct differences between microglia-like cells and co-cultured cells with 3D forebrain spheroids were observed, in particular, for immunoreceptors containing activating and inhibitory members ( Table 2) 40 , which indicates that the immunophenotype has heterogeneity for the MG and the D-MG groups. The genes related to ECMs (Fig. 8 and Supplementary Spreadsheet 3) were also differentially expressed. The microglia-like cells were enriched with collagens and laminin related genes, while the co-cultured cells expressed higher levels of Glypican 2 and Notch ligand. The co-cultured cells had higher levels of genes related to neurogenesis, synaptic function, as well as neuroprotection. ECMs and cytokines are important to activate microglia function 66 , thus the neural microenvironment provided by cortical spheroids should promote microglia function. To date, the transcriptome analysis of hiPSC-derived microglia showed the similarity to human fetal microglia 35,37,39 . However, the brain region-dependent diversity of microglia phenotype and the response to neural degeneration 40,43 may require microenvironment provided by brain region-specific spheroids. Therefore, our transcriptome analysis focus on the comparison of the presence and absence of 3D cortical microenvironment.   www.nature.com/scientificreports www.nature.com/scientificreports/ 3D spheroids experiences hypoxia and glucose starvation compared to 2D cultures 67 . Metabolic reprogramming may occur from oxidative phosphorylation to aerobic glycolysis in 3D spheroids 68 , which favors microglia M1 polarization 69 . Activation of mTOR (PIK3/AKT/mTOR pathway) and p53, repression of canonical Wnt and NF-kB pathway activities, and higher ATP levels could be observed in 3D spheroid culture 67 , as well as our genomics analysis summarized in a separate manuscript 68 . So the co-cultured 3D organoids, as shown in this study, more resemble physiological brain tissue environment than 2D neural cultures. Implication for disease modeling. Patient-derived 3D brain organoids from hiPSCs have been widely investigated recently as a powerful platform for drug screening for neurodegenerative disease. As the only resident macrophages and indispensable cellular component of blood-brain barrier (BBB), most recent brain organoids currently lack the microglia. A couple of recent studies demonstrated the importance of adding microglia into the neural constructs by observing up-regulated genes related to immune functions, such as IBA-1, CD14, and TREM2, compared to those without microglia 17,70 . However, the microglia-functionalized brain organoids in a diseased environment have not been well investigated. Our lab recently reported the derivation of AD-patient derived cortical organoids to recapitulate neurodegenerative microenvironment and investigate its response to potential drug treatment 46 . AD-related microenvironment including a higher level of Aβ42, elevated pro-inflammatory gene expression, and altered matrix remodeling proteins were recapitulated in the cortical organoids derived from AD-associated hiPSCs. This study performed the functional characterization of microglia within different regions of forebrain organoids. The long-term goal is to fabricate next-generation brain organoids with additional cellular components (e.g., microglia, BBB properties, astrocytes, and oligodendrocytes) from hiPSCs for disease modeling, drug screening, and possibly cell therapy 71 .

conclusions
In summary, the microglia-like cells were derived from hiPSCs and integrated with brain region-specific organoids. Co-culturing isogenic microglia-like cells with hiPSC-derived dorsal and ventral organoids showed differential migration ability, Ca 2+ transients imaging, and the response to pro-inflammatory stimuli. D-MG group showed higher anti-inflammatory cytokine secretion, while V-MG group showed higher TNF-α expression under Aβ42 stimulation. The co-culture (especially V-MG group) also stimulated cell proliferation (higher BrdU + cells) and reduced ROS expression, better resembling tissue-specific microenvironment. Transcriptome analysis exhibited microglia-related genes that were differentially expressed in MG and D-MG groups. This study should advance our understanding of the effects of microglia on brain tissue function towards engineering complex next generation of brain organoids.

Dorsal and ventral spheroid/organoid differentiation from hiPSCs. Undifferentiated iPSK3 cells
(2 × 10 5 cells) were seeded into low attachment 24-well plates in neural differentiation medium composed of DMEM/F-12 plus 2% B27. Y27632 (10 µM) was added during the seeding and removed after 24 h. For dorsal differentiation, the aggregates were treated with 10 µM SB431542 (Sigma) and 100 nM LDN193189 (Sigma) for 7 days. Then the spheroids were treated with FGF2 at 25 ng/mL for another 7 days 46,47,49 . To generate a ventral identity, the aggregates were treated with 10 µM SB431542 (Sigma), 100 nM LDN193189 (Sigma), and 5 µM IWP4 (Sigma) for 7 days. Then the spheroids were incubated with 5 µM IWP4 and 1 µM purmorphamine (Sigma) for another 7 days. For maturation, the spheroids were maintained in neural differentiation medium without growth factors for additional 14-38 days (total up to 52 days). The dorsal and ventral identity was characterized using histology, flow cytometry, and RT-PCR.
Preparation of gelatin-coated glass slides. A 60 °C gelatin solution (10 wt% in water) was dip-coated onto a glass slide. The solution was dried in air at room temperature for 12 hours.
Preparation of micro contact printing (µCP) stamps. Monomer and curing agent of Sylgard 184 PDMS kit (Dow Corning) were mixed at 10:1 weight ratio. The mixture was degassed and poured on a silicon master with micro-patterns prepared by photolithography. The mixture was then cured at 37 °C for 24 hours to form a PDMS slab. The PDMS slab was peeled off from the master and cut into small stamps. The stamp had an array of circular pillars with a diameter of 5 µm and a center-to-center distance of 10 µm in the square lattice.
Fabricating PAH/PSS/PAH/PPMA-R18 particles. Three layers of polyelectrolyte (PAH/PSS/PAH) were deposited onto the PDMS stamp via layer-by-layer technique. The stamp was soaked in polyelectrolyte solution for 15 min to deposit the specific layer. Polyelectrolyte solution used for the first layer: PAH solution (1 wt% in water, pH = 0, containing 150 mM NaCl); second layer: PSS solution (1 wt% in water); third layer: PAH solution (1 wt% in water, pH = 4, without NaCl). The stamp was washed with water and dried under a stream of nitrogen after soaking each layer. Then, an acetone solution of PPMA (5 wt%, containing 50 µg/mL R18) was spin-coated onto the PDMS stamp at 3000 rpm for 45 s. The stamp was brought into contact with a gelatin-coated glass slide which was placed on a hot plate at 100 °C. After 5 seconds, the stamp was peeled off, the particles were transferred onto the gelatin film.
Labeling microglial-like cell with particles. A PDMS chamber was placed onto a gelatin-coated glass slide which has the printed micro-arrays, forming a well whose bottom was covered by micro particles (each well has 7.85 × 10 5 particles on the bottom). The well was placed in 6 °C refrigerator for 30 min. Then 1 × 10 5 microglial cells (in 460 µL of phosphate-buffered saline, at 6 °C) was added into the well and kept at 6 °C for 30 min. The microglial cells precipitated and attached onto the particles. After 30 min, the well was brought into incubator (37 °C, 5% CO 2 ) and incubated for 2 hours. The gelatin film was dissolved so the particles released, forming microglial/particle complexes during the process. www.nature.com/scientificreports www.nature.com/scientificreports/ Immunocytochemistry. Briefly, the samples were fixed with 4% paraformaldehyde (PFA) and permeabilized with 0.2-0.5% Triton X-100. The samples were then blocked for 30 min and incubated with various mouse or rabbit primary antibodies (Supplementary Table S3) for four hours. For surface markers, no permeabilization was performed. After washing, the cells were incubated with the corresponding secondary antibody: Alexa Fluor ® 488 goat anti-Mouse IgG 1 , 488 or 594 goat anti-Rabbit IgG, or 594 donkey anti-Goat IgG (Life Technologies) for one hour. The samples were counterstained with Hoechst 33342 and visualized using a fluorescent microscope (Olympus IX70, Melville, NY). For 5-Bromo-2′-deoxyuridine (BrdU) assay, the cells were incubated in medium containing 10 µM BrdU (Sigma) for four hours. The cells were then fixed with 70% cold ethanol and denatured using 2N HCl/0.5% Triton X-100 for 30 min in the dark. The samples were reduced with 1 mg/mL sodium borohydride for 5 min and incubated with mouse anti-BrdU (1:100, Life Technologies) in blocking buffer (0.5% Tween 20/1% bovine serum albumin in PBS), followed by Alexa Fluor ® 488 goat anti-Mouse IgG 1 . The cells were counterstained with Hoechst 33342 and analyzed by a fluorescent microscope and the ImageJ software.

Co-culture of dorsal or ventral spheroids/organoids with microglia-like cells (MGs
Flow cytometry. To quantify the levels of various neural marker expression, the cells were harvested by trypsinization and analyzed by flow cytometry. Briefly, 1 × 10 6 cells per sample were fixed with 4% PFA and washed with staining buffer (2% FBS in PBS). The cells were permeabilized with 100% cold methanol, blocked, and then incubated with primary antibodies against NKX2.1, SATB2, KDR, CD45, CD31, CD11b, IBA-1, CX3CR1, and P2RY12 (Supplementary Table S1) followed by the corresponding secondary antibody Alexa Fluor ® 488 goat anti-Mouse IgG 1 , or 594 goat anti-rabbit IgG, or 594 donkey anti-goat IgG. For surface markers, no permeabilization was performed. The cells were acquired with BD FACSCanto ™ II flow cytometer (Becton Dickinson) and analyzed against isotype controls using FlowJo software. Histology. For histology, the dorsal, ventral, or the D-V spheroids/organoids containing microglia-like cells were fixed in 10% formalin, dehydrated, and embedded in paraffin wax. The sections of 10 μm were cut, de-paraffinized, and stained with Lerner-2 Hematoxylin (Lerner Laboratories, Pittsburgh, PA) and Eosin-Y w/Phloxine (Richard-Allan Scientific, Kalamazoo, MI) 79 . The sections were also stained with various neural markers, including anti-SATB2, TBR1, BRN2, NKX2.1, PAX6, Glutamate, vGAT, β-tubulin III (Supplementary  Table S1), to show different neural cell distribution and localization. Images were captured with an Olympus IX70 microscope with MagnaFire SP 2.1B software.

Whole
Reverse transcription polymerase chain reaction (RT-PCR) analysis. Total RNA was isolated from different cell samples using the RNeasy Mini Kit (Qiagen, Valencia, CA) according to the manufacturer's protocol followed by the treatment of DNA-Free RNA Kit (Zymo, Irvine, CA) 80 . Reverse transcription was carried out using 2 μg of total RNA, anchored oligo-dT primers (Operon, Huntsville, AL), and Superscript III (Invitrogen, Carlsbad, CA) (according to the protocol of the manufacturer). Primers specific for target genes (Supplementary  Table S2) were designed using the software Oligo Explorer 1.2 (Genelink, Hawthorne, NY). The gene β-actin was used as an endogenous control for normalization of expression levels. Real-time RT-PCR reactions were performed on an ABI7500 instrument (Applied Biosystems, Foster City, CA), using SYBR1 Green PCR Master Mix (Applied Biosystems). The amplification reactions were performed as follows: 2 min at 50 °C, 10 min at 95 °C, and 40 cycles of 95 °C for 15 sec and 55 °C for 30 sec, and 68 °C for 30 sec. Fold variation in gene expression was quantified by means of the comparative Ct method , which is based on the comparison of expression of the target gene (normalized to the endogenous control β-actin) between the compared samples.
Reactive oxygen species (ROS) assay. ROS detection was performed using Image-iT ™ Live Green Reactive Oxygen Species Detection kit (Molecular probes). Briefly, the spheroids and single cells were washed in Hank's Balanced Salt Solution (HBSS), and incubated in a solution of 25 µM carbioxy-H 2 DCFDA for 30 min at 37 °C. The samples (+/−Aβ42 oligomers stimulation) were then washed and analyzed under fluorescence microscope or by flow cytometry. As positive control, the cells were incubated in a 100 µM tert-butyl hydroperoxide solution, prior to staining with carbioxy-H 2 DCFDA. intracellular ca 2+ signaling assay. For calcium signaling, the samples were related on 1% Geltrex-coated 96-well plate and grown overnight. The growth medium was removed in each well and 100 μL of 1X Fluo-4 dye (Life Technologies) in assay buffer containing 1X HBSS and 20 mM HEPES (with 2.5 mM probenecid) was added into the wells and incubated at 37 °C for 30 min. The incubation was switched to room temperature for an additional 30 min. Baseline Ca 2+ signals (I 494 /I 516 ) were measured for more than 100 s, and then the calcium dye medium was replaced with 100 μL of 10 μM adenosine 5′-diphosphate (ADP, Sigma) solution in assay buffer www.nature.com/scientificreports www.nature.com/scientificreports/ (without probenecid). Ca 2+ recordings were read on a fluorescent plate reader (FLX800, Bioinstrument Inc., Winooski, VT) using instrument settings appropriate for excitation at 494 nm and emission at 516 nm.
RNA extraction and RNA-Seq cDNA Library Preparation. RNA was extracted from microglia-like cells only and D-MG group (4:1 ratio) at day 33 using the miRNeasy minikit (Qiagen). mRNA was isolated from the total RNA using an NEBNext Poly(A) mRNA Magnetic Isolation Module (New England Biolabs). cDNA libraries were generated from the isolated mRNA using an NEBNext Ultra RNA library prep kit for Illumina (New England Biolabs) and a unique 6 nucleotide index primer (NEBNext multiplex oligos for Illumina) was incorporated into each sample. The library construction was done according to the NEB manuals, modified for use with a Beckman Biomek 4000 at the Florida State University Biological Sciences core lab. The unique index (barcode) was added to each library to multiplex the six libraries in one lane of the sequencing run. The multiplexed sample was quantified with qPCR (Kapa Biosystems) specific for Illumina sequencing primers and the average fragment size was determined with a Bioanalyzer high sensitivity DNA chip (Agilent Technologies). 12 pM of the pooled sample was sequenced, with single end, 100 base reads on an Illumina HiSeq2500 located in the Translational Science Laboratory at the College of Medicine, Florida State University. The pooled data were demultiplexed into individual sample data and adapter primer sequences were removed 81 .
RNA-Seq data analysis. Initial quality control analysis of each sequenced library was performed using fastQC software (http://www.bioinformatics.babraham.ac.uk/projects/fastqc). The sequencing reads were further analyzed using RNA-Seq Alignment version 1.1.1 (Illumina BaseSpace application). The reads were aligned with Tophat 2 82 to the human genome (genome release GRCh38) using default parameters and counts for each gene were generated. This workflow uses Cufflinks to generate FPKM (fragments per kilobase per million reads) normalized values 83 . These normalized values account for differences in sequencing depth and the length of the gene. FPKM values were used to generate the heatmaps using Morpheus (Broad Institute; https://clue.io/morpheus). DESeq2 was used to determine statistically significant differentially expressed genes (a False Discovery Rate, FDR, of <0.05 was used). 15,585 genes were considered to be expressed in this study by the DESeq2 software due to low counts 84 . The top 500 genes that were upregulated and downregulated (1000 total genes) in the microglia culture versus the D-MG group were further assessed for GO, KEGG pathway and phenotype pathway analysis using Webgestalt 85,86 . The set of genes considered expressed in our dataset was used as the reference set to obtain significantly enriched pathways. Significant enrichment was determined in Webgestalt using the hypergeometric test and the Benjamini-Hochberg FDR method 87 for multiple testing adjustment.

Statistical analysis.
Each experiment was carried out at least three times. The representative experiments were presented and the results were expressed as [mean ± standard deviation]. To assess the statistical significance, one-way ANOVA followed by Fisher's LSD post hoc tests were performed. A p-value < 0.05 was considered statistically significant.