APOE ε4/ε4 diminishes neurotrophic function of human iPSC-derived astrocytes

Abstract

The ε4 allele of the APOE gene encoding apolipoprotein E (apoE) is a strong genetic risk factor for aging-related cognitive decline as well as late-onset Alzheimer’s disease (AD) compared to the common ε3 allele. In the central nervous system, apoE is produced primarily by astrocytes and functions in transporting lipids including cholesterol to support neuronal homeostasis and synaptic integrity. Although mouse models and corresponding primary cells have provided valuable tools for studying apoE isoform-dependent functions, recent studies have shown that human astrocytes have a distinct gene expression profile compare with rodent astrocytes. Human induced pluripotent stem cells (iPSCs) derived from individuals carrying specific gene variants or mutations provide an alternative cellular model more relevant to humans upon differentiation into specific cell types. Thus, we reprogramed human skin fibroblasts from cognitively normal individuals carrying APOE ε3/ε3 or ε4/ε4 genotype to iPSC clones and further differentiated them into neural progenitor cells and then astrocytes. We found that human iPSC-derived astrocytes secreted abundant apoE with apoE4 lipoprotein particles less lipidated compared to apoE3 particles. More importantly, human iPSC-derived astrocytes were capable of promoting neuronal survival and synaptogenesis when co-cultured with iPSC-derived neurons with APOE ε4/ε4 astrocytes less effective in supporting these neurotrophic functions than those with APOE ε3/ε3 genotype. Taken together, our findings demonstrate APOE genotype-dependent effects using human iPSC-derived astrocytes and provide novel evidence that the human iPSC-based model system is a strong tool to explore how apoE isoforms contribute to neurodegenerative diseases.

Introduction

Astrocytes are the most abundant cell type in the central nervous system (CNS) responsible for the maintenance of brain homeostasis (1); disturbances of their functions contribute to the pathogenesis of neurodegenerative diseases (2,3). In particular, astrocytes play a critical role in neurotransmitter trafficking and recycling, nutrient and ion metabolism, release of transmitters and growth factors, and protection against oxidative stress (4,5). They also exert profound impacts on neuronal networks during development and upon injury by supporting neuronal survival, axon and dendrite outgrowth, and synaptogenesis (6,7). Astrocyte-derived cholesterol has been shown to be a critical factor in promoting synaptogenesis in neurons (8). In the brain, apolipoprotein E (apoE) is a major cholesterol carrier, which is mainly synthesized and secreted by astrocytes (9,10). ApoE distributes lipids among cells in the CNS, thereby participating in neuronal repair and remodeling (10,11). In humans, the APOE gene has three major allelic variants: ε2, ε3, and ε4. Among them, carrying APOE ε4 allele is associated with significantly increased risk for age-related cognitive decline as well as Alzheimer’s disease (AD), whereas APOE ε2 is protective (9). AD is the most common form of dementia in the elderly accounting for 60-80% of all cases (12), where extracellular aggregates of amyloid-β (Aβ) and intracellular neurofibrillary tangles (NFTs) are major pathological hallmarks (13,14). Although several pathways have been proposed to explain the APOE genotype effects on cognition through animal and human studies, the underlying molecular mechanisms are not fully understood, in particular using relevant human cell models (15).

Toward the establishment of human cellular models for studying neurodegenerative diseases, recent advances in the induced pluripotent stem cell (iPSC) technology have provided a new opportunity to examine potential hypotheses (16,17). The iPSCs and iPSC-derived brain cells from individuals carrying specific gene variants or mutations enable the expression of endogenous genes at physiological levels, allowing disease modeling of relevant phenotypes in vitro. In this study, we obtained human skin fibroblasts from cognitively normal individuals with different APOE genotypes (ε3/ε3 or ε4/ε4) and reprogramed them to iPSC clones by episodic expression of defined transcription factors (18). By differentiating the iPSC clones into astrocytes, we found that apoE production was increased in a manner depending on astrocytic maturation. Importantly, human iPSC-derived astrocytes displayed APOE genotype-dependent effects on neuronal viability and synaptogenesis. To the best of our knowledge, we provide the first evidence that human iPSC-derived astrocytes can be used as a relevant in vitro human model in understanding the role of apoE in neurodegenerative diseases including AD.

Results

Generation and characterization of human iPSCs with different APOE genotypes

Fibroblasts from healthy individuals carrying APOE ε3/ε3 (n = 3) or APOE ε4/ε4 (n = 3) genotype were obtained from Mayo Clinic Neuroregeneration Lab (Fig. 1A). The fibroblasts were reprogrammed into iPSC clones, where two to three clones from each fibroblast line were selected for further studies.

Figure 1.

Generation and characterization of human iPSCs with APOE ε3/ε3 or APOE ε4/ε4 genotype. (A) List of fibroblasts from normal individuals with APOE ε3/ε3 or APOE ε4/ε4 genotype with information for age and sex. (B) Immunostaining of iPSC line APOE ε3/ε3 (MC0192) or APOE ε4/ε4 (MC0018) for pluripotency markers (Nanog, TRA-1-60 and SSEA4). Scale bars, 100 μm. (C) In vitro differentiation of iPSC lines into cell types of three germ layers. Cells were immunostained for Sox17 (endoderm), brachyury (mesoderm), or Nestin/Sox2 (ectoderm), together with DAPI (nucleus). Scale bars, 50 μm. (D) The expression of endogenous OCT3/4 and SOX2 at the mRNA level was assessed in all iPSC lines by qRT-PCR. A human ES line and an iPSC line obtained from ATCC were used as positive controls. Data are mean ± SEM from three independent experiments. (E) Karyotyping for the iPSCs are shown.

We confirmed that all iPSCs expressed the pluripotent stem cell-specific markers, including Nanog, TRA-1-60 and SSEA4 (Fig. 1B; Supplementary Material, Fig. S1A). In addition, they were capable of differentiating in vitro into endodermal, mesodermal and ectodermal cell types confirmed by immunostaining for Sox17 (endoderm marker), Brachyury (mesoderm marker) and Nestin/Sox2 (ectoderm marker), respectively, after 5-7 days differentiation in specific medium (Fig. 1C; Supplementary Material, Fig. S1B), indicating their pluripotency. Among iPSC lines with APOE ε3/ε3 or APOE ε4/ε4, no obvious differences were observed in their morphology and their ability to differentiated into three germ layers. In addition, expression of the endogenous pluripotent markers Oct3/4 and Sox2 at the mRNA level was also increased in all iPSC lines compared to fibroblasts when analyzed by RT-PCR, comparable to those from human embryonic stem (ES) cells (H5) and an iPSC line from ATCC (Fig. 1D). Karyotyping in each iPSC clone revealed the preservation of normal number and appearance of chromosomes after passaging 10 times (Fig. 1E; Supplementary Material, Fig. S1C).

Differentiation of human iPSCs into neurons and astrocytes

To derive neural progenitor cells from the iPSCs, iPSC clumps were first cultured in suspension in the neural induction medium to generate neurospheres for 5-7 days (Supplementary Material, Fig. S2A). Next, neurospheres were plated on Matrigel-coated plates onto generate neural rosettes expressing neural progenitor markers Nestin and PAX6 (Supplementary Material, Fig. S2A). Finally, neural rosettes were dissociated into single cells and differentiated into NPCs (Supplementary Material, Fig. S2A). We confirmed that more than 97% of iPSC-derived NPCs were Nestin-positive by fluorescence-activated cell sorter (FACS) (Supplementary Material, Fig. S2B).

While NPCs with low passage numbers spontaneously differentiate into neurons upon growth factor withdrawal (30,31), the gliogenic switch for differentiation is induced with a high passage number regardless of the presence of FGF and/or EGF (32,33). Therefore, we used the NPCs with passage number 2–4 for neuronal differentiation. After 2 weeks, differentiated cells exhibited a typical neuronal morphology and expressed neuronal marker MAP2 (Supplementary Material, Fig. S2A).

Astrocytic differentiation was induced in the NPCs with passage number 6. When the cells were stained with astrocytic markers (S100β and GFAP) at different time points during differentiation, S100β emerged as early as day 15 of differentiation from NPCs (Fig. 2A). Cell morphology changed along with differentiation, which was accompanied by a progressive increase in the number of GFAP-positive cells. After 45 days of differentiation, nearly all cells were positive for both S100β and GFAP (Fig. 2A). In addition, we confirmed that the differentiated astrocytes expressed AQP4 (a marker for the main water channel in the perivascular membranes) and Vimentin (an intermediate filament marker of reactive astrocytes) at day 45 (Fig. 2B). A major function of mature astrocytes is to eliminate excess glutamate from the extracellular space, which is required for the survival and normal function of neurons (34). Thus, to further validate the iPSC-derived astrocytes, we performed glutamate uptake assay in those cells. We found that the iPSC-derived astrocytes were able to take up glutamate more efficiently compared to HEK 293FT cells which lack specific glutamate uptake ability (35) (Fig. 2C). Together, these results indicate that the iPSCs were successfully differentiated into astrocytes with sufficient maturity and functionality. No significant differences were observed between iPSC clones with APOE ε3/ε3 or APOE ε4/ε4 genotype regarding the differentiation efficiency to neurons or astrocytes and their morphologies under these culture conditions.

Figure 2.

Differentiation and characterization of human iPSC-derived astrocytes. (A) The iPSC-derived cells (MC0192 #1) were stained for astrocytic markers (GFAP and s100β) at different time points after the start of differentiation from NPCs (day 0, 15, 30 and 45). Scale bars, 50 μm. (B) The cells (MC0192 #1) were stained with additional astrocytic markers AQP4 and Vimentin at day 45. Scale bars, 50 μm. (C) Glutamate uptake of the cells (MC0192 #1) was examined at day 45 of differentiation. (D) ApoE levels in the culture media from selected iPSC-derived astrocytes with APOE ε3/ε3 (MC0117) or APOE ε4/ε4 (MC0018) genotype were analyzed by ELISA at day 0, 10, 20, 30 and 45. (E) ApoE immunoreactivity in the culture media from those cells at day 45 and primary astrocytes from Apoe^-/- mice was analyzed by Western blotting after Native-PAGE and SDS-PAGE. Data are expressed as mean ± SEM (n = 3).

Human iPSC-derived astrocytes produce apoE along with their maturation

To address whether astrocytes derived from human iPSCs produce apoE, we measured the amount of apoE in serum-free conditioned media from those with APOE ε3/ε3 (MC0117) or APOE ε4/ε4 (MC0018) genotype by ELISA at day 0, 10, 20, 30 and 45 of differentiation from NPCs. Although the cells sparsely secreted apoE at day 0 and 10, we found that a robust increase of apoE secretion was observed starting from day 20 of astrocyte differentiation (Fig. 2D). The apoE levels were preserved in the subsequent differentiation stages (day 30 and 45). There was no evident difference between APOE ε3/ε3 astrocytes and APOE ε4/ε4 astrocytes in the pattern of apoE secretion during astrocyte differentiation (Fig. 2D). Consistently, we also confirmed the significant apoE production by astrocytes derived from other iPSC lines at day 50 of differentiation (Supplementary Material, Fig. S3). Furthermore, Western blotting detected apoE with different sizes in their condition media after Native-PAGE, while a single band for apoE was identified by SDS-PAGE (Fig. 2E). These results indicate that human iPSC-derived astrocytes secrete abundant apoE upon maturation, which likely forms lipoprotein particles.

ApoE produced by human iPSC-derived astrocytes with APOE ε4/ε4 has a lower lipidation status compared to those from astrocytes with APOE ε3/ε3

To investigate the lipidation status of apoE secreted by human iPSC-derived astrocytes with APOE ε3/ε3 or ε4/ε4 at day 45-55 of differentiation from NPCs, apoE particles in the conditioned media were assessed by Native-PAGE, followed by Western blotting. To evaluate the sizes of apoE particle, the apoE immunoreactivity was categorized into three groups; large particles (>669 kDa), medium particles (440–669 kDa) and small particles (<440 kDa) (Fig. 3A). While APOE genotypes did not influence the population of apoE particles with large size (Fig. 3B), we found significant APOE genotype-dependent effect on those with middle and small sizes; apoE particles from astrocytes with APOE ε4/ε4 were less abundant in the middle size category (Fig. 3C), but more abundant in the small size category (Fig. 3D), compared to those from the astrocytes with APOE ε3/ε3. These results from human iPSC-derived astrocytes indicate that apoE4 particles have a smaller size than apoE3 particles, suggesting a less lipidated property of apoE4 particles. To validate this, we next quantified the amounts of apoE-associated cholesterol in the media from human iPSC-derived astrocytes with APOE ε3/ε3 or ε4/ε4. After immunoprecipitation of apoE particles, the amounts of cholesterol co-precipitated with apoE were quantified. We found that apoE4 particles from human iPSC-derived astrocytes were associated with less amount of cholesterol (cholesterol/apoE; 6.208 ± 0.5718 mg/mg) than apoE3 particles (9.161 ± 0.7859 mg/mg) (Fig. 3E). Together, these results clearly indicate that apoE produced by human iPSC-derived astrocytes with APOE ε4/ε4 has a hypolipidated status compared to those from the astrocytes with APOE ε3/ε3, which is consistent with previous findings on apoE lipidation status in apoE-targeted replacement (TR) mice (36,37) and humans (38).

Figure 3.

Size and lipidation status of apoE lipoprotein particles produced by human iPSC-derived astrocytes with APOE ε3/ε3 or ε4/ε4 genotype. (A) ApoE in the culture medium was analyzed by Western blotting after Native-PAGE. ApoE immuno-reactivity was quantified in three particle sizes: Large particles (B; >669 kDa), Medium particles (C; 440–669 kDa) and Small particles (D; <440 kDa). The percentages of apoE particles in different size categories were quantified. (E) ApoE particles were isolated from the conditional media of iPSC-derived astrocytes with APOE ε3/ε3 or ε4/ε4 by immunoprecipitation, and the amounts of apoE-associated cholesterol were measured and calculated against apoE protein contents. Data are expressed as mean ± SEM (n = 8; 3 lines/genotype, 2–3 clones/line). N.S., not significant; **P < 0.01.

Human iPSC-derived astrocytes with APOE ε4/ε4 have inferior supportive effects on neuronal viability than those with APOE ε3/ε3

When iPSC-derived neurons with APOE ε3/ε3 (MC0192 #1) at day 14 of differentiation from NPCs were cultured in the growth factor-free medium for 7 days, the viability of neurons was substantially compromised, which was assessed by MAP2 staining and further confirmed by staining for live cells with calcein-AM and dead cells with ethidium homodimer-1 (Fig. 4). The number of MAP2 positive neurons was 109.3 ± 7.65 but reduced to 4.67 ± 0.82 after grow factor withdrawn for 7 days (Fig. 4A and B). Live/dead cell assay using FACS revealed that the population of live neurons was 60.06 ± 1.46% in normal growth medium but reduced to 4.09 ± 0.73% in growth factor-free condition (Fig. 4C). To investigate whether human iPSC-derived astrocytes have neuroprotective functions, we co-cultured human iPSC-derived neurons (day 14) and astrocytes (day 45) using Transwell chambers, which allowed secreted molecules to freely diffuse without direct astrocyte-to-neuron contact (Fig. 4A). The iPSC-derived neurons were placed in the bottom chamber of the co-culture system, whereas iPSC-derived astrocytes with APOE ε3/ε3 or ε4/ε4 were plated in the upper chamber. When those cells were similarly cultured under growth factor-free conditions, we found that iPSC-derived astrocytes significantly promoted neuronal survival in an APOE genotype-dependent manner. The MAP2 positive neuron number was significantly higher when cocultured with human iPSC-derived astrocytes with APOE ε3/ε3 (46.22 ± 2.73) compared to astrocytes with APOE ε4/ε4 (30.5 ± 1.41) (Fig. 4B). When analyzing the cell viability via FACS, the percentage of live neurons was significantly higher in the presence of human iPSC-derived astrocytes with APOE ε3/ε3 (31.83 ± 3.093%) compared to astrocytes with APOE ε4/ε4 (19.72 ± 1.365%) (Fig. 4C). Furthermore, FACS data also showed co-culturing with APOE ε3/ε3 astrocytes could restore the live/dead cell ratio of neurons in growth factor-free conditions from 0.06 ± 0.01 to 0.81 ± 0.13, in which the effect was stronger than that with APOE ε4/ε4 astrocytes (0.36 ± 0.03) (Fig. 4D). These results indicate that human iPSC-derived astrocytes have neuroprotective functions, which are less efficient in APOE ε4/ε4 than APOE ε3/ε3 astrocytes.

Figure 4.

Neuroprotective effects of human iPSC-derived astrocytes with APOE ε3/ε3 or ε4/ε4 genotype. (A) Human iPSC-derived neurons (day 14) with APOE ε3/ε3 (MC0192 #1) were co-cultured with or without iPSC-derived astrocytes carrying APOE ε3/ε3 or APOE ε4/ε4 genotype for 7 days in growth factor-free medium. Neurons were stained with neuronal marker MAP2, the number of MAP2 positive neurons were compared among different groups (A, B). Neuronal viability through Calcein AM and Ethi-D1 staining was analyzed via FACS (B, C). FACS data were analyzed and plotted differently as live cell percentage and live/dead cell ratio. Scale bars, 50 μm. Data are expressed as mean ± SEM (n = 6; 3 lines/genotype, 2 colonies/line, 6 fields/colony for MAP2+ neuron quantification). ^###P < 0.001 vs. growth factor free medium; *P < 0.05; ***P < 0.001.

APOE genotypes influence synaptic integrity in a co-culture system with human iPSC-derived astrocytes and neurons

To investigate the impact on synaptic integrity during neuronal maturation using human iPSC-derived cells, iPSC-derived NPCs (MC0192 #1) were stained for PSD95 (a post synaptic marker) and vGlut1 (a presynaptic marker) upon their differentiation. Immunoreactivity of PSD95 and vGlut1 became detectable in those cells accompanied with neural outgrowth after day 15 of differentiation from NPCs (Supplementary Material, Fig. S4A). ELISA measurements for PSD95 (Supplementary Material, Fig. S4B) and Western blotting for PSD95 and vGlut1 (Supplementary Material, Fig. S4C) confirmed their expression in iPSC-derived neurons at day 15, 21 and 30 of differentiation.

To assess whether human iPSC-derived astrocytes influence synaptic protein levels, human iPSC-derived neurons with APOE ε3/ε3 (MC0192 #1) (day 10) and astrocytes with different APOE genotypes (day 45) were co-cultured in Transwell chambers with normal growth medium. Western blotting showed that vGlut1 and PSD95 levels in iPSC-derived neurons were significantly increased in the presence of iPSC-derived astrocytes regardless of APOE genotypes after co-culture for 7 days (Fig. 5A) and 14 days (Fig. 5B). At day 20 of co-culture, however, we found that iPSC-derived astrocytes with APOE ε3/ε3 increased those synaptic proteins more significantly compared to those with APOE ε4/ε4 (Fig. 5C). Similar results were obtained when analyzed by immunostaining the iPSC-derived neurons for MAP2, vGlut1 and PSD95 after co-culture with astrocytes for 20 days (Fig. 5D). Taken together, these results demonstrate that human iPSC-derived astrocytes increase synaptic proteins in human iPSC-derived neurons, where APOE ε4/ε4 astrocytes have inferior effects compared to APOE ε3/ε3 astrocytes.

Figure 5.

Expression of synaptic proteins following iPSC-derived neurons co-cultured with iPSC-derived astrocytes carrying APOE ε3/ε3 or ε4/ε4 genotype. Human iPSC-derived neurons (day 10) with APOE ε3/ε3 (MC0192 #1) were co-cultured with iPSC-derived astrocytes carrying APOE ε3/ε3 or APOE ε4/ε4 in normal growth medium. (A–C) Expression of vGlut1 and PSD95 in the neurons was analyzed by Western blotting after co-culture for 7 days (A), 14 days (B), or 20 days (C). (D) The neurons were stained for PSD95 (red), vGlut1 (green) and MAP2 (purple) after co-culture for 20 days. Scale bars, 10 μm. Data are expressed as mean ± SEM (n = 6; 3 lines/genotype, 2 colonies/line). ^###P < 0.001 vs.no astrocyte group; ^##P < 0.01 vs.no astrocyte group; ^#P < 0.05 vs.no astrocyte group; N.S., not significant; *P < 0.05.

Discussion

Increasing evidence has demonstrated that apoE isoforms differentially regulate synaptic plasticity and repair both in physiological and pathological conditions (39,40). In AD and healthy aged controls, APOE ε4 allele dosage correlates inversely with dendritic spine density in the hippocampus (41). Animal experiments using apoE- TR mice have also shown that apoE4 lowered dendritic spine density and length compared with apoE3 (42). Thus, apoE isoform-dependent effects on neuronal integrity might relate to increased risk of cognitive decline in aged APOE ε4 carriers. In this study, we employed the co-culture system of human iPSC-derived astrocytes and neurons to investigate how APOE genotypes influence astrocyte-mediated functions on neuronal survival and synaptic integrity in human cells. Our findings have demonstrated that human iPSC-derived astrocytes provide beneficial effects on neuronal survival and synapses with astrocytes carrying APOE ε4/ε4 genotype being less efficient compared to those with APOE ε3/ε3.

Astrocytes play an important role in synaptic transmission and plasticity by regulating neurotransmitter trafficking and recycling, nutrient and ion metabolism, and releasing neuro-modulatory factors (7,43). In the adult brain, astrocytes intimately enwrap dendritic spines and presynaptic terminals forming the tripartite synapse, which in turn allows astrocytes to respond to synaptic activity and regulate synaptic transmission (43,44). Nonetheless, our co-culture system shows that the supportive role of astrocytes for neuronal survival and synaptic integrity does not require their direct contact. Consistently, recent studies have also found that astrocytes isolated from human brain could promote neuronal survival without their direct interaction (28). While astrocytes can secrete a variety of neurotrophic factors (45), cholesterol conjugated in apoE particles is an essential molecule secreted by astrocytes in regulating axonal growth, synaptic formation and remodeling in the brain (46,47). Cholesterol and other lipids are loaded onto apoE through a plasma membrane ATP-binding cassette transporter A1 (ABCA1) (34). Indeed, abnormal lipid metabolism is hypothesized to contribute to the pathogenesis of AD (9) and aging-related cognitive decline (9). It has been reported that cholesterol levels in serum, cell membranes of brains, and cerebrospinal fluid are decreased in AD patients compared with those in controls (48). Of note, we revealed that apoE particles secreted from iPSC-derived astrocytes with APOE ε3/ε3 carried more cholesterol than those from astrocytes with APOE ε4/ε4, which is consistent with results from primary astrocytes from apoE-TR mice (26). ApoE4-TR mice have also shown abnormal cholesterol levels and impaired lipid metabolism both in the brain and periphery (49). Therefore, the insufficient cholesterol supply from astrocytes with APOE ε4/ε4 to neurons likely contributes to compromised neuronal health, although further studies are needed to determine whether other growth factors are also involved in our observed phenotypes.

Because of the importance of apoE in the CNS, tremendous efforts have been devoted to defining the property and function of apoE isoforms through in vitro and in vivo experiments. In particular, animal models have provided fundamental insights into the role of apoE in neurodegeneration. However, given that significant differences exist between humans and rodents, there is an urgent need to establish a complementary platform to investigate apoE biology in human systems. The recent advance of iPSC technology enables human cell-based disease modeling. Human iPSCs are capable of self-renewal and differentiation to all cell types, including neurons (16) and astrocytes (50). By overcoming the limited accessibility to human brain cells, human iPSCs with specific genetic background represent an attractive strategy to generate highly enriched neuronal and astrocytic populations for modeling neurodegenerative diseases. While the generation of functional astrocytes from human iPSCs has been reported by several groups (23,51,52), we have in this study successfully differentiated iPSCs into astrocytes. Importantly, our human iPSC-derived astrocytes could produce abundant apoE particles and recapitulate the APOE genotype-dependent phenotypes. Although the variability of iPSCs depending on clones as well as culture conditions (53) should be carefully addressed, the usage of human iPSC-derived astrocytes with different APOE genotypes might be a promising approach for studying apoE biology and pathobiology.

In conclusion, our study indicates that astrocytes differentiated from human iPSCs can recapitulate certain phenotypes dependent upon APOE genotypes and therefore provide a valid in vitro human model to study the pathogenic mechanisms of neurodegeneration. We demonstrated apoE isoform-dependent properties using human iPSC-derived astrocytes; apoE3 has a more lipidated status than apoE4. We also showed that the astrocytes with APOE ε3/ε3 had greater supportive functions in neuronal survival and synaptogenesis compared to APOE ε4/ε4. Given that apoE and apoE-associated lipid metabolism are critically involved in regulating diverse aspects of brain functions, further studies using human iPSC-derived astrocytes might provide new insights into molecular mechanisms underlying the link among APOE genotypes, astrocytic biology, and aging-related cognitive decline. Our established platform composed of human iPSC-derived astrocytes with different APOE genotypes should also facilitate the development of novel apoE-targeted therapy for neurodegenerative diseases.

Materials and Methods

Generation of iPSCs from human skin fibroblasts

Human skin biopsies from normal individuals with APOE ε3/ε3 or ε4/ε4 genotype were obtained from Mayo Clinic patients under approved IRB protocols with patients’ consent for research purpose. The patient information was de-identified before skin biopsies were processed for fibroblasts in the Mayo Clinic Neuroregeneration Lab. The use of patient-derived fibroblasts, iPSCs and iPSC-derived cells is approved under a separate IRB protocol. APOE genotype was confirmed by Sanger sequencing using DNA samples from the fibroblast lines. Cells were cultured in DMEM (Invitrogen) containing 10% fetal bovine serum (FBS) (Gemini Bio-Products), and supplemented with 1% non-essential amino acids (NEAA) (Invitrogen), 1% Penicillin-Streptomycin (Invitrogen), and 1% Amphotecerin B (Gemini Bio-Products). The iPSCs were generated by electroporation of three episomal vectors containing, OCT3/4, SOX2, KLF4, L-MYC, LIN28, and p53-shRNA (Addgene) (18) into the fibroblasts using the NHDF nucleofector kit (Lonza) as described (19). After nucleofection, fibroblasts were plated onto a 100 mm dish coated with Matrigel (Corning). Cells were cultured in fibroblast medium for 5 days, then the medium was replaced with TeSR-E7 complete medium (Stemcell Technologies) and changed every day. The iPSC colonies were isolated and expanded after approximately 3–4 weeks in culture. The iPSC colonies were passaged using Dispase (Stemcell Technologies) and subjected to treatment with rock inhibitor Y27632 (Sigma-Aldrich) for the first 24 h. Karyotyping of the iPSC clones was performed by KaryoLogic, Inc. A human ES line (HUES2, Harvard Stem Cell Science) and an iPSC line (ATCC, ACS-3002) were used as positive controls.

Trilineage differentiation of human iPSCs

The pluripotency of iPSCs was confirmed by three germs layer differentiation using STEMdiff Trilineage Differentiation kit (Stemcell Technologies) according to the manufacturer’s instructions (20). Briefly, when cells were approximately 70% confluent, iPSCs were passaged with Accutase (Stemcell Technologies) and plated onto 24-well plates with mTeSR1 medium. Medium was replaced to specific differentiation medium for each lineage after 24 h, and cells were subjected to the differentiation into mesoderm and endoderm lineages for 5 days or an ectoderm lineage for 7 days. Differentiation was assessed with immunostaining for specific markers of each germ layer (Ectoderm: Nestin/Sox2; Mesoderm: Brachyury; Endoderm: SOX17).

RNA extraction and RT-PCR

Total cellular RNA was extracted from cells using RNeasy mini kit (Qiagen), according to the manufacturer’s instructions, and subjected to DNase I digestion to remove contaminating genomic DNA. Reverse transcription was performed using SuperScript III First-Strand Synthesis System (Invitrogen). Real-time qPCR was conducted with Universal SYBR Green Supermix (Bio-Rad) using an iCycler thermocycler (Bio-Rad). The 2exp (−ΔΔCt) method was used to determine the relative expression of each gene with β-actin used as a reference. The primers used to amplify target genes by RT-PCR and quantitative PCR were as follows; endogenous hOCT3/4 F (5’-GAC AGG GGG AGGGGA GGA GCT AGG-3’) and R (5’-CTT CCC TCCAAC CAG TTG CCC CAA AC-3’), endogenous hSOX2 F(5’-GGG AAA TGG GAG GGG TGC AAA AGA GG-3’) and R (5’-TTG CGT GAG TGT GGA TGG GAT TGGTG-3’), β-actin F (5’-CTG GCA CCA CAC CTT CTA CAA TG-3’) and R (5’-AAT GTC ACG CAC GAT TTC CCG C-3’).

Differentiation of human iPSCs into neurons and astrocytes

For neural differentiation, iPSCs were first cultured in commercial neural induction medium (Stemcell Technologies) (21) following manufacturer’s instruction with some modifications. In 6-well non-tissue culture treated plates (Corning), iPSC clumps were cultured in neural induction medium in suspension for 5-7 days to initiate neurosphere formation. Next, neurospheres were seeded onto Matrigel-coated dishes and cultured in neural induction medium for another 5-7 days to induce neural rosette formation. Neural rosettes were isolated as a single cell suspension and re-plated onto Matrigel-coated dishes in neural induction medium. To differentiate into neural progenitor cells (NPCs), the medium was replaced to neural progenitor cell medium (Stemcell Technologies) and cultured for additional 10-14 days. NPCs were amplified and frozen stocks were made for further experiments.

Neuron differentiation from NPCs was accomplished by culturing on PLO/Laminin-coated plates in neuronal differentiation medium, which was composed of DMEM/F12 + Neurobasal Medium (1: 1) supplemented with N2, B27, BDNF (20 ng/ml), GDNF (20 ng/ml), NT3 (10 ng/ml), IGF (10 ng/ml), ascorbic acid (200 μM) (all from Stemcell Ttechnologies) and dbcAMP (100 nM) (Sigma Aldrich) (22).

For astrocyte differentiation, NPCs were cultured on PDL-coated plates in astrocyte differentiation medium composed of astrocyte medium (ScienCell) with CNTF (10 ng/ml), BMP4 (10 ng/ml) and Heregulin-β (10 ng/ml) (all from Stemcell Technologies). Cells were passaged during the differentiation process when they reached 80% confluence (23,24).

Immunocytochemistry

Cells were fixed in 4% paraformaldehyde and then permeabilized with 0.25% Triton X-100 in PBS. After blocking with 1% BSA in PBS for 30 min, cells were incubated with primary antibodies overnight at 4 °C. After washing with PBS, cells were incubated with Alexa Fluor-488, 568 or 647 conjugated secondary antibodies for 1 h at room temperature. Fluorescent signals were detected by fluorescence microscopy (model IX71 Invert, Olympus) or confocal laser scanning fluorescent microscopy (model LSM510 Invert, Carl Zeiss), and images were processed using Photoshop. The information of primary antibodies and their dilutions used in this study were as follows; Nanog (Cell Signaling, Cat# 4903, 1: 300), TRA-1-60 (Abcam, Cat# ab16288, 1: 300), SSEA4 (Abcam, Cat# ab16287, 1: 300), Sox17 (Abcam, Cat# ab84990, 1: 300), Brachyury (R&D, Cat# AF2085, 1: 300), AQP4 (Santa cruz, Cat# sc20812, 1: 100), Vimentin (Abcam, Cat# ab92547, 1: 500), MAP2 (Novus biological, Cat# NB300-213, 1: 500), PSD95 (Abcam, Cat# AB2723, 1: 300), vGlut1 (Abcam, Cat# 77822, 1: 300), Nestin (Abcam, Cat# ab18102, 1:500), Sox2 (Abcam, Cat# ab97959, 1:500), GFAP (Abcam, Cat# ab7260, 1:300), s100β (Sigma, Cat# s2532, 1:500), and PAX6 (Abcam, Cat# ab5790, 1:300).

FACS of iPSC-derived NPCs

Cells were removed from the plate using Cell Dissociation Solution (Sigma), dissociated cells were fixed with 4% paraformaldehyde (PFA) for 15 min at room temperature (RT) and subsequently permeabilized with 0.2% Triton X-100 for 15 min. Cells were incubated with Alexa Fluor 647 anti-Nestin antibody (BD560393) at room temperature for 40 min. Cells were analyzed for fluorescence on a BD FACS Calibur machine (BD Biosciences, Sparks, MD). Cells incubated with mouse IgG1, κ isotype (BD557732) were used as a control.

Glutamate uptake assay

Human iPSC-derived astrocytes and HEK 293FT cells were treated with HBSS containing 20 μM glutamate for 10 min or 30 min at 37 °C, and the remaining glutamate in the medium was measured using the Glutamate Colorimetric Assay kit (Abcam) (25). The absorbance of the product was measured at 492 nm using a microplate reader (Biotek). Cellular glutamate uptake was calculated by subtracting the remaining glutamate in the medium after incubation with the cells. Cellular glutamate uptake measurements were normalized by protein concentration of cell lysates of each sample.

ApoE ELISA

ApoE ELISA was performed according to our published work (26). Briefly, 96-well plates were coated overnight with an apoE antibody (AB947, Millipore) in carbonate buffer at 4 °C overnight. The plates were blocked with 1% Block Ace in PBS, and then washed 3 times with PBS. Recombinant apoE3 and apoE4 (Fitzgerald) were used as standards for ELISA. Samples were diluted and incubated at 4 °C overnight. The plates were washed and incubated with biotin-conjugated goat anti-apoE antibody (Meridian Life Science) for 2 h at room temperature. After incubation with Horseradish Peroxidase Avidin D (Vector Laboratories) for 90 min at room temperature, the plate was developed by adding tetramethylbenzidine Super Slow substrate (Sigma). The reaction was stopped and read at 450 nm with a microplate reader (Biotek).

Quantification of apoE particles by native-PAGE

Human iPSC-derived astrocytes were cultured in serum-free astrocyte medium for 24 h, and the conditioned medium was concentrated 50 times via Amicon Ultra-15 Centrifugal Filter Units (10KD, Millipore) after filtration through a 0.45 µm membrane. The concentrated medium was loaded into Native PAGE™ Novex 4–16% Bis-Tris gels (Invitrogen) on ice following the manufacturer’s instructions, and transferred to PVDF Immobilon FL membranes (Millipore). Membranes were blotted with primary antibodies against apoE (K74180B, Meridian Life Science, 1:1000) in 5% non-fat milk containing 0.01% Tween-20. Odyssey IR680 or IR800 secondary antibodies against the species of the primary antibody were incubated for 1 h at room temperature. Immuno-reactive bands were detected and quantified using the odyssey infrared imaging system (LI-COR Biosciences). The Unstained Protein Standard from GE was used for estimation of particle sizes. Membanes were treated with Ponceau S Staining Solution (0.1% (w/v) Ponceau S in 5% (v/v) acetic acid) to visualize the molecular weight markers.

Immunoprecipitation of apoE particles and cholesterol assay

ApoE particles were immunoprecipitated as previously described (27). Briefly, streptavidin-conjugated agarose beads (Sigma) were first incubated with biotin-conjugated goat anti-apoE antibody (K74180B, Meridian Life Science) for 2 h at room temperature with shaking. The antibody-bound agarose beads were then mixed with the concentrated medium from iPSC-derived astrocytes at 4 °C overnight. After washing, the agarose beads were re-suspended in 100 µl of TBS with 0.1% Triton X-100. The amounts of cholesterol dissolved in the supernatant were measured by Amplex Red cholesterol assay (Invitrogen) according to the manufacturer’s protocol. Finally, the beads were suspended in 100 µl of 0.1 M glycine buffer (pH 2.5) followed by neutralization with 1 M Tris-HCl pH 8.8 plus 0.2% BSA, and the amount of apoE in the mixture was measured by ELISA. The amount of apoE-associated cholesterol was determined by normalizing total cholesterol quantities with total apoE quantities in the eluent.

Co-culture system of human iPSC-derived neurons and astrocytes

For neuron survival experiment, human iPSC-derived NPCs were differentiated into neurons on PLO/Laminin-coated plates for 10 days. Human-iPSC-derived astrocytes at day 45 of differentiation were plated on PDL-coated cell culture inserts with 0.4 µm diameter pores (Corning) and conditioned in neuronal differentiation medium with Ara-C (1 µM) for 24 h. Finally, the neurons were co-cultured with the insert containing astrocytes in growth factor-free medium without BDNF, GDNF, NT3 and IGF (28). 7 days upon coculture, neurons were stained with neuronal marker MAP2 (Abcam, Cat# ab5392, 1:500). Fluorescent signals were detected by fluorescent microscopy. 6 fields for each colony were randomly chosen. MAP2 positive cells were counted manually using ImageJ. The viability of neurons was also determined with the Live/dead Viability/Cytotoxicity kit (Invitrogen, L3224) according to manufacturer’s instructions at 7 days upon co-culture.

For the analysis of synaptic proteins, the iPSC-derived neurons and astrocytes were co-cultured in normal neural differentiation medium. Neurons were lysed in 1% Triton X-100 PBS containing 1% protease inhibitor cocktail (Roche), and centrifuged at 13,000 g for 10 min at 4 °C. Samples were subjected to Western blotting after SDS-PAGE. Primary antibodies against PSD95 (Abcam, Cat# ab2723, 1:1000), vGlut1 (Abcam, Cat# ab180188, 1:1000) and anti-Tuj1 (Sigma, Cat# T2200, 1:2000) were used. ELISA for PSD95 was conducted according to our published work (29).

Statistical analysis

All data were analyzed by student t-test or one-way analysis of variance (ANOVA) with a Tukey’s posthoc test using GraphPad Prism 5. Data were presented as Mean ± SEM. A P value of < 0.05 was considered statistically significant.

Supplementary Material

Supplementary Material is available at HMG online.

Conflict of Interest statement. None declared.

Funding

NIH grants RF1AG051504, P50AG016574, R01AG027924, R01AG035355, R01AG046205, and P01 NS074969 (to G.B.), and R21AG052423 (to T.K.), a grant from Cure Alzheimer’s Fund (to G.B. and T.K.), and Mayo Clinic Center for Regenerative Medicine grants (to G.B. and T.K.) and postdoctoral fellowship (to J. Z.). Z.K.W. was partially supported by the NIH/NINDS P50 NS072187, NIH/NIA (primary) and NIH/NINDS (secondary) 1U01AG045390-01A1, Mayo Clinic Center for Regenerative Medicine, Mayo Clinic Center for Individualized Medicine, Mayo Clinic Neuroscience Focused Research Team (Cecilia and Dan Carmichael Family Foundation, and the James C. and Sarah K. Kennedy Fund for Neurodegenerative Disease Research at Mayo Clinic in Florida), the philanthropic support to the Mayo Clinic Florida as the gifts from Carl Edward Bolch, Jr., and Susan Bass Bolch, The Sol Goldman Charitable Trust, and Donald G. and Jodi P. Heeringa.