An Overview of in vitro Methods to Study Microglia

1Neurodegeneration is defined as the progressive loss of functional neurons. This can be the selective loss of a particular neuronal subtype, such as occurs in diseases as Parkinson’s disease (PD) and amyotrophic lateral sclerosis (ALS), or the widespread loss of many neuronal subtypes, such as occurs in Alzheimer’s disease (AD) and Huntington’s disease (HD). Although all classified as neurodegenerative diseases, the underlying central nervous system (CNS) pathologies are different.

2PD pathology is characterized by the formation of Lewy bodies in dopaminergic neurons consisting of fibrillar α-synuclein, whereas ALS is characterized by protein-rich cytoplasmic inclusions in motor neurons of the spinal cord. AD pathology is characterized by the intracellular accumulation of hyper phosphorylated tau protein and neurofibrillary tangles and by the extracellular deposition of amyloid β (Aβ) in senile plaques. HD pathology is characterized by neuronal intranuclear inclusions consisting of mutant huntingtin protein. Although the progress, etiology and symptoms of these diseases differ, neuroinflammation is a common hallmark of all of them.

3How neuroinflammation contributes to the progression of neurodegenerative diseases is still unclear as it can either be the cause or the consequence of neuronal cell death. It is, however, generally accepted that persistent inflammation of the CNS is detrimental to neurons. Intriguingly, some molecules that are associated with the pathology of neurodegenerative diseases, such as Aβ and α-synuclein, can induce or modulate inflammatory responses via receptors of the innate immune system thereby providing a molecular link between both processes.

4Microglia express many receptors of the innate immune system and have a key role in neuroinflammation. Although microglial responses are thought to be primarily neuroprotective, they may also lead to tissue injury and neurodegeneration by the production of pro-inflammatory cytokines and reactive oxygen and nitrogen species (ROS/RNS). There is a large body of evidence for microglial activation in the pathogenesis of neurodegenerative disorders. Activation of microglia is characterized by an amoeboid morphology, by the production of cytotoxic molecules and pro-inflammatory cytokines, and by the increased expression of complement receptors and histocompatibility complex molecules. In the substantia nigra of PD patients, reactive microglia are found along with Lewy bodies and large numbers of activated microglia can be observed in the CNS and spinal cords of human ALS patients as well as in ALS mouse models. Microglia that surround plaques in AD change their morphology from ramified to amoeboid and stain positive for activation markers. Finally, many of the genes that were identified as risk factors for the development of AD in genome-wide association studies such as TREM2, ApoE, ABCA7, PICALM, or CD33 are expressed by microglia.

5Together, these observations fuel the thought that targeting microglia might provide benefit for those afflicted by neurodegenerative diseases. Detailed cellular biological knowledge of microglia is therefore crucial, and in vitro models are instrumental in obtaining such knowledge.

6Microglia were first described by Rio-Hortega early in the 20th century as non-neuronal elements that derive from oligodendroglia and astroglia. Despite intensive research, the origin of microglia has long remained a controversial issue. Researchers described microglia as cells derived from mesodermal pial elements, from pericytes and from neuroectodermal macroglia. Whereas it was already proposed that microglia derive from yolk-sac macrophages in 1999, conclusive evidence was only provided a decade later when it was shown that microglia originate from yolk-sac primitive myeloid progenitor cells.

7In mice, migration and colonization of yolk-sac derived macrophages to and into the brain starts between E8 and E10. During embryogenesis and throughout adult life, microglia are maintained by local self-renewal without replenishment from hematopoietic progenitors. Thereby, they form a distinct population from circulating blood monocytes and hematopoietic macrophages.

8Even under homeostatic conditions, microglia continuously sample the CNS environment with their highly motile processes and can hardly be described as “resting” but rather as neutral or “M0”. M0 microglia are subject to multiple inhibitory signals from the CNS environment including that of transforming growth factor beta (TGF-β). TGF-β is a soluble factor with general immunosuppressive properties that is constitutively expressed in the CNS. In addition, cell-cell interactions between microglia and neurons mediated by e.g., CD200R-CD200 contact provide constitutive inhibitory signals to microglia. Loss or disruption of constitutive inhibitory signaling leads to a different microglia phenotype, which is characterized by the increased expression of activation markers, such as CD11b and CD45. In addition, microglia lose their ramified morphology and can form aggregates, which are normally not observed in healthy CNS tissue. This implicates that inhibitory signals from the CNS environment are important to maintain the microglial M0 phenotype. It is important to realize that the M0 phenotype in itself might vary with age and there is also evidence for CNS regional differences.

9Microglia can switch phenotype when exposed to specific growth factors or cytokines. Classically, in vitro exposure to interferon gamma (IFNγ) and/or lipopolysaccharide (LPS) has been associated with morphological alterations from ramified to amoeboid and with the induction of an activated or “M1” phenotype. This phenotype has long been associated with neuroinflammation. Alternatively, in vitro exposure of microglia to anti-inflammatory cytokines like interleukin (IL)-4 can induce an alternative, “M2” phenotype. More recent studies have demonstrated that macrophages and microglia can display a wide spectrum of intermediate phenotypes, both in vitro as well as in vivo. The recent identification of the ApoE pathway as a driver of microglial phenotype alteration in AD has casted further doubt on the usefulness of the classical dichotomous M1-M2 phenotype description.

10As resident innate immune cells of the brain, microglia provide the first line of defense against invading pathogens, such as viruses, bacteria and prions. Like other macrophages, microglia are phagocytic cells that can secrete a wide range of chemokines and pro- and anti-inflammatory cytokines. While microglial inflammatory responses have been the focus of much research, there is increasing appreciation of the contribution of microglia to CNS development and homeostasis. Microglia secrete neurotrophic and growth factors that regulate the proliferation of oligodendrocytes, astrocytes and neuronal progenitors, and they contribute to the maturation of neural circuits. During CNS development and adult neurogenesis, microglia contribute to the clearance of superfluous neurons. Furthermore, microglia are involved in synapse elimination or synaptic pruning, which is required to establish efficient neuronal networks. Inappropriate synaptic connections are tagged by complement components C1q and C3, and can be recognized by microglia that express the complement C3 receptor.

11As most neurodegenerative diseases develop at adult age, in vitro models that aim to study the role of microglia in the pathogenesis of neurodegenerative disorders should ideally recapitulate the M0 phenotype of adult or elderly human microglia. This will allow exposure to different pathology-associated stimuli in order to characterize microglial responses and to test therapeutic interventions in such responses. Recent publications of RNA transcriptome profiles of ex vivo M0 microglia from different species and age have opened up previously shut doors. Not only have these studies led to the identification of microglia-specific markers such as P2RY12 and TMEM119, they have also provided a blueprint of the RNA transcriptome profile that in vitro microglia should ideally express.

12Over the years various microglia in vitro models have been developed, including microglia cell lines, stem cell-derived microglia cultures and primary dissociated cell cultures. Each of these models has specific advantages and limitations, which will be discussed in detail.

13Microglia cell lines are available from mouse, rat, macaque and human origin (Table 1). Most of these lines stem from primary microglia cultures derived of the brain or the spinal cord, which were immortalized by viral transduction with oncogenes (e.g., v-myc, v-raf, v-mil, SV40 T antigen). Non-transformed microglia cell lines that stem from primary microglia precursor cell cultures have been described as well. Advantages of cell lines include their ease of maintenance and their abundant availability due to their unrestricted proliferative capacity. A major disadvantage is their susceptibility to dedifferentiation. Furthermore, viral transformation or immortalization may alter the microglial phenotype. Recent studies have pointed out that microglia cell lines differ both genetically and functionally from primary microglia and ex vivo microglia. In addition, microglial cell lines obtained from neonatal or embryonic CNS sources are unlikely to reflect the phenotype of adult or elderly microglia. Despite these limitations, microglia cell lines are suitable for e.g., biochemical and molecular approaches as well as for high-throughput screening assays which all require high cell numbers.

14A large variety of mouse microglia cell lines has been generated over the years. We provide a non-exhaustive list in Table 1 and will briefly discuss the two most used mouse cell lines, BV2 and N9. The BV2 cell line was generated by transduction of neonatal primary microglia with the v-raf/v-myc carrying J2 retrovirus. Immortalized, transformed cells survived for more than 4 weeks in culture, while non-transformed cells died. BV2 cells express macrophage markers such as macrophage antigen (MAC) 1 and MAC2, and are negative for the astrocyte marker glial fibrillary acidic protein (GFAP) and the oligodendrocyte marker galactocerebroside (GalC). The expression of MAC2 (also known as galectin-3) in particular is important when considering murine microglial cell lines to model neuroinflammatory processes. BV2 cells have been used for many years to study neuroinflammation and neurodegenerative disorders, including AD and PD. BV2 cells are responsive to LPS, they have phagocytic capabilities and increase their expression levels of ROS/RNS and pro-inflammatory genes after exposure to Aβ fibrils or to α-synuclein.

15Embryonic primary microglial cultures that were transformed with v-myc or v-mil oncogenes of the avian MH2 retrovirus led to the generation of a series of clonally derived microglial cell lines, of which N9 is the most well studied. N9 cells express Immunoglobulin G receptors, glycoprotein F4/80 and MAC1, and are negative for GFAP and GalC. Similar to BV2 cells, N9 cells produce and secrete pro-inflammatory cytokines after LPS stimulation and can phagocytose Aβ fibrils.

16The highly aggressively proliferating immortalized (HAPI) cell line is derived from neonatal primary cultures enriched for microglia and was the first cell line recognized to be the result of spontaneous immortalization. The genetic mutation that is responsible for immortalization is unknown yet. HAPI cells express the microglial markers isolectin B4, OX-42 and fructose transporter GLUT5, and are negative for GFAP and the oligodendrocyte marker A2B5. Exposure to LPS induces secretion of tumor necrosis factor α and production of ROS/RNS, and HAPI cells are capable of phagocytosis as demonstrated by the uptake of fluorospheres that were added to the culture medium.

17Cell lines derived from rodents are frequently being used to study microglial functions. As recent studies have revealed important differences between rodent microglia cell lines and primary microglia as well as between rodent and human microglia both in terms of aging and function, results obtained with rodent microglia cell lines should always be extrapolated with care.

18The HMO6 cell line is a human microglial cell line generated by transduction of embryonic primary microglia from telencephalon tissue with a v-myc carrying PASK 1.2 retroviral vector. Of the transducted cells, 99% remained positive for the macrophage/microglia markers Ricinus communis agglutinin-1 lectin and CD11b, and negative for the neuronal marker neurofilament-medium, the astrocyte marker GFAP and the oligodendrocyte marker myelin basic protein. ATP responsiveness and phagocytotic capacity of HMO6 cells were grossly comparable to that of human embryonic primary microglia. However, detailed characterization of the induced protein profile after exposure to LPS or Aβ25-35 revealed that HMO6 cells secreted a markedly less diverse cocktail of soluble mediators compared to primary microglia. Although gene expression analysis demonstrated that after 6 h exposure to LPS or Aβ25-35, mRNA expression levels of most soluble inflammatory mediators were enhanced in HMO6 cells comparable to levels in primary microglia, exposure to LPS did not induce the expression of mRNA encoding for macrophage inflammatory protein (MIP)-1α or introduce tumor necrosis factor alpha (TNF-α) in HMO6 cells whereas it did in primary microglia. More importantly, exposure of primary human microglia to LPS induced the secretion of IL-1β, IL-6, IL-8, TNF-α, and MIP-1α proteins, and exposure to Aβ25-35 induced the secretion of IL-1β, IL-8, TNF-α, and MIP-1α proteins. Exposure of HMO6 cells to LPS or Aβ25-35 only triggered the secretion of IL-8 and TNF-α proteins. Whether this lack of responsiveness is attributable to the transformation process is unclear yet. Although the HMO6 cell line has long been the only human microglia cell line available, the patented status of the cells has most probably inhibited its widespread use.

19Other immortalized human microglial cell lines that have been described include HMC3 and C13NJ, which both originate from the CHME-5 cell line. Importantly, a recent study has found evidence that CHME-5 cells are not of human, but of rat origin, implying that also the HMC3 and C13NJ cell lines are of rat origin. This warrants further investigation and until this is elucidated these cell lines should be used with caution. Interestingly, the group that uncovered the origin of CHME-5 has itself developed a method to immortalize adult primary microglia from mouse, macaque and human origin, transforming the cells with either the SV40 large T antigen alone or in combination with human telomerase reverse transcriptase (hTERT).

20Stem cell technology not only holds great promise for regenerative medicine, it can also provide scientists with an unlimited availability of cells for in vitro purposes. The two types of stem cells most often described in the context of microglia are embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs). ESCs are derived from the inner cell mass of a blastocyst, whereas iPSCs can be generated from adult cells (e.g., fibroblasts) by reprogramming them via overexpression of just four transcription factors. A major advantage of the iPSC approach is that it allows comparisons of iPSC-derived cells from healthy donors and patients with neurological disorders. Thereby, the genetic background of these patients is recapitulated in their iPSC-derived neurons and glial cells.

21Many protocols have been established to differentiate ESCs and iPSCs to specific neuronal lineages, such as neurons, astrocytes and oligodendrocytes. Microglia have, however, proven to be amongst the most difficult cells to differentiate from stem cells, partly because their exact origin remained obscure until 2010. The first microglia-related stem cell studies described methods to differentiate mouse ESCs to microglia by directing ESCs through neuronal differentiation pathways. Since lineage tracing studies in mice revealed that microglia originate from primitive yolk-sac macrophages, more recent protocols direct differentiation of ESCs and iPSCs to embryonic macrophage precursors first, before skewing these toward a microglial phenotype.

22In 2016 Muffat and colleagues described a method in which they differentiated human ESCs and iPSCs into neuralized embryonic bodies (EB) and cystic EBs. When plated on poly-D-lysine for 14 days, the cystic EBs became positive for markers of early yolk-sac myelogenesis, such as PU.1. Yolk-sac EBs were collected and replated in polystyrene plates for 30 days after which semi-adherent round cells appeared that were highly phagocytic and highly motile. Subsequently, these cells were kept in culture for 30 days to mature into microglia-like cells, bringing the total culture time to around 75 days (Figure 1). Transcriptome analysis showed that the mRNA expression profile resembles that of human fetal microglia.

23Pandya and colleagues reported the differentiation of human iPSCs into microglia-like cells in around 30 days by exposure to defined factors followed by co-culture with astrocytes. First, iPSCs were cultured for 4 days in medium containing specific VEGF, BMP4, SCF, and ActivinA, after which at day 4, 7, and 10, cells were exposed to another type of defined medium. On day 15, cells expressed CD34, CD45, and CD43, which are markers for myeloid progenitor cells. These cells were further co-cultured with human astrocytes in medium containing IL-3, granulocyte-macrophage colony-stimulating factor (GM-CSF) and macrophage colony-stimulating factor (M-CSF). After another 1 to 2 weeks of culture, these cells differentiated into microglia-like cells (Figure 1). Transcriptome analysis demonstrated that the gene expression profile of these iPSC-derived microglia resembles that of human fetal microglia, but also that of dendritic cells and macrophages. This immature phenotype warrants caution when considering the use of these cells to model adult microglia in neurodegenerative diseases.

24Around the same time Abud and colleagues published a protocol describing how iPSCs can be differentiated into microglia-like cells in just over 5 weeks. First, iPSCs were differentiated into CD43+ myeloid progenitors for 10 days using defined medium and temporal exposure to low (5%) oxygen levels (a procedure often used for the generation of myeloid progenitor cells). After 10 days, medium was changed to serum-free microglia differentiation media containing M-CSF, IL-34, TGF-β1, and insulin. At day 35, microglia-like cells were exposed to CD200 and CX3CL1 for 3 days to induce maturation (Figure 1). Gene expression analysis demonstrate that these microglia-like cells cluster with human fetal and adult primary microglia.

25Douvaras and colleagues published a protocol in which ESCs and iPSCs were first differentiated into myeloid progenitor cells by serial exposure to defined media. Myeloid progenitor cells were purified, replated and exposed for 2 weeks to IL-34 and GM-CSF, which resulted in differentiation into microglia-like cells after around 60 days of total culture (Figure 1). The obtained ESC/iPSC-derived microglia express microglia markers and resemble human fetal primary microglia by gene expression profile.

26The simplest protocol published describes a period of around 6–8 weeks to generate, with high yields, microglia from iPSCs. The protocol is relatively simple and avoids repeated replating or cell sorting. iPSCs were grown into EBs, and after approximately a month embryonic macrophage precursors emerged in the culture supernatants. These cells were harvested and subsequently co-cultured for 2 weeks with iPSC-derived cortical neurons. The obtained microglia-like cells are phagocytic, adopt a highly dynamic ramified microglia-like morphology and have a transcriptional profile similar to that of human fetal primary microglia. On the neuronal side, co-culture with iPSC-derived macrophage precursors retained neuronal maturity and functionality for at least 42 days.

27As this is a new, rapidly emerging field, there is no consensus on methodology to generate iPSC or ESC-derived microglia yet. A variety of different culture media and culture conditions have been used, and comparative studies and harmonization are necessary to further validate the most reliable and reproducible approaches. Although stem cell technology can provide researchers with a readily available source of microglia, it should also be taken into account that these cells have never been exposed to the CNS microenvironment. How this lack of exposure to CNS-specific environmental cues might affect microglial differentiation and function will be discussed below. In addition, most neurodegenerative disorders develop at adult or elderly age and it is therefore important to recapitulate age-related characteristics in microglia when neurodegenerative disorders are studied. However, reprogrammed iPSCs from adult donors have had their aging signature, such as telomere attrition and cellular senescence, reset. Direct reprogramming of somatic cells to microglia might tackle this problem by avoiding passage through the stem cell phase. It has been demonstrated that direct reprogramming retains aging-associated transcriptomic signatures.

28Methods to generate dissociated single cell cultures of primary microglia have been described for mice, rats, non-human primates and humans (Table 2). Most methods start with mechanical and enzymatic dissociation of the donor brain tissue followed by a density gradient centrifugation step to separate the myelin from the cells. Dependent on the density gradient used, this can either be sufficient to obtain microglia cultures with a purity of > 99% or it is followed by additional purification steps. Other purification steps used to isolate microglia include labeling of cells with antibody-coated magnetic beads followed by magnetic-activated cell sorting (MACS), labeling of cells with fluorescently labeled antibodies followed by fluorescence-activated cell sorting (FACS) or specific shaking procedures. Primary microglia from mice and rats are generally derived from brain tissue of neonatal animals, although more studies are now reporting the use of adult animals as well. The advantage of using rodent primary microglia is that these animals form a genetic homogenous, specific pathogen free (SPF) population where ante-mortem conditions and post-mortem delay can be tightly controlled. The use of primary microglia derived from transgenic mice has been instrumental in delineating the role of specific genes in microglia activation. Limitations of rodent primary microglia include their evolutionary divergence from humans and lack of heterozygosity due to inbreeding and their aseptic housing conditions. Differences between rodents and humans have been described to hamper translation of rodent (neuro) immunological studies to the clinic.

29Dissociated cultures of human primary microglia can either be derived from fetal tissue that becomes available after abortion or from post-mortem brain tissue that becomes available from deceased human donors. Isolation from brain material of patients who suffered neurological disease may provide new insights into the role of microglia in the pathogenesis. To enable research using human brain tissue, brain banks have been set up worldwide. Human primary microglia are derived from different individuals reflecting the genetic variability within a population and translation of results is not hampered by the use of a genetically divergent species. Limitations of human primary microglia include the limited availability of (healthy) human brain tissue, and the limited control over the ante mortem conditions and post-mortem delay, which might affect the microglia phenotype. For example, CD11b expression, a marker for immunoreactivity, shows a significant positive correlation with post-mortem delay in gray matter microglia.

30To bridge the gap between rodents and humans, primary microglia cultures derived from non-human primates may be considered. Protocols have been developed for the rhesus macaque (Macacca mulatta). To isolate such cells, the presence of a research center with non-human primate facilities is a requirement, which might be considered as a limitation. Advantages on the other hand are that microglia are isolated from outbred individuals that are in close evolutionary proximity to humans with much control over ante mortem conditions and post-mortem delay. Comparison of primary microglia with primary bone marrow-derived macrophages from the same donors has been instrumental in uncovering microglia specific features of innate immune responses, demonstrating the utility of this methodology.

31At least equally important as the method of isolation, are the in vitro culture conditions. Both for stem cell-derived microglia as well as of primary dissociated microglia, many different cell culture media combined with a diversity of growth factors have been tested. Data from knock out mouse studies demonstrated the importance of the colony stimulating factor-1 (CSF-1) receptor for microglial survival and proliferation leading to the inclusion of CSF-1 (M-CSF) in most microglia media. The discovery of IL-34 as a second, brain-specific, ligand for the CSF-1 receptor has inspired researchers to experiment with this factor as well. In addition, RNA transcriptome comparisons of ex vivo microglia with in vitro microglia identified the TGF-β pathway as important. These data also demonstrated that the RNA transcript profiles of ex vivo microglia differed considerably from those of in vitro microglia. A recent publication describes that at least part of this difference can be attributed to culturing in the presence of serum. Besides the fact that in vivo, in a healthy CNS, microglia are not exposed to serum, serum exposure has more disadvantages. Serum is a poorly defined cell culture component and batch-to-batch variability negatively contributes to reproducibility. In vitro cultures of adult primary microglia under serum-free conditions are, however, characterized by low proliferative capacity and decreased survival rates. Identification of cholesterol as the minimal component to confer increased survival of microglia in the absence of serum is a major step forward toward better defined cell culture conditions. The RNA transcript profile of ex vivo M0 microglia will continue to provide us with clues on cell culture conditions to better mimic this profile in vitro, and it is reasonable to expect major modifications to both stem-cell derived as well as primary microglial cell culture conditions in the near future.

32Research on the importance of the CNS microenvironment in retaining the M0 phenotype of microglia has gained momentum over the past years. Regretfully, microglial cell lines have long lost their specific microenvironmental input and stem cell-derived microglia have never even received it. But also adult primary microglia, although they have been “educated” in the CNS microenvironment for the life span of the donor, are deprived of microenvironmental cues the moment they are brought in culture. Interestingly, loss of the CNS-specific signature is at least partly reversible by engraftment of isolated microglia into a CNS parenchyma that lacks microglia. Fascinatingly, also bone marrow-derived macrophages acquire microglia-like features when they are confronted with a CNS parenchyma. Together these findings demonstrate that signals from the CNS microenvironment are required to sustain microglial specification. High-throughput technologies for studying the transcriptome, epigenome and proteome have been used to study the effect of the local environment on microglial responses confirming that loss or disruption of homeostatic CNS microenvironment signals affects the microglia phenotype.

33Many of these inhibitory signals are provided by cell-cell interactions between microglia and neurons, astrocytes and the oligodendrocyte/myelin complex. CD200 e.g., is expressed by astrocytes and neurons and interacts with the CD200 receptor that is expressed by microglia. Ligand binding suppresses LPS-induced cytokine and IFN production by inhibition of extracellular signal-regulated kinase, mitogen-activated protein kinase p. 38 and c-jun N-terminal kinase. If this constitutive inhibitory signal is lost, microglia acquire an activated phenotype. Another example is CD172a (also known as SIRPα), an inhibitory receptor expressed by microglia. Its ligand CD47 is abundantly expressed in the brain, and CD172a-CD47 interactions provide constitutive suppressive signals to microglia. Finally, the triggering receptor on myeloid cells 2 (TREM2) that is expressed on microglia plays a role in inhibition of Toll-like receptor (TLR) mediated signaling. Loss or disruption of TREM2 expression results in increased secretion of pro-inflammatory cytokines in response to various TLR ligands. Interestingly, loss-of-function mutations of TREM2 cause the rare genetic disorder Nasu-Hakola disease (NHD). NHD is a leukodystrophy characterized by progressive presenile dementia. Pathological examination of the CNS of NHD patients shows atrophy of the cerebral white matter and sclerotic lesions that contain activated microglia.

34It is a major challenge to provide dissociated microglia with the mix of environmental cues that they receive in vivo. Culturing microglia in the presence of astrocyte-conditioned medium or on an astrocyte feeder layer supports microglia survival and the development of a ramified phenotype. An alternative approach is to use 3D cell culture models in which different CNS cells are included. For example, brain organoids can be used as a supportive environment. Abud and colleagues elegantly demonstrated that iPSC-derived microglia naturally integrate into the 3D structure of brain organoids. After integration in the organoid, microglia were able to mature, ramify, and to respond to injury similar to in vivo microglia. In addition, organotypic brain slice cultures from mouse or rat could be used to investigate the cellular and molecular processes of microglia in vitro. A final, more reductionistic approach would be to supply a 3D matrix with CNS microenvironmental cues or to coat the plastic or glass substrate with a surface that mimics the natural physicochemical properties that enhance microglial homeostasis.

35The neuroscience field is increasingly appreciating that modulation of neuroinflammation is a promising strategy to beneficially affect the disease course of neurodegenerative diseases. Microglia are key players in neuroinflammatory responses, leading to an intensive research interest in this particular cell type. Many different microglia in vitro models have been established with respective advantages and limitations as summarized above. For high throughput screening assays, the microglial cell lines that are available might suffice, using culture conditions with or without serum. In addition, it is of importance that doubts on the species origin of human cell lines will be unequivocally confirmed and, if appropriate, accepted and corrected for. For research questions where it is important that the biology of M0 microglia from adult or elderly humans is reflected in the in vitro system, there is a paucity of good models. ESC and iPSC-derived models, in combination with exposure to CNS microenvironmental cues, form a good basis to establish models for M0 microglia. Such microglia will, however, not reflect the effects of aging. To model the latter, direct reprogramming of somatic cells might offer future possibilities. Alternatively, protocols have been established to isolate primary microglia from post-mortem CNS tissue of different species. Human primary microglia have been exposed to an aging (and/or diseased) CNS microenvironment and are not hampered by interspecies translation of the results. Primary microglia of other species are more readily available and can be obtained from post-mortem CNS tissue under more controlled conditions and with different transgenic backgrounds. The main challenge for primary microglia models is, however, to generate M0 microglia. Exposure of primary adult microglia to CNS-specific cues, combined with optimized cell culture protocols that avoid the use of serum, provide a good platform to in vitro model M0 microglia from adult or elderly humans.

36The field is rapidly moving and we are gaining an ever better understanding of microglia biology. This will not only allow the development of better cell culture models, but will also enable the development of activation protocols that are more relevant for neurodegenerative disorders. The use of such refined approaches, will facilitate the development of innovative therapeutic approaches.

37RT, SB, and JB designed, researched, and wrote the paper together.

38The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.