A novel in vitro human microglia model: Characterization of human monocyte-derived microglia

Abstracts

Microglia are the innate immune cells of the central nervous system. They help maintaining physiological homeostasis and contribute significantly to inflammatory responses in the course of infection, injury and degenerative processes. To date, there is no standardized simple model available to investigate the biology of human microglia. The aim of this study was to establish a new human microglia model. For that purpose, human peripheral blood monocytes were cultured in serum free medium in the presence of M-CSF, GM-CSF, NGF and CCL2 to generate monocyte-derived microglia (M-MG). M-MG were clearly different in morphology, phenotype and function from freshly isolated monocytes, cultured monocytes in the absence of the cytokines and monocyte-derived dendritic cells (M-DC) cultured in the presence of GM-CSF and IL-4. M-MG acquired a ramified morphology with primary and secondary processes. M-MG displayed a comparable phenotype to the human microglia cell line HMC3, expressing very low levels of CD45, CD14 and HLA-DR, CD11b and CD11c; and undetectable levels of CD40, CD80 and CD83, and a distinct pattern of chemokine receptors (positive for CCR1, CCR2, CCR4, CCR5, CXCR1, CXCR3, CX3CR1; negative for CCR6 and CCR7). In comparison with M-DC, M-MG displayed lower T-lymphocyte stimulatory capacity, as well as lower phagocytosis activity. The described protocol for the generation of human monocyte-derived microglia is feasible, well standardized and reliable, as it uses well defined culture medium and recombinant cytokines, but no serum or conditioned medium. This protocol will certainly be very helpful for future studies investigating the biology and pathology of human microglia.

Introduction

Microglia are resident innate immune cells of the central nervous system (CNS) and play a crucial role in maintaining the healthy physiological homeostasis. Microglia also contribute substantially to inflammation in response to injury, toxins, pathogens and degenerative processes (Streit, 2001, Streit, 1996, Howell et al., 2010, Lee et al., 2010, Politis et al., 2011). Microglia are characterized by a quick response to various stimuli, resulting in activation with rapid changes in morphology, phenotype and function. Morphological changes include shortening of cell processes and cellular hypertrophy. To date, there is no specific microglia marker clearly separating this cell type from other cells of the monocyte/macrophage lineage. Some studies suggested that primary microglia could be distinguished from other tissue macrophages according to expression levels of markers such as CD11b and CD45 (Aloisi et al., 2000, Ford et al., 1995). Under healthy condition, resting microglia with typical ramified morphology show low expression levels of these common myeloid lineage markers (Nimmerjahn et al., 2005). However, microglia have been shown to share expression of surface markers common to other immune cells of the macrophage family such as CD45, CD14, MHC-class II, CD68, immunoglobulin Fc receptors and β2 integrins (Lambertsen et al., 2009, Wirenfeldt et al., 2005). Expression levels of these surface markers may depend on the inflammatory status of the CNS and on microglia activation. Similarly, functional inflammatory changes are dominated by secretion of pro-inflammatory cytokines, chemokines, neurotrophic factors and up-regulation of corresponding receptors, as well as production of nitric oxide and reactive oxygen intermediates (Aloisi, 2001, Tambuyzer et al., 2009).

Microglia have recently been shown in a mouse model to originate from a myeloid yolk sac population during embryogenesis (Ginhoux et al., 2010). In humans, the microglia origin is still not known. Furthermore, little is known about the turnover rate of microglia in the healthy CNS. It has been postulated that proliferation of the resident microglia population is rather low and there may be immigration of bone marrow derived precursor cells into the CNS. Invasion of bone-marrow derived microglia have been shown to migrate into the CNS using chimeric irradiated mice (Simard et al., 2006, Zhang et al., 2007). However, the proportion of bone-marrow progenitor cells in replenishment of microglia is controversial.

A variety of cytokines have been shown to contribute to microglia development and differentiation. Colony stimulating factor-1/M-CSF and its receptor CSF-1R play an important role in the development of macrophage populations (Chitu and Stanley, 2006). Recently, Ginhoux et al. (2010) reported the constant absence of microglia throughout life in CSF-1R deficient mice, indicating that M-CSF is essential for microglia development and differentiation. In addition, granulocyte-macrophage colony stimulating factor (GM-CSF) also contributes to microglia development and differentiation (Aloisi et al., 2000, Esen and Kielian, 2007). Both, M-CSF and GM-CSF have crucial effects on proliferation and survival of primary human fetal and adult microglia in culture with GM-CSF having a greater impact on proliferation (Esen and Kielian, 2007, Lee et al., 1994). M-CSF plays a crucial role in the final maturation stage of microglia. In op/op mice, a model of human osteopetrosis with a functional mutation the M-CSF gene, the number of microglia are decreased and the cells are smaller and have impaired activation ability in response to injury (Kalla et al., 2001, Węgiel et al., 1998). Importantly, GM-CSF or interleukin-3 does not substitute for M-CSF (Blevins and Fedoroff, 1995).

Nerve growth factor (NGF), a member of neurotrophins, has also been shown to act on the proliferation and survival of microglia (Zhang et al., 2003). NGF binds and acts through the P75 receptor, which has been shown to be expressed in microglia in multiple sclerosis lesions (Valdo et al., 2002). It has also been shown that microglia express the nerve growth factor receptor TrkA (Tonchev, 2011). NGF induces migration of the microglia cell through activation of the TrkA (TrkA activation) (De Simone et al., 2007). NGF is expressed by activated microglia, astrocytes and hippocampal neurons (Friedman, 2000, Saez et al., 2006, Tonchev, 2011). However, inflammation certainly induces neurotrophin secretion by human microglia (Heese et al., 1998, Nakajima et al., 2001).

Chemokines have also been shown to play an important role in microglia biology. Monocyte chemoattractant protein-1 (MCP-1), or CCL2, is one of the prominent chemokines in the regulation of the microglia migration to the site of inflammation in experimental models (Leonard et al., 1991, Zhang et al., 2007). CCL2 acts through its specific receptor CCR2 which is expressed by microglia and other cells of monocytic lineage, such as macrophages and dendritic cells (Rebenko-Moll et al., 2006). CCL2 acts also through CCR4, but little information is available about CCR4 expression in microglia (Craig and Loberg, 2006, Zhang et al., 2006). Studies on normal rat CNS showed that CCL2 is constitutively expressed by astrocytes and neurons and it can be found in various brain regions, including the cerebral cortex, the hippocampus and the hypothalamus (Banisadr et al., 2005). CCL2 is also secreted by astrocytes and neurons under inflammatory condition (Banisadr et al., 2005, Farina et al., 2007, Tanuma et al., 2006). It has been shown that CCL2 is secreted by both undamaged and damaged spinal sensory neurons in rat, resulting in activation of spinal microglia and initiating neuropathic-like pain (Thacker et al., 2009). Up-regulation of CCL2 expression during Alzheimer's disease and multiple sclerosis correlates with activation of microglia and may contribute to pathogenesis of neurodegenerative diseases (Conductier et al., 2010, Simpson et al., 2000a). In that respect, CCL2 contributes to recruitment of mononuclear cells into the inflamed CNS, followed by activation of the microglia. Zhang et al. (2007) showed that CCL2 can induce spinal microglia activation in mice. They also reported in chimeric mice that CCL2 recruits bone marrow-derived macrophages, which proliferate and differentiate into microglia in the spinal cord and induce inflammation and microgliosis after nerve injury. In experimental autoimmune encephalomyelitis (EAE), a murine model of multiple sclerosis, CCR2 deficient mice do not develop the mononuclear cell infiltration and inflammation. Indeed, no chemokines increase was detected in the CNS in that model (Izikson et al., 2000). On the other hand, CCR2 deficiency in a mouse model of Alzheimer's disease impairs significantly microglia accumulation at sites of plaque formation, resulting in a decrease of β-amyloid (Aβ) clearance and in accelerating early disease progression (El Khoury et al., 2007).

In vitro culture of mouse and human microglia has often been used as a model by various researchers. Thereby, astrocyte-conditioned medium (ACM) has been used to keep microglia in culture with a ramified resting morphology. ACM contains M-CSF, GM-CSF and transforming growth factor β (TGF β), all known to be secreted by astrocytes. Microglia lose their ramified morphology in the presence of antibodies against the mentioned cytokines, indicating that M-CSF and GM-CSF are essential for microglia culturing (Schilling et al., 2001). GM-CSF, even at low concentration in serum free medium, keeps microglia at a resting state with morphological ramifications (Fujita et al., 1996). A recent study reported the proliferative effect of CCL2 on neonatal and embryonic primary rat microglia in culture without inducing additional changes in morphology, phenotype and cytokine expression (Hinze and Stolzing, 2011).

Most interestingly, murine bone marrow-derived precursor cells can be differentiated towards microglia-like cells when cultured in the presence of astrocytes or in mixed glial cultures. Thereby, M-CSF alone is not sufficient for microglia differentiation, which indicates the importance of additional astrocyte-derived factors (Noto et al., 2010). Differentiation of murine bone marrow stem cells toward microglia-like cells has recently also been successful, using ACM supplemented with GM-CSF (Hinze and Stolzing, 2011). Culturing fetal and adult human microglia in the presence of GM-CSF induces functional maturation towards an antigen presenting cell type, especially in adult microglia. However, GM-CSF does not induce maturation of microglia to acquire the complete phenotype and function of mature dendritic cells (Lambert et al., 2008, Re et al., 2002).

In vitro studies on microglia used mainly primary microglia culture from embryonic or neonatal murine brains (Bassett et al., 2012, Hinojosa et al., 2011). In human, brain-derived microglia is difficult to obtain for ethical reasons. In addition, only low numbers of cells are collected. Also, fetal microglia seems to be quite different from adult microglia. Only few human microglia cell lines have been generated, including HMO6 cells (Nagai et al., 2005) and HMC3 cells (Janabi et al., 1995). These cell lines cannot be considered as an optimal model for microglia cells due the significant modification in morphology and function as result of genetic manipulation and long-term culture. Therefore, a more convenient and more appropriate model for in vitro microglia human studies is still missing.

Most importantly, Leone et al. (2006) showed that human blood-derived monocyte, cultured with astrocytes-conditioned medium (ACM), acquired the ramified morphology of microglia and expressed substance P, calcium binding protein Iba1 and dimly MHCII, three typical characteristics of microglia. There are also reports showing successful differentiation of rat blood monocytes towards microglia using ACM (Schmidtmayer et al., 1994, Sievers et al., 1994). Recently, Hinze and Stolzing (2011) showed the differentiation of murine bone marrow stem cells towards microglia-like cells in the presence of ACM with similar phenotypic and functional properties to primary brain-derived microglia cultures. In the light of these data, we developed a new in vitro human microglia model using human blood peripheral mononuclear cells, a serum-free culture condition and a panel of factors, including M-CSF, GM-CSF, NGF and CCL2. This new human microglia model has then been assessed for morphology, phenotype, and function.