Globoid cells are a defining pathological feature of Krabbe disease, a leukodystrophy currently lacking an effective long-term therapy. We have developed a cell culture model to study the innate biology and pathogenic potential of activated microglia and their transformation into globoid cells.
The precise function of multi-nucleated microglia, called globoid cells, that are uniquely abundant in the central nervous system of globoid cell leukodystrophy (GLD) is unclear. This gap in knowledge has been hindered by the lack of an appropriate in vitro model for study. Herein, we describe a primary murine glial culture system in which treatment with psychosine results in multinucleation of microglia resembling the characteristic globoid cells found in GLD. Using this novel system, we defined the conditions and modes of analysis for study of globoid cells. The potential use of this model system was validated in our previous study, which identified a potential role for matrix metalloproteinase (MMP)-3 in GLD. This novel in vitro system may be a useful model in which to study the formation and function, but also the potential therapeutic manipulation, of these unique cells.
Globoid cell leukodystrophy (GLD), also known as Krabbe disease, is a fatal demyelinating disease resulting from loss of function mutations in the galatocerebrosidase (galc) gene1. The most prevalent form of GLD is the infantile variant which is typified by onset in early childhood and characterized by an aggressive clinical course of motor and cognitive decline leading to premature death often before five years of age2,3. Genetic testing is used to verify a diagnosis of GLD4. Neuropathology of GLD reveals widespread demyelination, neuronal atrophy, astrogliosis and presence of engorged multi-nucleated microglia called globoid cells5-7. The identification of globoid cells, often containing tubulous inclusions in their cytoplasm, has been a defining feature of GLD for the past 97 years, although the specific function of these conspicuous cells has remained elusive.
The involvement of non-myelinating glia (microglia and astrocytes) in the pathogenesis of GLD has long been considered a secondary response to the profound demyelination in this disease8. Interestingly, the first description of this disease, made by Knud Krabbe in 19165, reported formation of multinucleated phagocytes containing lipid debris that have been named 'globoid cells' and are a defining characteristic of this disease.
Globoid cells are the hallmark feature of GLD pathology, although their role in GLD has long been ignored. Interestingly, these cells are among the earliest characteristic changes in CNS tissue of GLD. This lack of knowledge may have been due to the assumption that the formation of multinucleated phagocytes, called giant cells in other diseases, are typically considered as a consequence of pathology rather than an initial pathogenic driving force9. Therefore, there have been few studies investigating the mechanism by which globoid cells are formed from phagocytes, particularly in the CNS of GLD. The procedure described in this report focuses on the importance of globoid cell formation in the CNS and our previous demonstration that psychosine-induced multinucleation of microglia in vitro and these cells exhibited higher levels of phagocytic activity. Consistent with these observations, globoid cells in twitcher brains frequently contain PAS-positive debris, suggesting high levels of phagocytic activity. Globoid cells are also found to be immunopositive for ferritin (a microglia marker)10, KP-1/CD68 (a monocyte marker), and some are also positive for vimentin (an intermediate filament protein and marker of astrocytes and activated microglia)11, HLA-DRa (an MHCII surface receptor), and TNF-α7, and Iba-1 (a calcium binding protein used to identify microglia)12. Based on this collection of markers, globoid cells originate from microglia that develop a unique phenotype.
Despite their uniqueness, the specific function and contribution of GCs to GLD pathogenesis has been largely overlooked. Globoid cells have been thought to be a secondary consequence of chronic demyelination. However, past studies examining the temporal association of globoid cells to the white matter pathology of GLD have identified the presence of globoid cells in the late embryonic to early postnatal periods; times preceding oligodendrocyte apoptosis and overt demyelination13. Thus, the temporal sequence of development of the neuropathology in GLD suggests that globoid cells are formed in advance of demyelination in this disease14. This led to our hypothesis that the early formation of globoid cells in GLD may represent a defining pathogenic event rather than a secondary, reactive response to oligodendrocyte damage15. Additionally, dysregulation of microglial activity in GLD has been considered a factor limiting the long-term efficacy of hematopeotic stem cell therapies for treating this disease16. Thus, investigating the cellular functions and regulation of microglia, and globoid cells, in response to psychosine is expected to provide new insights in the pathogenesis of GLD.
Until recently, the lack of an appropriate model in which to study globoid cell formation had limited the understanding of the precise function and contribution of these cells to the pathology of GLD. In recent studies, it was determined that globoid-like cells can be formed in direct response to psychosine, a pathogenic lipid toxin that accumulates in GLD. We found that microglia, but not macrophages, are activated and transformed into globoid cells in primary glial cultures in response to psychosine15. This transformation into globoid cells was found to be mediated by the extracellular protease, matrix metalloproteinase (MMP)-315. More recently, we have extended these findings and determined that psychosine-activated microglia and globoid cells developed in this in vitro model system are potently toxic to oligodendrocytes and oligodendrocyte progenitor cells. Hence, when considered in the context of GLD, the early accumulation of psychosine and formation of globoid cells prior to demyelination would support an emerging primary and possibly pathogenic role for microglia in this disease.
We propose that study of globoid cell formation will reveal new information about the pathogenesis of GLD that will contribute to our understanding of this disease. Moreover, this new cellular model of GLD may provide a new format from which novel therapeutic approaches to address pathological changes in this disease could be tested. Hence, in this report we provide a detailed protocol for the in vitro development of psychosine-induced globoid cells from primary cultures of non-myelinating glia.
All procedures involving animals were performed in accordance with the Policy on Humane Care and Use of Laboratory Animals set forth by the Office of Laboratory Animal Welfare (NIH) and only with approval from the Institutional Animal Care and Use Committee (IACUC) of the University of Connecticut Health Center.
1. Preparation of Mixed Glial Cultures
2. Globoid Cell Induction in Mixed Glial Cultures
3. Immunocytochemistry (ICC) to Visualize Globoid Cells
4. Analysis and Characterization of Globoid Cells in Primary Culture
5. Induction of Globoid Cells from Purified Microglial Cultures
NOTE: An alternate approach to decipher intercellular signaling between the astrocyte and microglia is another advantage of this in vitro globoid cell model. Purification of primary microglial cells for replating or co-culturing can be achieved by their lower adherence property in culture, as described in the following steps (see Section 5.2).
This protocol, as written, is expected to take approximately 36 days to complete from start to finish (See Figure 1: Experimental Workflow Scheme). It has been our experience that the development of 'globoid-like' cells in this primary culture system is both reliable and reproducible: the formation of multinucleated cells in response to psychosine is consistently observed with 7 days of treatment.
Immunocytochemical staining of microglia using Iba-1 in conjunction with a nuclear counter-stain will enable identification of large, rounded multinucleated cells (Figure 1b). In some instances nuclear content in these globoid-like cells will appear with distinct nuclei. However, in many instances, the gross enlargement of the size of the nucleus reflects the multinucleated status of these cells (Figure 1b). The number of globoid-like cells can vary from visual field to visual field, but it is worth noting that the proportion of globoid cells formed in vitro represents <5-10% of the total microglial cell population. It is also important to point out that psychosine treatment also evokes a generalized activation of microglia (i.e. measurable increase in phagocytic activity) that occurs in response to psychosine treatment among mononucleated microglia and the multinucleated globoid cells (GCs) in these cultures. As shown in Figure 1c, phagocytically active profiles of Iba-1+ microglia can be readily identified when put in co-culture with oligodendrocyte progenitor cells (OPCs).
This cell culture system is amenable to experimental manipulation as a means to assess the biology of globoid cells. For example, we have recently demonstrated a novel role for the extracellular protease, matrix metalloproteinase-3 (MMP-3/stromelysin-1) as an important mediator of psychosine-induced GC formation using this culture assay15.
Figure 1: Experimental Workflow for Primary Culture in vitro Model of Globoid Cell Formation. (A) Flowchart depicting the primary steps and intervening times (in days) for development of globoid cells in culture. Each step is labeled with its associated text section from the protocol. This workflow provides both a logical progression from mixed (astrocyte and microglial cultures; Section 1) and an alternate starting culture of purified microglia (Section 5). (B) Representative immuncytochemical staining of a multinucleated microglial cell following 7 consecutive days of psychosine (10 µM) treatment as identified by the Iba1 (green) and DAPI (blue). (C) Example of presumptive phagocytic microglial cell (green) following psychosine treatment in co-culture with Olig2+ (red) OPC. Scale bar (in panel B) = 90 µm for 'B' and = 150 µm for 'C'. Please click here to view a larger version of Figure 1B, and here to view a larger version of Figure 1C.
The protocol described herein provides a new model system in which to study the development and functional characterization of activated microglia and globoid cells. Prior work by Im et al. using a HEK293 cell line provided a template for the development of the present protocol for the study of globoid cell formation21. It is also important to point out that the globoid cells derived in the model differ from the native globoid cells identifiable in GLD. For instance, we have routinely observed quadranucleation of microglia in our murine cultures, while globoid cells in GLD have been frequently observed containing in excess of 10 nuclei. Secondly, the globoid cells formed in the in vitro system are also small in diameter as compared to those observed in vivo. There are likely to be several reasons for these phenotypic differences which may include the relative simplicity of the culture system used. For instance, there are no neurons or oligodendrocytes present during the induction phase of globoid cell formation using this primary glial culture procedure. Also, the timing of the seven day protocol yields a significant density of globoid cells; however, longer treatment times may further enhance the similarity of murine globoid cells formed by this protocol with the native features of globoid cells observed in GLD.
An important feature of this model is that it uses primary mixed glial cultures, which include both astrocytes and microglia15,19. This is an important advantage of this globoid cell model for its utility in assessing the contributions and interplay between astrocytes and microglia. Both of these non-myelinating cell types are activated in GLD, though their roles in this disease are, as yet, poorly characterized. Importantly, we had previously determined that astrocytes in GLD express higher levels of MMP-3 expression and that this presumably astrocyte-derived factor was critical for the transformation of microglia in response to psychosine. Future applications of this in vitro model of globoid cell formation could employ chimeric cultures of astrocytes and microglia of differing genetic backgrounds, which could be employed to advance our understanding of the cell-specific contributions to the pathogenic effects of psychosine on microglia.
In summary, this protocol provides a new approach to study the pathogenesis of GLD. The role of the globoid cells has been enigmatic and hypotheses on their protective or deleterious functions in GLD have been proposed. The early identification of globoid cells in the natural history of GLD would further support our view at this time that activation of microglia in GLD are a primary pathogenic response and a likely mediator of myelin pathology in this disease. Adaptation of this in vitro model system is expected to further our understanding on the cellular etiology of GLD.
The authors have nothing to disclose.
This work was supported in part by grants from the National Multiple Sclerosis Society (RG 5001-A-3 to S.J.C.), the National Institutes of Health (NS065808 to E.R.B.; NS078392 to S.J.C.), start-up funds from the UConn Health Center (to SJC) and the Kim Family Fund (UCHC in support of K.I.C.).
Hank’s balanced salt solution (HBSS) containing no cations (Mg2+ and Ca2+). | Life technologies | 14175-095 |
Neural Tissue Dissociation Kit | Miltenyi | 130-092-628 |
40 uM cell-strainer | Fisherbrand | 22363547 |
Hank’s balanced salt solution (HBSS) containing cations (Mg2+ and Ca2+). | Gibco | 14025-092 |
Dulbecco's modified eagle medium (DMEM) | Gibco | 11995-065 |
fetal bovine serum (FBS) | Atlanta Biologicals | S11150 |
Penicilin/Streptomycin | Life technologies | 15070-063 |
Laminin | Sigma | L2020 |
Trypsin-EDTA solution | Life technologies | 25299-056 |
Psychosine | Sigma | P9256 |
Dimethyl sulfoxide (DMSO) | Sigma | D2650 |
Paraformaldehyde (PFA) | Electron Microscopy Science | 19208 |
Normal Goat Serum (NGS) | Invitrogen | PCN5000 |
Iba-1 | WAKO | 019-19741 |
Alexa Fluor conjugated antisera | Life Technologies | Various |
Mounting Media | Southern Biotech | OB100-01 |
Phagocytic Assay Kit | Cayman Chemicals | 500290 |
HEPES | Sigma | BP310-500 |