Microglia have long been the focus of much attention due to their strong proliferative response (microgliosis) to essentially any kind of damage to the CNS. More recently, we reached the realization that these cells play specific roles in determining progression and outcomes of essentially all CNS disease. Thus, microglia has ceased to be viewed as an accessory to underlying pathologies and has now taken center stage as a therapeutic target. Here, we review how our understanding of microglia’s involvement in promoting or limiting the pathogenesis of diseases such as amyotrophic lateral sclerosis, Alzheimer’s disease, Huntington’s disease, multiple sclerosis, X-linked adrenoleukodystrophy (X-ALD) and lysosomal storage diseases (LSD) has changed over time. While strategies to suppress the deleterious and promote the virtuous functions of microglia will undoubtedly be forthcoming, replacement of these cells has already proven its usefulness in a clinical setting. Over the past few years, we have reached the realization that microglia have a developmental origin that is distinct from that of bone marrow-derived myelomonocytic cells. Nevertheless, microglia can be replaced, in specific situations, by the progeny of hematopoietic stem cells (HSCs), pointing to a strategy to engineer the CNS environment through the transplantation of modified HSCs. Thus, microglia replacement has been successfully exploited to deliver therapeutics to the CNS in human diseases such as X-ALD and LSD. With this outlook in mind, we will discuss the evidence existing so far for microglial involvement in the pathogenesis and the therapy of specific CNS disease.

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by upper and lower motoneuron death. Mutations in the gene for superoxide dismutase 1 (SOD1) cause a familial form of ALS and have been used to develop transgenic mice which overexpress human mutant SOD1 (mSOD) and these mice exhibit a motoneuron disease which is pathologically and phenotypically similar to ALS. Neuroinflammation is a pathological hallmark of many neurodegenerative diseases including ALS and is typified by the activation and proliferation of microglia and the infiltration of T cells into the brain and spinal cord. Although the neuroinflammatory response has been considered a consequence of neuronal dysfunction and death, evidence indicates that manipulation of this response can alter disease progression. Previously viewed as deleterious to neuronal survival, recent reports suggest a trophic role for activated microglia in the mSOD mouse during the early stages of disease that is dependent on instructive signals from infiltrating T cells. However, at advanced stages of disease, activated microglia acquire increased neurotoxic potential, warranting further investigation into factors capable of skewing microglial activation towards a neurotrophic phenotype as a means of therapeutic intervention in ALS.

Amyotrophic lateral sclerosis (ALS) is associated with increased numbers of microglia within the CNS. However, it is unclear to what extent bone marrow (BM)-derived cells contribute to this microgliosis. We have studied the adoptive transfer of green fluorescent protein (GFP)-labeled whole BM cells and BM from mice that express GFP only in CX(3)CR1+ cells (CX(3)CR1(+/GFP)) into the CNS of a murine model of ALS having over-expression of mutant superoxide dismutase (mSOD), and wt littermates. We find that most GFP+ and CX(3)CR1(+/GFP) cells are found adjacent to the microvasculature within the CNS, both in mSOD and wt mice. GFP+ and CX(3)CR1(+/GFP) cells within the CNS have a variety of morphologies, including cells with an elongated appearance, weak Iba-1 immunoreactivity, and often mannose receptor immunoreactivity, indicating that these cells are perivascular microglia. Typically, less than 10% of BM-derived cells had a stellate-shape and expressed strong Iba-1 immunoreactivity, as expected for parenchymal microglia, indicating that BM-derived cells uncommonly generate parenchymal microglia. Adoptive transfer of BM-derived cells from CX(3)CR1(+/GFP) mice revealed that many elongated cells are GFP+, demonstrating that some perivascular cells are derived from BM cells of the CX(3)CR1+ lineage. The significantly greater numbers of BM cells in mSOD than in control mice indicate that the presence of these BM cells in the spinal cord is regulated by conditioning stimuli that may include irradiation and inflammatory factors within the CNS.

Myeloablative preconditioning using irradiation is the most commonly used technique to generate rodents having chimeric bone marrow, employed for the study of bone marrow-derived cell accumulation in the healthy and diseased central nervous system. However, irradiation has been shown to alter the blood-brain barrier, potentially creating confounding artefacts. To better study the potential of bone marrow-derived cells to function as treatment vehicles for neurodegenerative diseases alternative preconditioning regimens must be developed. We treated transgenic mice that over-express human mutant superoxide dismutase 1, a model of amyotrophic lateral sclerosis, with busulfan to determine whether this commonly used chemotherapeutic leads to stable chimerism and promotes the entry of bone marrow-derived cells into spinal cord. Intraperitoneal treatment with busulfan at 60 mg/kg or 80 mg/kg followed by intravenous injection of green fluorescent protein-expressing bone marrow resulted in sustained levels of chimerism (∼80%). Bone marrow-derived cells accumulated in the lumbar spinal cord of diseased mice at advanced stages of pathology at both doses, with limited numbers of bone marrow derived cells observed in the spinal cords of similarly treated, age-matched controls; the majority of bone marrow-derived cells in spinal cord immunolabelled for macrophage antigens. Comparatively, significantly greater numbers of bone marrow-derived cells were observed in lumbar spinal cord following irradiative myeloablation. These results demonstrate bone marrow-derived cell accumulation in diseased spinal cord is possible without irradiative preconditioning.