Destroying Existing Microglia is Necessary for Replacement Strategies to Work

Today's open access research is a demonstration in mice of approaches to replace near all microglia in the central nervous system. Microglia are innate immune cells of the brain, involved not just in destroying pathogens and errant cells, but also in ensuring the correct function of neural connections. With the progression of aging, their behavior shifts to become more harmful and inflammatory, and their numbers include ever more senescent cells. Senescent cells generate tissue dysfunction and chronic inflammation via the senescence-associated secretory phenotype, but beyond that microglia tend to adopt a more aggressive and inflammatory set of behaviors even when not senescent. This detrimental change is the consequence of some mix of persistent infection, protein aggregates, and other forms of the underlying molecular damage that drives aging.

Microglial dysfunction contributes meaningfully to age-related neurodegeneration, as illustrated by the benefits produced in animal models by the selective destruction of senescent microglia. That approach has turned back the tau pathology characteristic of Alzheimer's disease in mice, for example. There is also evidence for inflammatory microglia to be involved in the progression of Parkinson's disease.

More than just the senescent cells need to be replaced, or otherwise have their behavior changed for the better, however. Approaches involving clearance of a large fraction of microglia, and allowing them to regenerate thereafter, have seemed viable. Efforts to replace microglia with transplanted cells have proven challenging, however: even hematopoietic stem cell transplantation, such as via a bone marrow transplant, doesn't replace more than a small fraction of the existing microglia. As researchers here demonstrate, it is necessary to first destroy near all microglia in order to leave an empty niche in the brain that will generate signals telling the body to replace these cells. Will replacement be necessary for the treatment of age-related microglial dysfunction, rather than genetic dysfunction? It seems plausible that hematopoietic stem cell replacement will be adopted as an approach to immune system rejuvenation, so why not pair it with clearance of cell populations that should be replaced?

Efficient Strategies for Microglia Replacement in the Central Nervous System

Microglia are important immune cells in the central nervous system (CNS). Dysfunctions of gene-deficient microglia contribute to the development and progression of multiple CNS diseases. Microglia replacement by nonself cells has been proposed to treat microglia-associated disorders. However, some attempts have failed due to low replacement efficiency, such as with the traditional bone marrow transplantation approach.

Engrafted cells in previous transplantation approaches do not extensively proliferate in the recipient brain, which explains the low efficiency of transplantation. Indeed, the proliferation-dependent turnover rate of microglia is rather slow in homeostatic conditions. In contrast, we have demonstrated that residual microglia exhibit an astonishing proliferation capacity after pharmacological depletion (~99%). This potentially suggests that microglial proliferation relies on an empty microglial niche. We therefore reasoned that the microglia-free niche is a vital prerequisite for successful engraftment of nonself microglia (or microglia-like cells). Colony-stimulating factor 1 receptor (CSF1R) is essential for microglia survival. PLX5622 is a CSF1R inhibitor with improved specificity compared to its analog, PLX3397. To create the microglia-free niche, we utilized PLX5622 to inhibit CSF1R.

We then developed highly efficient approaches for nonself microglia replacement that are effective in the adult normal mouse at the CNS-wide scale. First, microglia replacement by bone marrow transplantation (mrBMT) is capable of inducing allografted bone marrow cells (BMCs) to differentiate into microglia-like cells in the entire CNS, replacing 92.66% of resident microglia in the brain, 99.46% in the retina, and 92.61% in the spinal cord, respectively. Second, microglia replacement by peripheral blood (mrPB) is able to induce peripheral blood cells (PBCs) to microglia-like cells and replace 80.74% of resident microglia in the brain and 74.01% in the retina. Third, to precisely manipulate microglia in a specified brain region without affecting the rest of the brain, we further developed microglia replacement by microglia transplantation (mrMT). The engrafted microglia via mrMT resemble the natural characteristics of naive microglia.

When determining superiority of a strategy, replacement efficiency and source availability are the two most important dimensions to take into consideration. Among the three microglia replacement approaches, mrBMT achieves the highest replacement efficiency - 92.66% in the brain, 99.46% in the retina, and 92.61% in the spinal cord. However, mrBMT uses the BMC as the donor cell, which is clinically hard to acquire due to the invasive nature of the procedure and the aversive response from the donor. Such scarce availability of the source is likely to restrict its potential of becoming a widely used standard clinical method for microglial replacement. On the other hand, mrPB greatly broadens the donor source by using PBC, the largest donor cell pool, while maintaining high replacement efficiency CNS wide, just slightly inferior to mrBMT. Abundant availability of donor cells and the relatively high efficiency of cell replacement make mrPB an ideal approach to manipulate microglia at the whole-CNS scale.


How dangerous could such almost complete cleaning of microglia be? As much as an almost complete T-cell cleaning? Or even worse, since they are needed for synapsis function?

Posted by: Antonio at August 25th, 2020 1:50 AM

To Antonio,

Thank you for your concerns. However, different from the T cell depletion, microglia depletion is a relatively safe approach. In animal models, there are no obvious aversive effects after microglia depletion by pharmacological ablation via CSF1R inhibition [1-3], even for a long-term depletion. In fact, the safety of CSF1R inhibitors, including PLX3397 and PLX5622, are FDA approved by clinical trials [4-7]. Even without mrBMT, mrPB or mrMT treatment, when you remove the CSF1R inhibition, microglia can repopulation to the same density and similar phenotype through rapid self-proliferation of surviving cells (~1% in the brain)[2]. This might be against your gut feeling; however, it is safe to deplete microglia by CSF1R inhibition.


1. Huang, Y., et al., Dual extra-retinal origins of microglia in the model of retinal microglia repopulation. Cell Discov, 2018. 4(1): p. 9.
2. Huang, Y., et al., Repopulated microglia are solely derived from the proliferation of residual microglia after acute depletion. Nat Neurosci, 2018. 21(4): p. 530-540.
3. Elmore, M.R., et al., Colony-stimulating factor 1 receptor signaling is necessary for microglia viability, unmasking a microglia progenitor cell in the adult brain. Neuron, 2014. 82(2): p. 380-97.
4. Butowski, N.A., et al., A phase 2 study of orally administered PLX3397 in patients with recurrent glioblastoma. Journal of Clinical Oncology, 2014. 32(15_suppl): p. 2023-2023.
5. Cannarile, M.A., et al., Colony-stimulating factor 1 receptor (CSF1R) inhibitors in cancer therapy. J Immunother Cancer, 2017. 5(1): p. 53.
6. Papadopoulos, K.P., et al., First-in-Human Study of AMG 820, a Monoclonal Anti-Colony-Stimulating Factor 1 Receptor Antibody, in Patients with Advanced Solid Tumors. Clin Cancer Res, 2017. 23(19): p. 5703-5710.
7. Lee, J.H., et al., A phase I study of pexidartinib, a colony-stimulating factor 1 receptor inhibitor, in Asian patients with advanced solid tumors. Invest New Drugs, 2020. 38(1): p. 99-110.

Posted by: Bo Peng at August 25th, 2020 6:32 PM

@Antonio It could be more dangerous. Microglia are one of multiple cell types necessary for the survival of neurons, secreting survival signals such as GDNF (though other cell types do the same thing, and perhaps there is some redundancy?) - while providing at least some energy to astrocytes through lactate (And these astrocytes then provide lactate for neurons)

Worst-case, you'll see some brain damage as a result. If this procedure is done, you may want the microglia to be restored quickly, or for some additional therapeutics (such as injected BDNF) to compensate for the loss. If it could be done in multiple weaker 'chunks', that would probably also control for the problem (e.g. kill off and replace (arbitrary figures ahead) 10% of the cell population 10x instead of 50% of the cell population once). Still, generally, tough to say!

Posted by: Patrick Deane at August 25th, 2020 10:00 PM

Self-correction, I got the direction of lactate travel incorrect, ignore the mention of astrocytes gaining energy from microglia via lactate.

However, I stand by the point on survival signals. That *could* have some off-target effects, unless controlled for via combination therapy. It may not be that extreme however, after all, redundancy from astrocytes... It could still be worth controlling it, regardless.

Posted by: Patrick Deane at August 25th, 2020 10:11 PM

@Bo Peng, @Patrick Deane: Thanks!

Posted by: Antonio at August 26th, 2020 1:21 AM

Dear Dr. Peng: thank you very much for lending your insights to this forum. Your input is very welcome. Congratulations on the new work.

Posted by: Michael at September 1st, 2020 11:56 AM
Comment Submission

Post a comment; thoughtful, considered opinions are valued. New comments can be edited for a few minutes following submission. Comments incorporating ad hominem attacks, advertising, and other forms of inappropriate behavior are likely to be deleted.

Note that there is a comment feed for those who like to keep up with conversations.