Identification of Microglia Stem Cell Markers Through Single-cell RNA Sequencing Analysis in the Postnatal Brain

Parastoo Amlashi¹, Victoria Neckles¹, Dr. David Feliciano¹

¹ Department of Biological Sciences; Clemson University, Clemson SC


Microglia are immune cells that perform important functions in the central nervous system, such as facilitating the formation of synaptic connections, responses in the innate and adaptive immune system, and maintaining brain homeostasis. During embryonic development, microglial progenitor cells migrate to the brain from the peripheral nervous system and undergo changes that ultimately affect their gene expression and morphology. The maturation of round, immature microglia cells plays a role in neurodevelopment as they become stellate, adult microglia cells. Mature microglia have phenotypes that can be characterized into four main states; such as neuroinflammatory responses (M1), repair and regeneration (M2) with M2 having two distinct subtypes (M2a, M2b) that have varying function in the CNS. Immunohistochemical analysis demonstrates that a heterogeneous, immature microglia population may act as microglia stem cells during neonatal development. Microglia have unique transcriptomes based on region and age as shown in single-cell RNA sequencing. Using publicly available information from single cell RNA sequencing microglia databases, we sought to identify additional markers of this diverse stem cell population. Differential gene expression analysis was performed using a bioinformatics approach to identify genes expressed in two target populations, IBA1low/CD11Bhigh and IBA1high/CD11Blow. We identified a network of co-expressed genes that may allow this population to function as a microglia progenitor in the postnatal brain.



Although microglia cells have been characterized for their innate and adaptive functions in brain development and CNS maintenance, little is known about the specific region diversity as the brain ages. Studies on microglia on a single-cell level have been carried out to look at the heterogeneity across different regions of the brain during development. Single-cell RNA sequencing is a useful genomic technique that detects spatiotemporal heterogeneity and assesses gene expression in a sample of cells in tissue. This technique has revealed that microglia have unique transcriptomes in early postnatal timepoints with rising regional heterogeneity and decreasing cellular heterogeneity throughout the lifespan. Understanding microglial heterogeneity through single cell RNA sequencing will give a comprehensive view of microglia and their relationship to other immune cells. By identifying markers that are pertinent for microglia function, we will have more insight on how cell populations iare affected by neurodegenerative diseases. Differential gene expression on the RNA-seq data will allow us to seek a population of markers that can be used for in vivo experiments using a transgenic mouse model. The markers that are currently being used are Iba1 (Aif1) and Cd11b (ITGAM). The IBA1 gene is found within round and stellate microglia that may or may not be immunologically active. Meanwhile, the CD11B marker represents microglia that are immunologically active and are therefore considered a marker for stem cells, which are immature.  Cell imaging using fluorescence microscopy demonstrates that microglia cells that are high in CD11B are round cells and stellate cells are rich in IBA1. The aim of this project is to find a subset of microglia markers in early time embryonic development which can be used for further investigation into immune system development in the neonatal brain.

Materials and Methods

Utilized the publicly available microglia genome database and performed differential gene expression analysis using the “DESeq2” package in Rstudio. This package enables quantitative analysis based on the total counts in the RNA-seq data. DESeq2 uses shrinkage estimation for dispersions and fold changes to improve the reliability of estimates. It does this by calculating the geometric mean across all the genes and then the counts for that gene in each sample is divided by that mean. DESeq2 uses the Wald test as a hypothesis test. It establishes the null hypothesis that there is no difference between the two target groups.

Once the results were generated with DESeq2, the list of genes identified based on the log2 fold change of the gene expression analysis were put into GeneMANIA, a program that creates network analyses using a large set of functional association data.


Workflow in Rstudio:

→Download raw data from publicly available database (FASTQ file)

→ create subset “treatments” for groups of interest

→ bind treatments together and assign conditions to treatments

→ apply DESeq2 analysis in code



Screen Shot 2020-08-16 at 7.21.26 PM

Figure 1. Microscopy images of round cells at age P2 and stellate cells at age P21

Figure 2. PCA plot identifies the two components (PC1 and PC2) which capture the most variance in our data.

Figure 2. PCA plot identifies the two components (PC1 and PC2) which capture the most variance in our data.


Figure 3. Heatmap with top varied genes in “round” and “stellate” condition groups

Figure 5. CD11b(Itgam) network analysis showing 50 total related genes with similar function in color

Figure 4. CD11b(Itgam) network analysis with 50 total related genes with similar function in color

Figure 6. Iba1(Aif1) network analyses showing 50 total related genes with similar function in color

Figure 5. Iba1(Aif1) network analyses with the 50 total related genes with similar function in color


  1. We can reject the null hypothesis that all the cells in the sample behave similarly. Based on the bioinformatic analyses and cell sorting we were able to pull out different cell types that express a different set of genes.
  2. By doing the networking analysis, we are demonstrating between the two cell types by showing they have different functions.
  1. Identifying a subset of genes specific to that cell type that demonstrate the cell type we have identified but also could use them as markers instead (a marker such as CD11B as seen in immature cells).


    • Li, Q., Cheng, Z., Zhou, L., Darmanis, S., Neff, N. F., Okamoto, J., … Barres, B. A. (2018). Developmental Heterogeneity of Microglia and Brain Myeloid Cells Revealed by Deep Single-Cell RNA Sequencing. Neuron. doi:10.1016/j.neuron.2018.12.006
    • Turano, A., Lawrence, J. H., & Schwarz, J. M. (2017). Activation of neonatal microglia can be influenced by other neural cells. Neuroscience letters, 657, 32–37.
    • M. I. Love, W. Huber, S. Anders: Moderated estimation offold change and dispersion for RNA-Seq data with DESeq2.bioRxiv (2014). doi:10.1101/002832 [1
    • Ginhoux, F., Greter, M., Leboeuf, M., Nandi, S., See, P., Gokhan, S., … Merad, M. (2010). Fate Mapping Analysis Reveals That Adult Microglia Derive from Primitive Macrophages. Science, 330(6005), 841–845. doi:10.1126/science.1194637
    • Matcovitch-Natan, O., Winter, D. R., Giladi, A., Vargas Aguilar, S., Spinrad, A., Sarrazin, S., … Amit, I. (2016). Microglia development follows a stepwise program to regulate brain homeostasis. Science, 353(6301), aad8670–aad8670. doi:10.1126/science.aad8670
    • Karperien, A., Ahammer, H., & Jelinek, H. F. (2013). Quantitating the subtleties of microglial morphology with fractal analysis. Frontiers in cellular neuroscience, 7, 3.
    • Meeta Mistry, R. (2017, October 13). Gene-level differential expression analysis with DESeq2. Retrieved August 17, 2020, from
    • Masuda, T., Sankowski, R., Staszewski, O., & Prinz, M. (2020). Microglia Heterogeneity in the Single-Cell Era. Cell Reports, 30(5), 1271–1281.