We have primarily been using THP1 (a human monocytic line) and RAW264.7 (a mouse macrophage line) as models since these cells are reasonably well characterized with substantial gene expression profiles already available in the public domain. In addition, they are also being studied by our collaborators at the Laboratory of Systems Biology where synergy can be achieved by integrating information from additional levels (e.g., protein-protein interactions) and time scales (e.g., signaling). Our goal is to assay and understand the regulation of phenotypic diversity and plasticity and infer gene networks using data from both the cell population and single cell levels. Based on prior work on bacteria, yeast and mammalian cell lines, it is clear that even for clonal cell populations, substantial heterogeneity can be pervasive at both basal and stimulated states. We will also extend our studies using primary cells to characterize phenotypes at the cell population level. To achieve our goals, we are developing systematic approaches in several critical areas: 1. Develop and apply physiologically relevant perturbations to assay phenotypic diversity. For network inference purposes, systematic perturbations that can generate expression variations across genes and components in the underlying network are also needed (e.g., systematic genetic alterations). The Alliance for Cellular Signaling (AfCS) has already generated publicly available data on single and multi-ligand Toll-like receptor (TLR) stimulations. Our collaborator, Iain Fraser and his group, have also been applying similar types of perturbations to RAW cells. Our own effort centers on developing a set of physiologically relevant environmental perturbations, including monocyte-to-macrophage differentiation signals, combinations and quantitative titrations of cytokines and chemokines, and exposure to other immune cells;we are currently in the process of testing these strategies. 2. Develop experimental and computational protocols and pipelines for measuring genomic phenotypes before and after perturbations. Our initial focus will be on gene expression, including that of a class of non-coding RNAs (microRNAs). We are using both RNA-seq and microarrays the former can provide more detailed information, including the abundance of non-coding genes, alternative splicing isoforms and rare transcripts;microarrays can be applied to a larger number of conditions/samples because they are less expensive and analysis methods are significantly more mature for microarrays than for RNA-seq. By stimulating RAW cells with LPS, a prototypical TLR activator, we have generated pilot RNA-seq data with deep coverage (85 million reads per sample). Using this data, we have developed computational pipelines for processing and analyzing RNA-seq data and have also examined basal and LPS-stimulated phenotypes at the splice isoform, intergenic, and rare transcript levels. We are also using this data to determine the appropriate sequencing coverage to use. In addition to gene expression, we also plan to measure and develop data processing methods for other types of genomics phenotypes, especially methylation and chromatin states. 3. Develop single-cell phenotyping assays for obtaining gene expression information of a large panel of genes before and after perturbations. We are currently testing combining flow cytometry with single-cell based PCR to assay the expression of hundreds of genes, including transcription factors and microRNAs that are either known to regulate monocyte/macrophage phenotype polarization or are derived based on information from cell population-based experiments (see above). 4. Develop integrative computational data analysis approaches and methods. In addition to computational methods for processing and analyzing individual data types, a key goal is to integrate the data obtained from different perturbation conditions to infer gene-gene interaction network(s). Another goal is to understand how innate immune cells process environmental information. Since in vivo stimulations typically involve combinations of cytokines and foreign molecular patterns, one question we aim to address is whether responses to complex stimuli can be predicted and understood based on responses to simpler constituent stimulus. By using data from AfCS and data generated by us and the Fraser group, we are developing a computational approach to predict genes and proteins that facilitate cross-talk between signaling subnetworks. This approach can also lead to better understandings of how networks evolve to process complex information and general network features for generating phenotypic diversity.
| Matsuyama, Masashi; Martins, Andrew J; Shallom, Shamira et al. (2018) Transcriptional Response of Respiratory Epithelium to Nontuberculous Mycobacteria. Am J Respir Cell Mol Biol 58:241-252 |
| Vodovotz, Yoram; Xia, Ashley; Read, Elizabeth L et al. (2017) Solving Immunology? Trends Immunol 38:116-127 |
| Martins, Andrew J; Narayanan, Manikandan; Prüstel, Thorsten et al. (2017) Environment Tunes Propagation of Cell-to-Cell Variation in the Human Macrophage Gene Network. Cell Syst 4:379-392.e12 |
| Gottschalk, Rachel A; Martins, Andrew J; Angermann, Bastian R et al. (2016) Distinct NF-?B and MAPK Activation Thresholds Uncouple Steady-State Microbe Sensing from Anti-pathogen Inflammatory Responses. Cell Syst 2:378-90 |
| Narayanan, Manikandan; Martins, Andrew J; Tsang, John S (2016) Robust Inference of Cell-to-Cell Expression Variations from Single- and K-Cell Profiling. PLoS Comput Biol 12:e1005016 |
| Crompton, Joseph G; Narayanan, Manikandan; Cuddapah, Suresh et al. (2016) Lineage relationship of CD8(+) T cell subsets is revealed by progressive changes in the epigenetic landscape. Cell Mol Immunol 13:502-13 |
| Zhang, Hongbin; Guan, Meiping; Townsend, Kristy L et al. (2015) MicroRNA-455 regulates brown adipogenesis via a novel HIF1an-AMPK-PGC1? signaling network. EMBO Rep 16:1378-93 |
