The large and unique database and accompanying bioinformatic analysis of the Immunological Genome Project provide novel insight into macrophage populations isolated from various organs of mice. A striking initial revelation was that macrophage populations from different organs are considerably diverse, and it is likely that further profiling in macrophages will expand upon this diversity. Only a very small group of mRNA transcripts were associated with all macrophages but not DCs. Proteins previously predicted to distinguish macrophages from other cell types, such as F4/80, CD68 and CD115 (C-fms/Csf1r), did not emerge as the most powerful markers of macrophages. However, many canonical genes did, including those encoding CD14, the high-affinity FcγR FcγR1 (CD64), the Mer tyrosine kinase involving in efferocytosis MerTK, cathespin D, and a fms/fps protein kinase FERT2 that may strongly impact CD115 signaling (but which has not yet been studied in macrophages). The identification of these genes as selectively macrophage-associated reinforce the key role of macrophages in innate immunity, efferocytosis, and clearance of debris, whereas genes associated with antigen presentation and migration to lymphoid tissues were more associated with DCs (Miller at el., In Press). However, our data do suggest that macrophages may have a greater role in activation of MAIT cells than DCs. Based upon follow-up protein expression analysis of MerTK and FcγRI in macrophages from six different tissues, we propose that analysis of MerTK and FcγRI should serve as a starting point for identifying macrophages in tissues, as staining for these markers appears to identify F4/80hi
macrophages and other macrophages that express somewhat lower levels of F4/8016
in all tissues. We believe staining for MerTK and FcγRI has advantage over, but can also powerfully be used in addition to, traditional staining for F4/80, CD11c, and MHC II. The levels of F4/80 and CD11c are often overlapping between macrophages and DCs in nonlymphoid tissues, but it appears that DCs do not co-express MerTK and FcγRI.
Beyond these cell surface markers closely associated with macrophage identity, we uncover other transcripts associated only with macrophages among hematopoietic cells. In particular, immGen module 161 identified a group of genes (A930039a15Rik, Akr1b10, Blvrb, Camk1, Glul, Myo7a, Nln, Pcyox1, Pla2g15, Pon3, Slc48a) that are co-expressed across all the ImmGen dataset and whose functions with the established broad roles of macrophages, but none of them have previously been considered macrophage markers. Both the genes from this module and their predicted regulators would deserve considerable attention in the future.
The Ontogenet algorithm makes it possible to extend the macrophage-associated genes we identified to regulatory programs that may control them. Induced expression of a single module (#330) in red pulp macrophages over all other macrophages and the predictions generated by the algorithm that this module is regulated by SPI-C support the reliability of the algorithm predicted regulatory programs, as SPI-C is already known to be required selectively for red pulp macrophage development or maintenance15
. Exciting new information also emerged, such as strong association of modules unique to peritoneal macrophages that are predicted to be regulated by GATA6.
Gene transcripts that were highly expressed in multiple macrophage populations but not highly expressed in DCs were associated with predicted transcriptional regulatory programs that strongly differed from those uncovered in DCs (Miller et al., Submitted). The predicted regulatory programs of modules enriched for macrophage-associated genes include several members of the MiT family of transcription factors that has been recognized to be specifically expressed in macrophages3
as well as transcription factors not previously associated with macrophages, such as BACH1 and CREG-1. BACH1 has been little studied in macrophages but has recently been linked to osteoclastogenesis28
and is a regulator of heme oxygenase 129
. CREG1 (cellular repressor of E1A-stimulated genes) is a secreted regulator30,31
associated broadly with differentiation32
and cellular senescence33
that was strongly associated with macrophage-enriched gene modules, though it has never been studied in the context of macrophage biology. The Ontogenet algorithm predicts RXRα as the most prominent key activator of the highly specific and universal macrophage module genes, module # 161. Future analysis of these predictions is expected to be highly fruitful in revealing how macrophage identity and function is controlled.
To date, the Immunological Genome Project has mainly focused on cells recovered from resting, uninfected mice, where macrophages mainly derive from the yolk sac12
. Macrophage polarization in the context of infection and inflammation is a topic of great interest that this study has scarcely been able to address. beyond finding that monocytes recruited to the peritoneum in response to thioglycollate upregulate mRNA transcripts observed in resting tissue macrophages, even though monocytes are not precursors for resting tissue macrophages as they are for inflammatory macrophages. The foundations laid herein suggest that future additions to the ImmGen database of macrophages recovered during disease states will add enormously to our understanding of how to manipulate these crucial cells to favor desired outcomes in disease. Based on the great diversity of macrophages in different organs, which we anticipate will hold up even in inflamed organs, such studies may be expected to ultimately generate therapeutic approaches to selectively target macrophages in diseased organs without affecting others cell types.