. Transcriptome-based network analysis reveals a spectrum model of human macrophage activation. Immunity. 2014 Feb 20;40(2):274-88. Epub 2014 Feb 13 PubMed.


Please login to recommend the paper.


Make a Comment

To make a comment you must login or register.

Comments on this content

  1. In the last decade, there has been a fundamental reevaluation of the origins and biology of microglia. Recent work has provided substantial new understanding about the critical roles these cells play in central nervous system disease and injury. In the past, it was recognized that these cells exist in a quiescent or resting state, but upon insults to the brain they acquire an “activated” proinflammatory phenotype. This dichotomy was also used to describe peripheral macrophage activation states. Siamon Gordon was the first to formulate the notion that macrophages exhibit specific phenotypes associated with chronic parasitic infections, various disease states, and the resolution of inflammation (Gordon, 2003). These “alternatively activated” macrophages were termed “M2,” while the classical proinflammatory phentotype was termed “M1.” This nomenclature, based on T-cell phenotypes, has been useful to delineate two distinct polarization states of macrophages. In an influential review, Mosser and Evans postulated that there are different populations of macrophages that possess a spectrum of activities, reflective of the various influences in their environment (Mosser and Edwards, 2008). This paper by Xue et al. validates this view and adds substantial detail to the transcriptional signatures that typify M1/M2 states. They ultimately identify a modular organization of nine “programs,” and the genes that serve to govern the phenotypic conversion to these states. Ultimately, they conclude that a “spectrum model” best describes their data set, a refined and updated version of the thesis advanced by Mosser and Edwards. The Xue et al. study is remarkable for the size and complexity of the data sets that were generated and analyzed.

    This work is directly relevant to the evolving understanding of microglial biology. It only recently has been recognized that microglia are embryologically derived from yolk sac progenitors and thus arise from a different lineage from peripheral monocytes and macrophages. Recent studies by Butovsky et al. (Butovsky et al., 2014)and Hickman et al. (Hickman et al., 2013) have now clearly identified genes that are uniquely expressed in microglia, providing a transcriptional signature for these cells and allowing a clear distinction from macrophages/monocytes. The field has been bedeviled by an inadequate nomenclature linked to the biology of macrophages (M1/M2), which is neither descriptive nor informative of microglial phenotypes. Butovsky et al. nicely show how TGFβ, typically thought to be associated with “alternative activation” states, regulates microglial gene expression. This is a solid start to a broader and more comprehensive understanding of the plasticity exhibited by this cell type and the development of a new nomenclature that allows us to accurately talk about their phenotypic states.


    . Identification of a unique TGF-β-dependent molecular and functional signature in microglia. Nat Neurosci. 2014 Jan;17(1):131-43. Epub 2013 Dec 8 PubMed.

    . Alternative activation of macrophages. Nat Rev Immunol. 2003 Jan;3(1):23-35. PubMed.

    . The microglial sensome revealed by direct RNA sequencing. Nat Neurosci. 2013 Dec;16(12):1896-905. Epub 2013 Oct 27 PubMed.

    . Exploring the full spectrum of macrophage activation. Nat Rev Immunol. 2008 Dec;8(12):958-69. PubMed.

  2. Twenty-two years ago, Stein, Keshav, Harris, and Gordon reported that elicited peritoneal macrophages differentially regulated the macrophage mannose receptor in response to IL-4 as compared to IFN-γ (Stein et al., 1992). These investigators inaugurated the concept of the alternatively activated macrophage to account for differential responses to different pathogens, and to distinguish acute inflammatory host defense responses as compared to subsequent tissue remodeling. Thereafter, these concepts, now called M1 and M2 activation-states, were extended to take additional stimuli into account, so M2a, M2b, and M2c pathways were proposed. The relevant pathobiology included the action of macrophages infiltrating tumors and chronically inflamed tissues as well as infected sites, and the behavior of every myeloid cell was in time lumped together under the M1/M2 rubrics. Myeloid cells isolated from the autopsy brain tissues of Alzheimer’s disease (AD) patients or from mouse models of AD were also classified in this fashion. 

    It must be emphasized that many investigators, led by Gordon and coworkers, proposed a subtle, intricate spectrum of gene-expression changes that would underlie the responses to the enormously varied circumstances confronted by infiltrating monocytes and tissue macrophages. However, the M1/M2 split was too convenient to ignore for some scientists, including many investigating neurodegenerative disease. The result was that the field unfortunately did not capitalize on the explanatory and discovery potential of expression profiling. 

    When directly assessed by comprehensive sequencing and informatics analysis, this M1/M2 classification was inadequate, as shown by Chiu, Maniatis, and colleagues in studies of cells from an ALS mouse model (Chiu et al., 2013). In the present study by Schultze et al., the dichotomous paradigm is laid finally to rest. Expression-profiling informatics analysis of human macrophages (blood monocytes incubated with GM-CSF or M-CSF) exposed to M1 and M2 or closely related stimuli showed the familiar polarization. As the evaluation was expanded to comprise 17 and then 29 stimuli (dynamically rendered in authors’ movie S1), it unmistakably could be discerned that there were spectra of expression profiles (the authors distinguished nine response patterns) that failed to align along the M1/M2 polarization axes. When kinetic patterns of gene expression were incorporated into the analysis, no polarization at all was evident but rather tightly-packed networks of separate expression profiles.

    It’s rather unlikely that the specific molecules delineated in this study will be relevant to the behavior of microglia in brains of individuals with ongoing neurodegeneration. Indeed, the authors found none of their inflammatory signatures in alveolar macrophages from subjects with chronic obstructive pulmonary disease (COPD), suggesting that the challenge of extracting expression signatures from those with distinct pathological conditions will be formidable. Nevertheless, comprehensive, unbiased profiling enabled by informatics analysis will certainly help us understand how myeloid cells, including microglia, respond to disease and simultaneously condition affected tissue. This paper, titled “Transcriptome-Based Network Analysis Reveals a Spectrum Model of Human Macrophage Activation,” is classified as a “Resource” and it unambiguously fulfills that mission.



    . Interleukin 4 potently enhances murine macrophage mannose receptor activity: a marker of alternative immunologic macrophage activation. J Exp Med. 1992 Jul 1;176(1):287-92. PubMed.

    . A neurodegeneration-specific gene-expression signature of acutely isolated microglia from an amyotrophic lateral sclerosis mouse model. Cell Rep. 2013 Jul 25;4(2):385-401. PubMed.