Original Article

Genes and Immunity (2011) 12, 445–456; doi:10.1038/gene.2011.17; published online 31 March 2011

Monocyte polarization: the relationship of genome-wide changes in H4 acetylation with polarization

Z Zhang1, L Song2, K Maurer2, A Bagashev2 and K E Sullivan2

  1. 1Center for Bioinformatics, Children's Hospital of Philadelphia, Philadelphia, PA, USA
  2. 2Division of Allergy Immunology, Children's Hospital of Philadelphia, Philadelphia, PA, USA

Correspondence: Dr KE Sullivan, Division of Allergy Immunology, Children's Hospital of Philadelphia, 3615 Civic Center Boulevard., Philadelphia, PA 19104, USA. E-mail: sullivak@mail.med.upenn.edu

Received 23 August 2010; Revised 17 December 2010; Accepted 17 January 2011; Published online 31 March 2011.



The character of monocytes is both molded by and contributes to ongoing immune responses. We hypothesized that monocyte polarization could have durable qualities and these would be mediated partly by changes in the chromatin. We defined genome-wide expression and histone H4 acetylation (H4ac) changes after γ-interferon (IFN), α-IFN and interleukin-4 treatment. To identify genes with altered potential for expression, we stimulated polarized monocytes and identified genes up- or downregulated after polarization and stimulation but not either treatment alone. We also defined durability after an 18-h or 3-day washout. Genes uniquely regulated after the combination of polarization and stimulus were durably altered, with 51% of the effects being durable. This gene set was highly enriched for cytokine-induced alterations in H4ac, with P-values ranging from 10−24 to 10−37. Certain regulons defined by patterns of expression were also associated with altered H4ac, with P-values ranging from 10−4 to 10−29. Networking software revealed a high density of mitogen-activated protein (MAP) kinase nodes in these clusters. Therefore, some changes in monocyte gene expression were sustained over a 3-day period. These durably altered gene sets were enriched for changes in H4ac and were associated with potential MAP kinase effects.


chromatin; epigenome; immune complex; cytokine



Monocytes and their tissue counterpart, the macrophage, are known to participate in innate defenses against pathogens and to signal and mold the adaptive immune response.1 The role of monocytes and macrophages in molding host responses is becoming clearer, and recent descriptions of the importance of anti-inflammatory monocytes in recovery from sepsis2 and the role of anti-inflammatory macrophages in the local immune suppression associated with tumors3 have highlighted the importance of this cell type. Understanding the mechanisms governing the qualitative features of monocytes has also become important, as targeting of these cells has become a therapeutic reality. In autoimmune diseases, tissue macrophages have been shown to correlate with the severity of renal disease in lupus and the severity of synovitis in rheumatoid arthritis, leading to interest in modulating their effects.4, 5 Glatiramer acetate has been shown to mediate its therapeutic effects in multiple sclerosis via the generation of anti-inflammatory monocytes, which secondarily induce Treg development and suppression of Th17 cells.6 There is also substantial interest in developing interventions to modulate macrophage function in the oncology field to enhance antitumor immune responses.3, 7 Therefore, understanding the molecular pathways involved in monocyte/macrophage polarization is of critical importance.

Macrophage plasticity is extreme, and several studies have demonstrated convincingly that macrophages can even transdifferentiate into smooth muscle cells or lymphatics.8, 9 Generally, macrophage function has both plastic and stable qualities.3, 7, 10, 11, 12, 13, 14 It would be valuable to identify features of monocyte/macrophage function that are durable and those which are subject to subsequent modification. In designing macrophage-oriented therapeutics, delineation of durable and plastic features will be important.

Traditional efforts to polarize monocytes use interleukin (IL)-4 or IL-13 for the ‘alternative’ M2 polarization and γ-interferon (IFN) or microbial ligands for ‘classical’ M1 polarization.15 M1 monocytes/macrophages are generally characterized as having IL-12hi, IL-23hi, tumor necrosis factor (TNF)-αhi and IL-10lo responses to stimulation.16, 17, 18 They produce abundant reactive oxygen species and drive Th1 responses.1 M2 monocytes/macrophages typically have IL-12lo, IL-23lo, TNF-αlo and IL-10hi responses.19, 20, 21 These cells are similar to myeloid suppressor cells and are thought to perform important regulatory functions and to participate in Th2 responses.1, 22 The classification of M1 and M2 has been valuable model; however, monocyte and macrophage phenotypes in vivo more likely represent points on a spectrum.23 Conversion of M2 monocytes into M1 monocytes could improve host responses to tumors, and conversion of M1 monocytes into M2 monocytes could benefit patients with autoimmune disease, therefore, understanding that the degree of plasticity after polarization is of significant value.3, 6, 7 Our study examined the effects of a strong M1 agent, a strong M2 agent and the use of an agent designed to induce a dendritic cell-like phenotype.

This study examined H4 acetylation (H4ac) changes associated with polarization. Epigenetic changes regulate gene expression, and the epigenetic landscape of the cell sets the developmental program of the cell.24, 25, 26 Macrophage polarization has been specifically associated with epigenetic pathways.27, 28 H4ac specifically is associated with increased competence for transcription, and a large number of studies have confirmed its importance in regulating gene expression.29, 30 The ability of epigenetic changes to sustain a differentiation program make it an attractive target for investigation of the mechanisms underlying monocyte polarization, and a recent study of altered histone modifications regulating M2 macrophage differentiation suggested that polarization in some cases is regulated epigenetically.28

We have previously demonstrated that the TNF-α locus is regulated by histone modifications, both in response to acute stimulation and in response to polarizing signals.31, 32 The TNF-α locus demonstrated some persistence of polarizing effects after γ-IFN but not after IL-4 polarization, and these persistent effects were mediated largely by increased H4ac. To begin to address the important question of polarization durability at a genome-wide level, we utilized an unbiased approach to characterize the relationship of H4ac and gene expression after polarization. We then defined the durability of the polarization effects by treating monocytes with polarizing cytokines, removing the cytokines and stimulating with immune complexes at various times after cytokine exposure. In this in vitro model system, specific subsets of durably altered gene expression patterns were associated with polarization-induced chromatin changes. The gene set with altered potential for expression, as revealed by the acute immune complex stimulus, was particularly enriched for cytokine-induced H4ac changes.



Cellular responses to polarization

We initially validated our polarization strategy using common cell surface markers of polarization.33 We polarized the primary human monocytes for 18h. Three contrasting polarizers were used: α-IFN, γ-IFN and IL-4. IL-4 and γ-IFN are often considered opposite polarizers, with γ-IFN driving an M1 phenotype and IL-4 driving an anti-inflammatory M2 phenotype.34, 35, 36, 37, 38 The effect of α-IFN is less well characterized, but is thought to be associated with dendritic cell-like qualities.39 This strategy allowed us to examine diverse effects on monocyte biology. Owing to some inherent heterogeneity in human peripheral blood monocytes,40 we characterized the percent of cells positive for each cell surface marker. The M1 phenotype, seen after γ-IFN treatment, was associated with increased expression of FcγRI (CD64) and CCR7 (CD197). The M2 phenotype, seen after IL-4 treatment, was associated with increased expression of FcεRII (CD23) and the macrophage mannose receptor (CD206). We then characterized the durability of the changes in expression of these cell surface markers after a washout of the cytokine polarizer (Figure 1). The durability of the polarization effects after a washout was heterogeneous, with IL-4 stably altering CD16 expression but transiently affecting CD23 expression. The γ-IFN had transient effects on CD197 and CD16 expression. Note that the maturation of primary human monocytes in culture is associated with alterations in some cell surface markers. At 3 days in culture, the monocytes have acquired some macrophage-like characteristics with partial adherence (see below).

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Flow cytometric identification of polarization-induced cell surface molecules. Primary monocytes were polarized with the indicated cytokines for 18h. The cells were studied using flow cytometry immediately and after a 3-day washout period. The cells were gated on physical parameters and CD14. CD16, CD11c, CD23, CD32, CD64, CD80, CD197, CD206 and CCR2 were studied with the graphs demonstrating the statistically significant changes after polarization. The first set of points represents the changes immediately after cytokine treatment. The arrows denote the statistically significant changes due to polarization (P<0.05). The second set of points reflects the cell surface staining after the 3-day washout. The graphs represent an average of four experiments with four separate donors. Representative flow plots are shown below.

Full figure and legend (276K)

To provide additional insight into the durability of cytokine polarization in monocytes, we polarized the cells as above and observed the physical changes in the cell morphology. These effects were heterogeneously durable with endosomes being more evident after IL-4 and α-IFN but only α-IFN-polarized cells retaining that feature after the washout (Figure 2). The γ-IFN polarization drove increased actin content that was retained after the washout.

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Cell morphology is altered by polarization. Primary monocytes were polarized with the indicated cytokines for 18h. Cells were stained with Alexafluor 488-phalloidin (green) to detect intracellular actin and anti-EEA1 (early endosomal antigen-1) (red) to detect endosomes. The cells were counter stained with 4’,6-diamidino-2-phenylindole to define the nucleus. Monocytes were studied immediately after the 18-h polarization treatment and after a 3-day washout. Some aspects of cell morphological changes are retained after the 3-day washout. The confocal shown is representative of three different donors.

Full figure and legend (305K)

Gene expression and H4ac in monocytes

H4ac at the promoter is known to be associated with increased competence for transcription.41 The acetylation mark can be deposited acutely by histone acetyl transferases or can be deposited during a differentiation-induced process. In both cases, the gene is rendered more transcriptionally competent, but the acute changes are transient while the differentiation-induced H4ac changes are more durable. In untreated monocytes, the H4ac levels within the (−7.5kb, +2.5kb) promoter region of 11152 Ensembl genes exhibited a location-specific pattern, with two peaks immediately before and after the transcription start site (Figure 3a). The correlation between H4ac and expression was strongest around the transcription start site (Figure 3b). Nevertheless, the H4ac as far, as −7.5kb was still significantly correlated with gene expression (P=4.7e–11), suggesting that H4ac could indirectly regulate the expression of the downstream gene via adjusting accessibility of transcription factors and other regulators. We then summarized the H4ac measurements within the promoter region of each gene into a single value. The representative H4ac measurement had a highly significant correlation with gene expression with Spearman's ρ=0.56 and P-value <10e–300 (Figure 3c). Although the overall correlation is prominent, many genes had discordant H4ac and expression measurements. In some cases, this could be because of gene-mapping issues. For example, ARHGAP11A had very high H4ac and very low expression according to the expression arrays. A closer look at the data revealed that all of the expression probes of ARHGAP11A were located on one of its alternative 3′-UTRs and probably failed to detect the actual mRNA.

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Gene expression correlates with H4ac at the promoter. (a) The H4ac modification was distributed primarily around the transcription start site (TSS). (b) The correlation of H4ac with expression was highest around the TSS. (c) H4 content in the promoter window correlates with gene expression. The solid line was generated by the Loess method.

Full figure and legend (89K)

Gene expression changes related to polarization

To examine the hypothesis that polarization-induced changes in gene expression are associated with alterations to chromatin, we first identified changes in gene expression related to polarization. Our model utilized polarization of primary monocytes with IL-4, γ-IFN or α-IFN. Our initial analysis compared gene sets from primary human monocytes from three donors polarized for 18h. Baseline was established by parallel cultures that were mock polarized. Each cytokine induced the expression of a unique set of genes (Figure 4a). There was little overlap between IL-4-treated cells and the two other polarized cell types but there was substantial overlap between γ-IFN- and α-IFN-treated cells. When comparing polarization-induced changes in gene expression and polarization-induced changes in H4ac, there was a statistically significant overlap for all three cytokines (Figure 4b). The extent of direct concordance ranged from 15 to 30%; however, this is dramatically increased over the background concordance, and P-values ranged from 10−24 to 10−37. The overlap among downregulated genes was less significant. We conclude that genes with increased expression after cytokine polarization are enriched for increased H4ac, but that increased H4ac does not dictate the majority of expression changes.

Figure 4.
Figure 4 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Cytokine-induced changed in gene expression and H4ac. Three monocyte donors were used for all manipulations. Array results were normalized and averaged for the three donors. (a) The three cytokines induced and repressed the expression of different, but partly overlapping, gene sets. The numbers indicate genes with at least a twofold change and a P-value of Student's t test <0.05 after cytokine treatment. (b) The cytokine polarization led to increased H4ac and increased gene expression. There was significant overlap in each case between increased gene expression and increased H4ac. The P-values and odds ratios for the concordance are given at the top of each Venn diagram.

Full figure and legend (63K)

Durability of polarization effects

To define the durability of the polarization effect, we examined the effect of polarization on a subsequent stimulus. Cells were polarized as above, and after varying periods of washout, the cells were stimulated with immune complexes. This stimulus engages Fc receptors and drives a proinflammatory cytokine response, similar to processes in immune complex-mediated diseases such as lupus.42, 43 The cellular pathways activated by immune complexes are well characterized and represent a physiological model system. RNA was harvested immediately after polarization or at 4h after immune complex stimulation to minimize secondary effects from cytokine release. This model allowed us to define not only gene expression changes directly related to the cytokine polarization but also to detect changes in the potential for gene expression due to polarization. The goal was to identify genes with altered potential for response as revealed by the immune complex stimulation. An important issue in the study of macrophages is whether they retain a ‘memory’ of their previous exposures. We specifically queried the array data for gene sets that fulfilled the following criteria: genes with expression altered by the combination of cytokine polarization plus immune complex stimulation but with minimal alteration due to polarization alone or stimulation alone. Figure 5 demonstrates the heatmaps of the genes fulfilling these criteria for each type of polarization. The left lanes represent the various controls, demonstrating that cytokine alone or immune complexes alone do not markedly alter expression in these gene sets. The three right-hand lanes demonstrate the sets of genes uniquely regulated by the combination of cytokine and immune complexes. The last two lanes are the same genes displayed after the 18-h and 3-day washout. This subgroup of genes, characterized by altered potential for expression as manifested by a response to the combination of cytokine and immune complexes, is generally durably altered. Overall, 51% of the genes demonstrated a stable effect of polarization after a washout.

Figure 5.
Figure 5 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Altered potential for expression is strongly associated with increased durability of expression. The heatmaps demonstrate that the sets of genes with altered competence for expression, as defined by altered expression after the combination of polarization and immune complex stimulation but much less alteration with each agent alone, represent extremely durably altered sets of genes. This was true for downregulated genes (lower group) as well as upregulated genes (upper group). These heatmaps represent the average of three donors, with all treatments being applied to all three donors. For each cytokine, the last three lanes represent the gene set defined by response to the combination of polarization and immune complex stimulation, the same genes after an 18-h washout and the same genes after a 3-day washout. The first five lanes in each case represent the relevant controls.

Full figure and legend (211K)

Chromatin correlates with gene expression changes

We hypothesized that the gene set with altered potential for expression would correspond, in part, with a gene set with altered H4ac after polarization. The role of histone modifications in defining transcriptional competence suggests there will be an association between H4ac and gene expression. Each Venn diagram in Figure 6 demonstrates the overlap in H4ac changes related to polarization and gene expression. As was seen in Figure 3, the overlap is incomplete but reveals statistically significant associations in H4ac and gene expression changes in this model system. The small orange circles indicate the durably altered gene set. Therefore, the gene set defined by altered potential for expression was significantly associated with altered H4ac, but the subset defined by durability was not additionally enriched for H4ac.

Figure 6.
Figure 6 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

The association of durably altered gene expression with altered H4ac. We analyzed only genes with altered potential for expression as defined by altered expression after the combination of cytokine and immune complex stimulation (Figure 5). Overall, 51% of the genes retained their altered expression after the washout. The green circles indicate the genes with altered H4ac after cytokine treatment and the yellow circles represent the genes with altered expression after cytokine polarization and immune complex stimulation. The small orange circle represents the genes with durably altered expression after the washout and subsequent stimulation. There is significant overlap of the genes with altered potential for expression and the cytokine-induced alteration to H4ac according to Fisher's exact test (see P-values in figure). The durably altered gene set was not statistically associated with altered H4ac; however, the sample size was small in this analysis.

Full figure and legend (112K)

To further understand the pathways regulating the durably altered sets of genes, we performed unsupervised clustering of genes based on their expression pattern after polarization and washout (Table 1 and Figure 7). We hypothesized that specific groups of related genes might have a higher correlation with increased H4ac. We reexamined each of the clusters defined by the pattern of expression (Figure 7). By segregating the clusters, it was clear that certain clusters had a high correlation with altered H4ac and this association was much stronger for upregulated genes and H4ac. There was no consistent pattern of expression in those clusters strongly associated with increased H4ac. The clusters with a strong association with increased H4ac were selected for further evaluation (focus clusters). In this analysis, P-values ranged from 10−4 to 10−29.

Figure 7.
Figure 7 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Individual regulons defined by pattern of expression are highly enriched for H4ac. The pattern of expression of the highly concordant regulons identified in Table 1 is shown in this figure. Three clusters from each cytokine polarization were selected for demonstration based on their association with H4ac (Table 1). The clusters were defined by the pattern of changes in gene expression after immune complex stimulation and washout. The red lines track changes in upregulated genes and the green lines track changes in downregulated genes.

Full figure and legend (149K)

Ingenuity was utilized to define gene categories of related genes within the focus clusters. The two clusters of genes upregulated after IL-4 polarization and associated with increased H4ac were significantly associated with cell–cell signaling, cell morphology, cell development, cell assembly and gene expression (Table 2). There were five clusters identified as upregulated after γ-IFN polarization. These categories were similar to those seen after IL-4 polarization;, however, they were more diverse. Four clusters of genes were identified after α-IFN polarization. These four clusters exhibited features distinct from those of γ-IFN and IL-4. Therefore, the focus clusters were not clearly related to conserved gene categories across the three polarizing conditions.

To gain insight into the signaling pathways related to these regulons, a systems biology approach was used to derive networks relating genes from the 11 focus clusters. In each case, nodes of the polarizing cytokine were identified. Regulons identified after γ-IFN were associated with multiple networks centered on nodes of mitogen-activated protein (MAP) kinases, nuclear factor-κB, hepatocyte nuclear factor-4A and STAT1 (Table 3 and Supplementary Data). IL-4-polarized cells had significant nodes of peroxisome proliferator-activated receptor-γ, nuclear factor-κB and MAP kinases. The α-IFN-polarized cells had significant nodes of nuclear factor-κB, IL-12, MAP kinases, TRAIL and STAT1.

The role of MAP kinases in durably altered gene expression

The consistent identification of MAP kinases was of interest because we have previously reported that MAP kinase inhibitors altered the ability of γ-IFN to elicit chromatin changes at the TNF-α locus.31 Furthermore, recent studies suggest that MAP kinases directly target histone modification pathways to regulate gene expression.44, 45, 46 To better understand potential mechanisms, we identified the effect of polarization on signaling molecules downstream of immune complex stimulation (Figure 8). Primary monocytes were polarized with cytokine for 18h, and cells were examined at various points after the immune complex stimulation. Phosphorylation of activating transcription factor-2, ERK and p38 was detected by western blotting after immune complex stimulation using phospho-specific antibodies. Cytokine polarization led to basal changes in the phosphorylation of ERK and p38. The effect of α-IFN polarization on p38 and ERK was clearly visible as sustained phosphorylation after stimulation. The effect of γ-IFN was subtler, although it appeared to lead to a more sustained phosphorylation of activating transcription factor-2 compared with untreated cells. IL-4 had its most marked effect on activating transcription factor-2 phosphorylation and little effect on p38 or ERK. These data support the link between cytokine polarization, MAP kinase activation and H4ac.

Figure 8.
Figure 8 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

MAP kinase phosphorylation. MAP kinases were implicated in our systems biology approach. Cytokine polarization induced baseline changes in phosphorylation ERK and p38. The acute effect of immune complex stimulation drives phosphorylation of activating transcription factor-2, ERK and p38. Cytokine pretreatment for 18h modifies both the duration and magnitude of the response to immune complexes. The western blots shown are representative of two independent experiments.

Full figure and legend (78K)

We similarly examined STAT1 and STAT6 phosphorylation as additional pathways related to cytokine polarization. STAT6 exhibited increased phosphorylation after IL-4 polarization but was not affected by immune complex stimulation (Supplementary Data). STAT1 phosphorylation was transient after α-IFN and γ-IFN; however, polarization led to altered phosphorylation after immune complex stimulation with α-IFN, inhibiting STAT1 phosphorylation (Supplementary Data). Overall, these data support our pathway identification using a systems biology approach.



This study was designed to clarify the mechanisms underlying the recognized heterogeneity in monocyte polarization durability.7, 13, 14, 31 We hypothesized that chromatin changes could be responsible for some aspects of durability. Our use of three different monocyte-polarizing conditions allowed us to examine diverse effects on gene expression and the epigenome. The non-biased approach afforded an opportunity to characterize the gene subsets falling into distinct regulons, based on expression patterns in response to cytokines and washout. Several of the regulons were found to be strongly associated with alterations in H4ac but many regulons were not associated with H4ac changes. These data suggest that certain signaling pathways may be more likely to drive durable changes in gene expression via chromatin mechanisms. A systems biology approach identified MAP kinases as central to the polarization process. The role of MAP kinases in histone modifications has recently been highlighted.44, 45, 46 They participate in differentiation-induced changes to the epigenome.

Gene expression and H4ac are clearly linked; however, the strongest association of coordinately altered gene expression and H4ac occurred in sets of genes where cytokine altered the potential for expression. This reinforces the point that histone modifications set the competence for expression and are not simply a mirror of gene expression.47 The finding that not all genes with altered potential for expression were associated with altered H4ac prompts the question of what else could be contributing to the altered gene expression. A limitation of our study was the evaluation of a single-histone mark, a limitation imposed by the high cell numbers required for this type of assay. Therefore, other epigenetic modifications that were not measured could contribute to the altered potential for expression and durability. Nevertheless, this study clearly demonstrated that genes with altered potential for expression after polarization were generally enriched for cytokine-induced H4ac, and that certain clusters/regulons were highly enriched for cytokine-induced H4ac.

The two significant findings of this study are that monocyte polarization is heterogeneous with respect to durability and that in specific sets of genes, durability of changes in gene expression are related to cytokine-induced alteration in the H4ac pattern. Certain gene clusters identified by unsupervised clustering were extremely enriched for coinduced H4ac changes. These clusters were also enriched for MAP kinase nodes identified by networking analysis. MAP kinases have traditionally been thought to regulate transcription by phosphorylating transcription factors.48, 49 They have also been recognized for many years to phosphorylate H3, and thereby regulate the expression of immediate-early genes.49, 50, 51, 52 An additional layer of complexity has been added whereby MAP kinases are now also known to regulate polycomb binding and interact with histone acetyltransferases.44, 45 These recent data provide a potential link between MAP kinases and monocyte polarization.

The polarization process was remarkable for significant effects on cell behavior. Cell morphology, cell responses to stimuli and cell surface markers were all altered by polarization. Although many of the polarization-induced changes were reversible after washout, many were not. The durability of the changes described in this study is not antigen-specific memory but could serve to mold subsequent responses and to have an impact on the T-cell responses to pathogens. This echo or footprint of a previous exposure has important implications for the hygiene hypothesis. Although this study relied on an in vitro model to assess durability, it will also be important to know whether the durable changes in gene expression are maintained after macrophage differentiation in tissues.

This study examined the persistence of the polarization effect after exposure to a neutral environment. Whether offering an alternative stimulus might be able to reverse the durable polarization effects is not yet known. This is of importance, as monocyte-oriented therapeutics become a reality. Understanding the mechanisms underlying the durable gene expression changes and how to reverse them could have relevance in tumor immunotherapy, treatment of inflammatory diseases and recovery from infections. The identification of MAP kinases as a potential target represents an initial step in the process.


Materials and methods


Primary human monocytes were obtained by elutriation and rested before cytokine treatment.31 Donors gave written informed consent, and the protocol was approved by the IRB. Monocytes were greater than or equal to95% pure by flow cytometry and cultured in RPMI with 10% cosmic calf serum. Each set of monocytes from a single donor received all manipulations for the arrays to limit the effect of interindividual variability. Ten million monocytes were used for each chromatin immunoprecipitation/RNA preparation, and three biological replicates from different donors were used for each analysis. Monocytes were treated for 18h with 50ngml−1 of γ-IFN, 50ngml−1 of IL-4 (R&D Systems Minneapolis, MN, USA) or 500Uml−1 of α-IFN (PBL Interferon Source, Piscataway, NJ, USA). The monocytes were washed extensively and rested for varying times as indicated in the figures and in some cases, stimulated or mock stimulated with rabbit anti-bovine serum albumin immune complexes at 1μgml−1 for 4h.53 The immune complexes were free of endotoxin and had been previously titrated for effect. The cytokine concentrations have been previously shown to elicit strong polarization with no increase in cell death.31

For flow cytometric evaluations of changes in cell surface markers, four color analyses were performed on a FacsCalibur Instrument (BD Biosciences, San Jose, CA, USA). Isotype controls were used to define background staining, and the following antibodies were used in combination with PE-CD14 as an anchor: CD16, CD32, CD64, CD80, CD206 (all fluorescein isothiocyanate and all from BD Biosciences, San Jose, CA, USA) and AlexaFluor 647 CCR2 (BD Biosciences). APC-CD11c from BD Biosciences, FITC-CD23 from Biolegend (San Diego, CA, USA) and FITC-CD197 from R&D Systems were also used. The monocytes were often heterogeneous with respect to their surface expression, as has been previously described.54 Therefore, the changes after cytokine treatment are reported as percent of the population.

Confocal microscopy utilized a Zeiss Axiovert 200M microscope (Zeiss, Thornwall, NY, USA) with an argon laser (488nm wavelength) and a 1543nM HeNe coherent chameleon fs-pulsed NIR laser. Z-sections at the depth of 0.25–0.45μm were generated. Cells were fixed at various time points in 4% paraformaldehyde in PBS, permeabilized with 1% Triton X-100 SigmaUltra, blocked with 1% BSA in PBS at room temperature and counterstained in DAPI (Sigma). AlexaFluor 488-phalloidin was used to define actin (Invitrogen, Carlsbad, CA, USA) and anti-EEA1 (Santa Cruz, CA, USA) was used to define endosomes. The western blots utilized the following antibodies: anti-phospho-ATF2, anti-phospho ERK, anti-phospho P38, anti-tubulin or actin from Santa Cruz Biotechnology (Santa Cruz, CA, USA), pSTAT6 (Santa Cruz Biotechnology) and pSTAT1 (Cell Signalling Technology, Danvers, MA, USA).


The U133A 2.0 platform was used for the expression analyses. cRNA was prepared according to the recommendations of the manufacturer (Affymetrix, Santa Clara, CA, USA). The H4ac immunoprecipitation was performed as in previous studies.31 Purified DNA from the immunoprecipitation was amplified, cleaved and labeled using the GeneChip WT double-stranded DNA terminal labeling kit (Affymetrix). DNA preparation and hybridization were all performed according the recommendations for the GeneChip Human Promoter 1.0R array (Affymetrix).

Array data preprocessing

The CEL files of Human Promoter 1.0R arrays were processed by MAT,55 which adjusted the raw measurements according to probe sequences. The processing generated the average difference between three samples in the same treatment group and four common control samples where the DNA was precipitated by the non-specific antibody, anti-glutathione-S-transferase (GST). The processed data of more than 4.4 million probes were retrieved by CisGenome56 and input into R (http://www.R-project.org) for analyses. The expression arrays were processed by the RMA method57 using a custom library file that remapped the probes to the current release of Ensembl genes.58 The two types of array data were mapped to Ensembl genes and their transcription start site. In total, there were 11152 unique Ensembl genes that had at least 3 probes on the expression arrays and at least 30 probes within the (−7.5kb, +2.5kb) promoter region on the chromatin immunoprecipitation arrays. These genes were used as the ‘universe’ of all data analyses in this study.

Array data analyses

The raw data of the tiling arrays were processed by MAT program (Bioconductor, Fred Hutchinson Cancer Center, Seattle, WA, USA) by setting the bandwidth to 250bp and the P-value cutoff to 1e–6. Triplicate samples of each treatment were grouped together and compared with four GST control samples. MAT identified about 5000 H4ac peaks from each treatment group. The MAT-processed measurements on array probes were retrieved by the bar2txt command of the CisGenome program. The systematic difference of the scale of the measurements was removed by a quantile–quantile normalization of all measurements across treatment groups. The rest of the data processing, statistical analysis and result plotting were performed within R.

Tiling array probes were grouped into 100-bp bins, and the measurements of probes in the same bin were averaged to represent the H4ac level at that bin. The H4ac-expression correlation was calculated for each bin. To summarize the H4ac level of each Ensembl gene within −7.5kb to +2.5kb of transcription start site, the measurements at all 100 bins were weighted by their H4ac-expression correlation coefficients and then averaged. Thus, a single measurement was used to represent the general H4ac level at the promoter region of each gene in each treatment group.

To identify genes whose H4ac at promoter was changed by cytokine treatment, we used D to designate the difference of the single-value H4ac measurements of each gene before and after treatment. Given M and s.d. as the mean and standard deviation of Ds of all 11152 genes, we called that a gene had up- or downregulated H4ac if its D was greater than M+2 × s.d. or less than M−2 × s.d.

Genes altered by the combination of cytokine and immune complexes (IC) were identified by the similar strategy, but the samples treated by both were compared with three groups: no treatment, cytokine only and IC only. Only genes satisfying the M+2 × s.d. or M−2 × s.d. cutoffs in all three comparisons were selected. The durability of the change of these genes after washout was evaluated by comparing the six samples treated by both cytokine and IC, and rested for 18 or 72h after washout to the three corresponding sample treated by IC only. If the average difference was more than M+2 × s.d. or less than M−2 × s.d., the expression change was considered durable.

We applied the following steps for gene-clustering analysis. First, we identified significant genes that were differentially expressed across any of eight sample groups: four groups without cytokine polarization (one without IC, one with IC and two with 18 and 72h of rest after washout) and four corresponding groups with one cytokine treatment. All selected genes had a P-value less than 0.01 according to analysis of variance and at least 25% difference of average expression between any two sample groups. These genes were used to build a hierarchical-clustering tree. The tree was cut by height of 0.8 to get a number of gene clusters. The overlapping of those clusters to genes with H4ac change was evaluated by Fisher's exact test.

Gene Expression Omnibus (GEO) submission of data in process.


Conflict of interest

The authors declare no conflict of interest.



  1. Mills CD, Kincaid K, Alt JM, Heilman MJ, Hill AM. M-1/M-2 macrophages and the Th1/Th2 paradigm. J Immunol 2000; 164: 6166–6173. | PubMed | ISI | ChemPort |
  2. Mehta A, Brewington R, Chatterji M, Zoubine M, Kinasewitz GT, Peer GT et al. Infection-induced modulation of m1 and m2 phenotypes in circulating monocytes: role in immune monitoring and early prognosis of sepsis. Shock 2004; 22: 423–430. | Article | PubMed | ISI |
  3. Sinha P, Clements VK, Ostrand-Rosenberg S. Reduction of myeloid-derived suppressor cells and induction of M1 macrophages facilitate the rejection of established metastatic disease. J Immunol 2005; 174: 636–645. | PubMed | ISI | ChemPort |
  4. Hamilton JA, Tak PP. The dynamics of macrophage lineage populations in inflammatory and autoimmune diseases. Arthritis Rheum 2009; 60: 1210–1221. | Article | PubMed | ISI |
  5. Schiffer L, Bethunaickan R, Ramanujam M, Huang W, Schiffer M, Tao H et al. Activated renal macrophages are markers of disease onset and disease remission in lupus nephritis. J Immunol 2008; 180: 1938–1947. | PubMed | ISI | ChemPort |
  6. Weber MS, Prod’homme T, Youssef S, Dunn SE, Rundle CD, Lee L et al. Type II monocytes modulate T cell-mediated central nervous system autoimmune disease. Nat Med 2007; 13: 935–943. | Article | PubMed | ISI | ChemPort |
  7. Watkins SK, Egilmez NK, Suttles J, Stout RD. IL-12 rapidly alters the functional profile of tumor-associated and tumor-infiltrating macrophages in vitro and in vivo. J Immunol 2007; 178: 1357–1362. | PubMed | ISI | ChemPort |
  8. Maruyama K, Ii M, Cursiefen C, Jackson DG, Keino H, Tomita M et al. Inflammation-induced lymphangiogenesis in the cornea arises from CD11b-positive macrophages. J Clin Invest 2005; 115: 2363–2372. | Article | PubMed | ISI | ChemPort |
  9. Ninomiya K, Takahashi A, Fujioka Y, Ishikawa Y, Yokoyama M. Transforming growth factor-beta signaling enhances transdifferentiation of macrophages into smooth muscle-like cells. Hypertens Res 2006; 29: 269–276. | Article | PubMed | ISI | ChemPort |
  10. Mylonas KJ, Nair MG, Prieto-Lafuente L, Paape D, Allen JE. Alternatively activated macrophages elicited by helminth infection can be reprogrammed to enable microbial killing. J Immunol 2009; 182: 3084–3094. | Article | PubMed | ISI |
  11. Biswas SK, Sica A, Lewis CE. Plasticity of macrophage function during tumor progression: regulation by distinct molecular mechanisms. J Immunol 2008; 180: 2011–2017. | PubMed | ISI | ChemPort |
  12. Guiducci C, Vicari AP, Sangaletti S, Trinchieri G, Colombo MP. Redirecting in vivo elicited tumor infiltrating macrophages and dendritic cells towards tumor rejection. Cancer Res 2005; 65: 3437–3446. | PubMed | ISI | ChemPort |
  13. Stout RD, Jiang C, Matta B, Tietzel I, Watkins SK, Suttles J. Macrophages sequentially change their functional phenotype in response to changes in microenvironmental influences. J Immunol 2005; 175: 342–349. | PubMed | ISI | ChemPort |
  14. Stout RD, Suttles J. Functional plasticity of macrophages: reversible adaptation to changing microenvironments. J Leukoc Biol 2004; 76: 509–513. | Article | PubMed | ISI | ChemPort |
  15. Mantovani A, Sica A, Locati M. Macrophage polarization comes of age. Immunity 2005; 23: 344–346. | Article | PubMed | ISI | ChemPort |
  16. Goerdt S, Politz O, Schledzewski K, Birk R, Gratchev A, Guillot P et al. Alternative versus classical activation of macrophages. Pathobiology 1999; 67: 222–226. | Article | PubMed | ISI | ChemPort |
  17. Edwards JP, Zhang X, Frauwirth KA, Mosser DM. Biochemical and functional characterization of three activated macrophage populations. J Leukoc Biol 2006; 80: 1298–1307. | Article | PubMed | ISI | ChemPort |
  18. Baetselier PD, Namangala B, Noel W, Brys L, Pays E, Beschin A. Alternative versus classical macrophage activation during experimental African trypanosomosis. Int J Parasitol 2001; 31: 575–587. | Article | PubMed | ChemPort |
  19. Raes G, Van den Bergh R, De Baetselier P, Ghassabeh GH, Scotton C, Locati M et al. Arginase-1 and Ym1 are markers for murine, but not human, alternatively activated myeloid cells. J Immunol 2005; 174: 6561–6562. | PubMed | ISI | ChemPort |
  20. Martinez FO, Gordon S, Locati M, Mantovani A. Transcriptional profiling of the human monocyte-to-macrophage differentiation and polarization: new molecules and patterns of gene expression. J Immunol 2006; 177: 7303–7311. | PubMed | ISI | ChemPort |
  21. Lumeng CN, Bodzin JL, Saltiel AR. Obesity induces a phenotypic switch in adipose tissue macrophage polarization. J Clin Invest 2007; 117: 175–184. | Article | PubMed | ISI | ChemPort |
  22. Biswas SK, Gangi L, Paul S, Schioppa T, Saccani A, Sironi M et al. A distinct and unique transcriptional program expressed by tumor-associated macrophages (defective NF-kappaB and enhanced IRF-3/STAT1 activation). Blood 2006; 107: 2112–2122. | Article | PubMed | ISI | ChemPort |
  23. Ghassabeh GH, De Baetselier P, Brys L, Noel W, Van Ginderachter JA, Meerschaut S et al. Identification of a common gene signature for type II cytokine-associated myeloid cells elicited in vivo in different pathologic conditions. Blood 2006; 108: 575–583. | Article | PubMed | ISI |
  24. Bird A. DNA methylation patterns and epigenetic memory. Genes Dev 2002; 16: 6–21. | Article | PubMed | ISI | ChemPort |
  25. Reik W. Stability and flexibility of epigenetic gene regulation in mammalian development. Nature 2007; 447: 425–432. | Article | PubMed | ISI | ChemPort |
  26. Jones PA, Baylin SB. The epigenomics of cancer. Cell 2007; 128: 683–692. | Article | PubMed | ISI | ChemPort |
  27. Satoh T, Takeuchi O, Vandenbon A, Yasuda K, Tanaka Y, Kumagai Y et al. The Jmjd3-Irf4 axis regulates M2 macrophage polarization and host responses against helminth infection. Nat Immunol 2010; 11: 936–944. | Article | PubMed | ISI | ChemPort |
  28. Ishii M, Wen H, Corsa CA, Liu T, Coelho AL, Allen RM et al. Epigenetic regulation of the alternatively activated macrophage phenotype. Blood 2009; 114: 3244–3254. | Article | PubMed | ISI | ChemPort |
  29. Agalioti T, Chen G, Thanos D. Deciphering the transcriptional histone acetylation code for a human gene. Cell 2002; 111: 381–392. | Article | PubMed | ISI | ChemPort |
  30. Schubeler D, MacAlpine DM, Scalzo D, Wirbelauer C, Kooperberg C, van Leeuwen F et al. The histone modification pattern of active genes revealed through genome-wide chromatin analysis of a higher eukaryote. Genes Dev 2004; 18: 1263–1271. | Article | PubMed | ISI |
  31. Garrett S, Dietzmann-Maurer K, Song L, Sullivan KE. Polarization of primary human monocytes by IFN-gamma induces chromatin changes and recruits RNA Pol II to the TNF-alpha promoter. J Immunol 2008; 180: 5257–5266. | PubMed | ISI | ChemPort |
  32. Sullivan KE, Reddy AB, Dietzmann K, Suriano AR, Kocieda VP, Stewart M et al. Epigenetic regulation of tumor necrosis factor alpha. Mol Cell Biol 2007; 27: 5147–5160. | Article | PubMed | ISI | ChemPort |
  33. Strauss-Ayali D, Conrad SM, Mosser DM. Monocyte subpopulations and their differentiation patterns during infection. J Leukoc Biol 2007; 82: 244–252. | Article | PubMed | ISI | ChemPort |
  34. Gordon S. Alternative activation of macrophages. Nat Rev Immunol 2003; 3: 23–35. | Article | PubMed | ISI | ChemPort |
  35. Munder M, Eichmann K, Moran JM, Centeno F, Soler G, Modolell M. Th1/Th2-regulated expression of arginase isoforms in murine macrophages and dendritic cells. J Immunol 1999; 163: 3771–3777. | PubMed | ISI | ChemPort |
  36. Raes G, De Baetselier P, Noel W, Beschin A, Brombacher F, Hassanzadeh Gh G. Differential expression of FIZZ1 and Ym1 in alternatively versus classically activated macrophages. J Leukoc Biol 2002; 71: 597–602. | PubMed | ISI | ChemPort |
  37. Welch JS, Escoubet-Lozach L, Sykes DB, Liddiard K, Greaves DR, Glass CK. TH2 cytokines and allergic challenge induce Ym1 expression in macrophages by a STAT6-dependent mechanism. J Biol Chem 2002; 277: 42821–42829. | Article | PubMed | ISI | ChemPort |
  38. Martinez FO, Gordon S, Locati M, Mantovani A. Transcriptional profiling of the human monocyte-to-macrophage differentiation and polarization: new molecules and patterns of gene expression. J Immunol 2006; 177: 7303–7311. | PubMed | ISI | ChemPort |
  39. Then Bergh F, Dayyani F, Ziegler-Heitbrock L. Impact of type-I-interferon on monocyte subsets and their differentiation to dendritic cells. An in vivo and ex vivo study in multiple sclerosis patients treated with interferon-beta. J Neuroimmunol 2004; 146: 176–188. | Article | PubMed |
  40. Cros J, Cagnard N, Woollard K, Patey N, Zhang SY, Senechal B et al. Human CD14dim monocytes patrol and sense nucleic acids and viruses via TLR7 and TLR8 receptors. Immunity 2010; 33: 375–386. | Article | PubMed | ISI | ChemPort |
  41. Clayton AL, Hazzalin CA, Mahadevan LC. Enhanced histone acetylation and transcription: a dynamic perspective. Mol Cell 2006; 23: 289–296. | Article | PubMed | ISI | ChemPort |
  42. Sullivan KE, Jawad AF, Piliero LM, Kim N, Luan X, Goldman D et al. Analysis of polymorphisms affecting immune complex handling in systemic lupus erythematosus. Rheumatology 2003; 42: 446–452. | Article | PubMed | ISI |
  43. Barrionuevo P, Beigier-Bompadre M, Fernandez GC, Gomez S, Alves-Rosa MF, Palermo MS et al. Immune complex-FcgammaR interaction modulates monocyte/macrophage molecules involved in inflammation and immune response. Clin Exp Immunol 2003; 133: 200–207. | Article | PubMed | ISI |
  44. Gehani SS, Agrawal-Singh S, Dietrich N, Christophersen NS, Helin K, Hansen K. Polycomb group protein displacement and gene activation through MSK-dependent H3K27me3S28 phosphorylation. Mol Cell 2010; 39: 886–900. | Article | PubMed | ISI |
  45. Suganuma T, Mushegian A, Swanson SK, Abmayr SM, Florens L, Washburn MP et al. The ATAC acetyltransferase complex coordinates MAP kinases to regulate JNK target genes. Cell 2010; 142: 726–736. | Article | PubMed | ISI |
  46. Drobic B, Perez-Cadahia B, Yu J, Kung SK, Davie JR. Promoter chromatin remodeling of immediate-early genes is mediated through H3 phosphorylation at either serine 28 or 10 by the MSK1 multi-protein complex. Nucleic Acids Res 2010; 38: 3196–3208. | Article | PubMed | ISI |
  47. Heintzman ND, Stuart RK, Hon G, Fu Y, Ching CW, Hawkins RD et al. Distinct and predictive chromatin signatures of transcriptional promoters and enhancers in the human genome. Nat Genet 2007; 39: 311–318. | Article | PubMed | ISI | ChemPort |
  48. Edmunds JW, Mahadevan LC. MAP kinases as structural adaptors and enzymatic activators in transcription complexes. J Cell Sci 2004; 117(Part 17): 3715–3723. | Article | PubMed | ISI | ChemPort |
  49. Thomson S, Mahadevan LC, Clayton AL. MAP kinase-mediated signalling to nucleosomes and immediate-early gene induction. Semin Cell Dev Biol 1999; 10: 205–214. | Article | PubMed | ISI | ChemPort |
  50. Nowak SJ, Corces VG. Phosphorylation of histone H3: a balancing act between chromosome condensation and transcriptional activation. Trends Genet 2004; 20: 214–220. | Article | PubMed | ISI | ChemPort |
  51. Strelkov IS, Davie JR. Ser-10 phosphorylation of histone H3 and immediate early gene expression in oncogene-transformed mouse fibroblasts. Cancer Res 2002; 62: 75–78. | PubMed | ISI | ChemPort |
  52. Thomson S, Clayton AL, Mahadevan LC. Independent dynamic regulation of histone phosphorylation and acetylation during immediate-early gene induction. Mol Cell 2001; 8: 1231–1241. | Article | PubMed | ISI | ChemPort |
  53. Xiao S, Xu C, Jarvis JN. C1q-bearing immune complexes induce IL-8 secretion in human umbilical vein endothelial cells (HUVEC) through protein tyrosine kinase- and mitogen-activated protein kinase-dependent mechanisms: evidence that the 126 kD phagocytic C1q receptor mediates immune complex activation of HUVEC. Clin Exp Immunol 2001; 125: 360–367. | Article | PubMed | ISI |
  54. Belge KU, Dayyani F, Horelt A, Siedlar M, Frankenberger M, Frankenberger B et al. The proinflammatory CD14(+)CD16(+) DR(++) monocytes are a major source of TNF. J Immunol 2002; 168: 3536–3542. | PubMed | ISI | ChemPort |
  55. Johnson WE, Li W, Meyer CA, Gottardo R, Carroll JS, Brown M et al. Model-based analysis of tiling-arrays for ChIP-chip. Proc Natl Acad Sci USA 2006; 103: 12457–12462. | Article | PubMed | ChemPort |
  56. Ji H, Jiang H, Ma W, Johnson DS, Myers RM, Wong WH. An integrated software system for analyzing ChIP-chip and ChIP-seq data. Nat Biotechnol 2008; 26: 1293–1300. | Article | PubMed | ISI | ChemPort |
  57. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res 2003; 31: e15. | Article | PubMed | ChemPort |
  58. Dai M, Wang P, Boyd AD, Kostov G, Athey B, Jones EG et al. Evolving gene/transcript definitions significantly alter the interpretation of GeneChip data. Nucleic Acids Res 2005; 33: e175. | Article | PubMed | ChemPort |


This work was supported, in part, by NIH R01 AI 0511323 and R01 ES 017627.

Supplementary Information accompanies the paper on Genes and Immunity website