Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jul 15.
Published in final edited form as: J Immunol. 2021 Jun 30;207(2):569–576. doi: 10.4049/jimmunol.2001324

Early-derived murine macrophages temporarily renounce tissue identity during acute systemic inflammation

Radika Soysa *, Jonathan C Bean †, Xia Wu *, Sarah Lampert *, Sebastian Yuen *, Ian N Crispe *
PMCID: PMC8516673  NIHMSID: NIHMS1704536  PMID: 34193604

Abstract

In mice, a subset of cardiac macrophages and Kupffer cells derive from fetal precursors, seed the developing tissues, self-renew locally and persist into adulthood. Here we investigated how these cells survive acute systemic inflammation. In both tissues, early-derived subsets rapidly responded to acute systemic inflammation by assuming a temporary non-classical activation state featuring upregulation of both pro-inflammatory (Il1b, Tnf, Nfkb1), and anti-inflammatory (Il10, Il4ra, Nfkbiz) genes. During this process, transcription factor genes associated with myeloid identity (Spi1,Zeb2) were upregulated while those associated with tissue specificity (Nr1h3 for Kupffer cells and Nfatc2, Irf4 for cardiac macrophages) were down regulated, suggesting that the cells reasserted their myeloid identity but renounced their tissue identity. Most of these changes in gene expression reverted to steady state levels post resolution. We conclude that these early-derived macrophage subsets are resilient in the face of acute stress by temporarily loss of adaptation to local tissue-specific niches, while reasserting their generic myeloid identity.

Introduction

Cardiac macrophages and Kupffer cells participate in tissue stability, repair, inflammation and its resolution(1, 2). At steady state, a subset of these macrophages is embryo-derived and long-lived(3–8). These macrophages originate from yolk sac erythromyeloid progenitors and fetal monocytes that seed the organs during the embryonic or neonatal stages of development(5, 9). Once seeded these macrophages self-renew and live into the adult stage. Thus, these embryo-derived macrophages need mechanisms to remain resilient for an extended time through developmental changes and external stresses.

These embryo-derived cardiac macrophages and Kupffer cells are immunologically active and survive stress(10, 11). During cardiomyocyte injury, the embryo-derived macrophage subset in the neonatal heart expanded and supported repair(12). Ablation of this subset limited cardiac infarct healing in adult mice. We have previously shown that fetal-derived Kupffer cells survive ionizing radiation injury via cell cycle arrest mediated by the cyclin-dependent kinase inhibitor p21Cip1/WAF1 (13). Both those macrophage subsets have the capacity to respond to systemic inflammation such as sepsis, and may form an interconnected network; for example, Kupffer cells were activated during myocardial infarction(11). While these studies further show that the embryo-derived macrophages self renew after stress, the state of these cells post recovery remains elusive.

We therefore investigated these fetal-derived macrophages, evaluating their gene expression at steady state and post-exposure to bacterial lipopolysaccharide (LPS) endotoxin-induced systemic acute inflammation and resolution. Using the CX3CR1 inducible fate mapping system we traced the CX3CR1-marked early-derived macrophage subset in each organ and evaluated the changes in actively translating mRNAs. In response to LPS-induced inflammation both subsets mounted a common response as well as tissue specific responses. The acute inflammatory response included upregulation of both pro- and anti-inflammatory genes (Tnf, Il10) and transcription factor genes (Nfkb1, Nfkbiz). Immediately post LPS-stimulation, tissue specific transcription factor genes (Nr1h3, Nfatc2) and those associated with inhibition of self-renewal (Mafb, Zfp90, Zbtb4) were down regulated in each macrophage subset. At this early time point, the essential transcription factor genes for macrophage identity (Spi1, Zeb2) were upregulated. Post resolution both macrophage subsets resumed their steady state gene expression. These findings reveal that these early-derived resident macrophage subsets mounted a temporary immune response, renouncing their tissue identity to acute inflammation, which was reversed once the stress was removed.

Material and Methods

Mice

Wild-type (CD45.2), RiboTag (stock number 011029), and Cx3cr1tm2.1(cre/ERT2)Litt/WganJ (stock number 021160) mice were purchased from The Jackson Laboratory (CA). Emr1-Cre (+/+) breeders were a gift from Dr. Klaus Pfeffer (Heinrich Heine University, Düsseldorf, Germany). All mice were housed in a specific pathogen-free environment. The experiments described were performed under Institutional Animal Care and Use Committee approval.

RiboTag immunoprecipitation and RNA isolation

The mice were anesthetized with Avertin (T48402; Sigma,MO). Heart and liver were separated. For hearts only the lower 2/3 were used in analysis. Samples for immunoprecipitation were collected using a 2.0 mm Harris Uni-Core (CA) into 1.6ml tubes immediately placed in ethanol slurry, and stored at −80 C until processing. Polysome immunoprecipitation was performed and mRNA isolated, as described by Sanz et al(14)

Cell isolation and flow cytometry

Cardiac macrophage isolation was performed, as described previously(3). Cells were treated with 1:50 Fc block (BioLegend, CA) for 10 minutes at 4 C followed by staining with antibodies for surface antigens at 4 C for 30 minutes. Intracellular staining was performed on surface-antigen stained cells using Cytofix/Cytoperm (BD Biosciences, CA) according to the manufacturer’s protocol. Stained cells were analyzed using an LSR II (BD Biosciences, CA) and FlowJo software version10.5.0 (FlowJo). The antibodies and clones were purchased from BioLegend or BD Bioscience and used at the indicated dilutions; Ly6C (HK1.4, 1:300), F4/80 (BM8, 1:400), CD11b (M1/70, 1:400), IA/IE (M5/114.15.2, 1:400), Ly6G (1A8, 1:300), Cx3cr1 (SA011S11, 1:300), CD45.2 (104, 1:300), CD11c (N418, 1:300), HA (16B12, 1:300). Dead cells were excluded using Fixable Violet or Far-red dye at 1:1000 dilution (ThermoFisher)

Tamoxifen inductions

Up to 2-days old postnatal mice were gavaged with a single dose of tamoxifen, 1 mg (T5648; Sigma, MO) dissolved in corn oil in a total of 50 ul using a 24G-100 straight 1.25-mm-ball stainless gavage needle (Braintree Scientific, MA). Control mice received only corn oil.

Lipopolysaccharide (LPS) administration

Eight to ten week old male mice were administered with 1mg/kg dose of lipopolysaccharides from Escherichaia coli 0111:B4 (L2630, sigma, MO) diluted in sterile 1xPBS intraperitoneally. Control mice received the same volume of sterile 1xPBS.

Gene expression analysis using qRT-PCR

A selected set of the genes provisionally identified as important by RNAseq gene expression analysis was validated using multiplex qRT-PCR for N=6 independent samples. Briefly, mRNA was isolated as described. cDNA was generated using the QuantiTect Reverse Transcription Kit (Qiagen), and pre-amplified using BIO-X-ACT Short Mix (Bioline) and the TaqMan assays of interest (Thermo Fisher, USA). Microfluidic quantitative RT–PCR was performed on a BioMark HD microfluidic system (Fluidigm Corp, South San Francisco, CA, USA). The Fluidigm Gene Expression software was used to calculate Ct thresholds, and Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) was used to calculate relative expression levels and fold change value. Gene expression was normalized to the house keeping genes Gapdh and Hprt.

RNAseq library preparation and analysis

cDNA libraries were constructed using NuGEN Trio RNA-Seq (Catalog # 0507)(Tecan Genomics, Redwood City, CA) and 2–10 ng of total RNA. Libraries were quantified with NEBNext Library Quant Kit for Illumina (Catalog #E7630, New England BioLabs, Ipswich, MA) before sequencing them on a NextSeq500 (Illumina, San Diego, CA) using NextSeq500 High-Output v2 75 cycles consumables (FC-404–2005)(Illumina). Libraries were sequenced 75 cycles using single ended reads with 8 additional cycles for barcodes. Sequencing reads were converted from BCL to FASTQ format and demultiplexed using Bcl2fastq2 software v2.20.0.422 (Illumina). The first 5 base calls were then trimmed and reads were filtered for quality scores over Q25 on 80% or more base calls on any given read using FASTX-Toolkit v0.0.14 (Hannon Lab). Reads were then aligned to GRCm38/mm10 UCSC mouse genome with TopHat2 v2.1.1 (Trapnell 2012) using -b and -u options . Gene counts were normalized to fragments per kilobase million (FPKM) with Cuffnorm from Cufflinks v2.2.1 using --output-format cuffdiff option. Non-coding genes were not included in any analysis herein. Genes were classified as enriched, in any given tissue, if both IP samples had greater than 5 FPKM and were both 5-fold or greater than their respective input samples. Data from ‘genes.read_group_tracking’ cuffnorm output file were reformatted for subsequent downstream analyses using dplyr v.0.8.5 and tidyr v1.0.2. Principle component analysis (PCA) was calculated with prcomp from the stats package in R v3.6.3 "Holding the Windsock & quot; using center = TRUE and scale = TRUE options. The PCA was graphed using ggfortify v0.4.9 and ggplot2 v3.3.0.

Data analysis

Bar graphs and box plots present data from more than three independent experiments.

Exact numbers and repetitions are noted in the figure legends. No randomization or blinding of the investigator was performed. Groups were compared with paired or unpaired Mann-Whitney U test with p values adjusted for multiple comparisons using Holm-Bonferroni method. Significant p values are represented by asterisks (p <0.05). All statistical analyses were performed in either GraphPad Prism 7 or Microsoft Excel (Microsoft Corporation). Pathway analysis was performed using metascape.org (15). Heatmaps in Figure 4 and Supplementary figure S4 were generated using heatmapper.ca(16). Hierarchical clustering with the log2(gene fold change) was analyzed with Cluster 3.0(17). Genes included were required to be FPKM>5 in both IP replicates in the RNAseq data, and to be 2-fold or greater enriched in two IP replicates of the same tissue samples compared with input. A color scale of up to 10-fold was shown.

Results

Lineage-tracing of early-derived long-lived cardiac macrophages in adult mice

No phenotypic markers clearly identify the early-derived cardiac macrophages or Kupffer cells in adult tissues, and the study of these cells depends on lineage tracing. The chemokine receptor gene Cx3cr1 is expressed in yolk sac macrophage precursors(6, 18), subsets of dendritic cells and monocytes, and subsets of tissue resident macrophages, including cardiac macrophages in mice(3, 8). The Cx3cr1 promoter driven Tamoxifen-inducible Cre recombinase, combined with fluorescent reporters has been used to trace these long-lived tissue-resident macrophages during embryonic life and in the adult(3, 6, 8, 18). This system allowed FACS-based isolation of fluorescent-labeled macrophages and the analysis of their transcriptome. However, there is a risk that cell isolation may induce changes in gene expression, as was recently documented in the case of microglia(19). To avoid this concern we adopted the RiboTag reporter, a tool that facilitates the recovery of ribosome-associated mRNA (ie: the translatome) from specific cells within a complex tissue(14). We previously evaluated gene expression in the fetal-derived subset of Kupffer cells using this strategy(13), and here we apply it also to cardiac macrophages.

Cre recombinase, driven by the Emr1 promoter in mature tissue macrophages, activated the RiboTag transgene, which encodes an influenza hemagglutinin (HA) tag(13). Using this Emr1Cre+/−::RiboTag+/−approach we were able to label ~70% of the cardiac macrophages (Supplementary Fig S1) while both Kupffer cells and microglia were labeled at ~80% and ~67% respectively. While this validated the use of RiboTag as a reporter, only an inducible Cre system allows tracing cells from an early development stage. We selected the Cx3Cr1-CreER because the Cx3Cr1 promoter is active in early progenitor of both Kupffer cells and cardiac macrophages. This approach only makes sense for labeling early progenitor cells since CX3CR1 is expressed on adult cardiac macrophages (3, 8) but not on adult Kupffer cells (6, 13). In the neonatal heart ~90% of the cardiac macrophages expressed CX3CR1 and the expression level was greater compared to monocytes (Supplementary Fig S2, Fig 1 A). We previously reported that in 2 day old neonatal livers, only ~42% of the Kupffer cells express CX3CR1 (13) supporting evidence that CX3CR1 expression wanes rapidly after birth on Kupffer cells (6, 18). Therefore, only in the analysis of early progenitors can the Tamoxifen-inducible Cx3Cr1-CreER be used in Kupffer cell and cardiac macrophage lineage tracing; in adult mice it would also label already-mature macrophages in some tissues including the heart. Thus, we crossed Cx3Cr1-CreER+/+ (20) to RiboTag +/+ (14) mice and generated heterozygous Cx3cr1-CreER+/−::RiboTag+/− mice. We administered Tamoxifen within two days of birth to these neonatal mice to induce RiboTag activation in these early-derived CX3CR1 expressing macrophages and their precursors (Fig 1B).When mice reached adulthood we analyzed heart tissue using intracellular FACS to detect RiboTag labeling via the HA tag (Fig 1C and Supplementary Fig S2).

Figure 1: Lineage-tracing of early-derived long-lived cardiac macrophages in adult mice.

Figure 1:

Open in a new tab

A. FACS analysis of CX3CR1 expression levels in monocytes and cardiac macrophages in neonatal hearts B. Diagram Illustrates the lineage tracing of early-derived cardiac macrophages using Tamoxifen inducible labeling in Cx3Cr1-CreER+/−::RiboTag+/− mice. C-E. Intracellular FACS analysis of, percent HA(RiboTag)-labeled, early-derived cardiac macrophages compared to corn-oil controls (C), surface expression of MHC Class II (1A/1E), TIMD4, and CCR2 on early-derived cardiac macrophages (E), percent early-derived macrophages in liver and the heart (D) in 8–10 week old mice. Data represent three independent experiments with n = 4 mice/group. (A, C-E) boxplots with boxes representing interquartile range, median indicated with a line within the box and whiskers indicating minimum to maximum . ****p ≤0.0001; ns:Non significant, Mann-Whitney U test; MFI-Mean fluorescent intensity; MQs: Macrophages; HA,hemagglutinin; Tam: Tamoxifen

Within the CD45+ hematopoietic cells only a subset of macrophages expressed the HA tag, representing the early-derived long-lived cardiac macrophages (Supplementary Fig S2 C). Among cardiac macrophages, this subset accounted for ~40% in 8–12 weeks adult male mice (Fig 1C). Thus only a subset of the CX3CR1 expressing neonatal cardiac macrophages survived into the adult stage. There were no other RiboTag labeled cells in the single cell suspensions, indicating that any other cells that were labeled during the neonatal period did not survive in the adult (Supplementary Fig S2 D). About ~80% of these early-derived cells expressed CX3CR1 and MHC class II on the surface mirroring previous reports (Fig 1D)(3, 8, 10). However only about ~20% of the early-derived cardiac macrophages expressed TIMD4 (Fig 1D), a surface receptor for apoptotic cell engulfment, that is previously shown to express on long-lived self-renewing cardiac macrophages(10). These disparities suggest that not all long-lived cardiac macrophages in the adult heart originated from early precursors, and others have shown that blood monocytes has the capacity to seed adult tissues(21) and become long-lived self-renewing macrophages(22). In the same mice the frequency of HA-tagged, early-derived Kupffer cells was 37% in agreement with our previous report (Fig 1E)(13).

Collectively, using the Cx3cr1 inducible fate mapping system combined with RiboTag reporter we show that about 40% of the cardiac macrophages and 37% of the Kupffer cells in the adult mice derive from cells that were clonally marked through their expression of Cx3Cr1 in the neonates, and represent early-derived macrophage subsets in each tissue.

Translatome analysis of early-derived cardiac and hepatic macrophage subsets at steady state

Transcriptome comparisons of adult tissue resident macrophages have clearly indicated the tissue specific gene expression of macrophages(23, 24). A study comparing the transcriptome of sorted cells and translatome of microglia concluded that the transcriptome analysis contained genes that were artifacts of the cell isolation process and contaminations, while RiboTag-based translatome analysis provided an enriched gene expression profile, specific for microglia(19). Thus, we sought to understand the, early-derived cardiac and Kupffer cell specific translatomes at steady state in adult mice.

Once neonatally Tamoxifen treated Cx3cr1-CreER+/−::RiboTag+/− mice reached adulthood we harvested heart and liver tissues and performed ribosome immunoprecipitation (IP) to isolate actively translating mRNA from the RiboTag-labeled early-derived macrophages. Using RNA sequencing we compared all isolated mRNA in total tissue versus the IP fractions representing early-derived macrophages (Fig 2A). The enriched genes in each early-derived macrophage subset included genes expressed in common as well as tissue specific genes (Fig 2B, D). Examples of shared genes included canonical macrophage genes (Adgre1, Fcgr1, Itgam), Toll-like receptor signaling pathway components (Tlr2, Tlr4, Cd180, Ly86), phagocytosis-associated genes (Cd68, Mrc1) and antigen presentation-associated molecules (H2-Aa, H2-Ab1, Cd40, Cd86) indicating common functions of these early derived macrophages in each tissue.

Figure 2: Translatome analysis using RNAseq of early-derived cardiac and hepatic macrophages (eCMQ and eKC) at steady state.

Figure 2:

Open in a new tab

A. Heat maps indicating the 8-fold enriched genes in Ip fraction representing early-derived macrophage subset (red) compared to Input representing the total tissue mRNA in RNAseq analysis. Gene criteria: FPKM≥5 in at least one sample input or Ip B. Heat map indicates the 8-fold or more enriched genes in eKC and eCMQ in RNAseq analysis. C. Gene ontology and Pathway analysis of genes that had FPKM > 5 in the IP of at least one group and showed 10 fold difference between eKC and eCMQ IPs, using metascape.org (15). D. venn-digram comparing the 5 fold- enriched genes and representative genes specific or common to each macrophage subset in RNASeq dataset. Bar graphs indicate the RT-qPCR validation of the genes. RT-qPCR was performed on N=6 independent samples. A-C: data represents N=2 mice per group. D: Bars represent two independent experiments for total of N=6 mice. bars indicate the mean ± SD, *p ≤0.05,**p ≤0.01, ,***p ≤0.001, ****p ≤0.0001. Mann-Whitney U test with p values adjusted for multiple comparisons using Holm-Bonferroni method. APC: Antigen presenting cells

Early-derived cardiac macrophages strongly expressed chemokine genes Ccl7, Ccl9, Ccl12, and Cx3cr1 suggesting a role in steady state myeloid cell recruitment, and S100a4 and Mmp13 providing evidence for a regenerative function. Early-derived Kupffer cells strongly expressed Cxcl13, a chemokine for B cell recruitment(25) and the gene Ptgs1, important in immunosuppression. Pathway analysis of the genes that were ten-fold enriched in each early-derived subset suggested tissue repair and vascular development functions for cardiac macrophages and innate immune defense for Kupffer cells (Fig 2C). Collectively these gene expression data suggest the early-derived subsets serve the tissue specific functions previously reported for hepatic and cardiac macrophages(1, 2). Furthermore our data highlights that these early-derived macrophages have similar functions across tissues, despite the microenvironment differences between tissues.

Early-derived cardiac macrophages and Kupffer cells reversibly respond to LPS-induced acute inflammation

The presence of adult tissue macrophages that were derived from cells expressing Cx3cr1 in the neonatal period argues that these are resilient cells. The question arises whether these cells are resilient because they do not participate in immune events such as inflammation, or whether they are able to respond but then revert to their prototypical form. We sought evidence for such tissue macrophage plasticity in an acute systemic inflammation model.

We administered bacterial LPS intraperitoneally to mice bearing labeled early-derived macrophages and harvested tissue for analysis of the translatome 4 hours and 2 weeks post stimulation (Fig 3A). The 4-hour and the 2-week time points were chosen to highlight the rapid onset of inflammation, and the state of the macrophage populations post-resolution (Supplementary Fig S3 A). First we analyzed the early immune response in these macrophages by comparing those genes enriched five fold or more in each subset. Of the genes that were enriched in either tissue, about one third were shared while each tissue had about one third that were exclusively enriched (Fig 3B). As expected, pathway analysis of genes enriched at the early time point highlighted inflammatory pathways with significant hits on gene ontology (GO) terms like, “response to molecule of bacterial origin”, “Tumor necrosis factor production” and “myeloid leukocyte activation” (Fig 3C). Genes exclusively enriched in one or the other subset suggested that the pathways leading to these early immune responses are not identical, though it is not clear whether this is due to differences between the macrophages, or secondary effects due to other cells in the tissue.

Figure 3: Early-derived cardiac macrophages and Kupffer cells (eCMQ and eKC) reversibly respond to LPS-induced acute inflammation.

Figure 3:

Open in a new tab

A. Diagram illustrates the reversible systemic inflammation model B. Venn-diagram comparing common vs tissue specific genes among the 5-fold enriched genes in each macrophage subset at 4 hours post LPS stimulation. Bar graphs indicate RT-qPCR validation of several commonly upregulated genes compared to steady state for N=6 independent samples. C. Gene ontology and Pathway analysis of genes that are commonly and tissue specifically enriched at 4 hours post-LPS stimulation. D. Heat map comparing the expression of tissue specific genes at steady state, 4 hours and at 2 weeks for both macrophage subsets. E. Venn diagrams showing 5 fold or more enriched genes at steady state, 4 hours and 2 in each early-derived macrophage subset. Gene criteria: Greater than 5 FPKM in both Ip samples, and both Ip samples 5-fold greater than their respective input samples. F. Principle component analysis demonstrating the majority of the variation in the dataset is due to LPS treatment (PC1) while the second most can be attributed to the difference in tissues (PC2). IP samples only in graph. B: Bars represent two independent experiments for total of N=6 mice. Bars indicate the mean ± SD, *p ≤0.05,**p ≤0.01. Mann-Whitney U test with p values adjusted for multiple comparisons using Holm-Bonferroni method. APC: Antigen presenting cells

The common LPS-responsive genes in the early-derived Kupffer cells and cardiac macrophages included MAP-kinase pathway components (Ptk2b, Rasgrp1), cytokines and chemokines (Il1b, Tnf, Il18, Ccl3, Ccl5 and Il10), their receptors (Il4ra, Csf1r, Ccr5), and components of the TLR4 pathway (Ly96, Cd14, Acod1, Sash1) that engages LPS (Fig 3B and Supplementary Fig S3 B). The up-regulation of Il10 and Il4ra genes, that produces IL10 (an anti-inflammatory cytokine) and IL4ra (a receptor for IL4 and IL13 anti-inflammatory cytokines) together with pro-inflammatory genes (Il1a, Il1b,Tnf, Il8) suggest that these early-derived macrophages activate both positive and negative regulators of inflammation. This contrasts with the LPS-induced response of in vitro cultured macrophages, which was exclusively pro-inflammatory(26). These early-derived macrophages therefore appear to employ both pro- and anti-inflammatory circuits from the beginning to contain the injury in vivo, as previously suggested by others(27).

Next we analyzed the post-resolution translatome at 2 weeks after LPS administration, to document the changes in these macrophages after recovery. Heat-maps (Fig 3D) and principal component analysis (Fig 3F and Supplementary Fig S3 C), demonstrate marked changes in the translatome at 4 hours and 2 weeks post-LPS in both subsets with a return to near baseline by 2 weeks. Comparison of those genes enriched 5-fold or more in each subset at steady state, 4 hours and 2 weeks indicated that 12 (2%) and 128 (25%) of the genes remained differentially expressed in early-derived macrophage subsets in liver and heart respectively (Fig 3 E). The handful of genes that remained different from baseline at 2 weeks within the early-derived Kupffer cell subset suggests a near complete reversal to baseline for these cells while a higher number of genes that remained differentially expressed for cardiac macrophage subset by 2 weeks indicate that, in these cells, reversion was slower.

Transcription factor dynamics of the early-derived cardiac macrophages and Kupffer cells during acute inflammation and post-resolution

So far our data indicated that the early-derived cardiac macrophages and Kupffer cells mounted a temporary response to systemic LPS stimuli and regained steady state gene expression patterns, post resolution. To understand how this temporary activation state was achieved we focused on the expression of transcription factors.

First we identified transcription factor genes in our RNAseq dataset that were enriched five fold or more in early-derived tissue macrophage subsets, and found both common (Spi1, Ikzf1, pou2f2, Irf5, Irf8, Batf) and tissue specific factors (Fig 4A, Supplementary Fig S4 A). We identified known tissue specific factors (in Kupffer cells:Nr1h3(23), and in cardiac macrophages:Irf4(27)) as well as new candidates in each subset. These included Nfatc2, Phf20, Maf, Runx1 in cardiac macrophages and Cebpb, Hic1, Hey1 (Supplementary Fig S4D) in Kupffer cells of early-origin. These data further document tissue specific gene expression in each subset. Next, we analyzed the transcription factor dynamics for each early-derived macrophage subset over the time course of the LPS response. Both early-derived cardiac macrophages and Kupffer cells showed increased expression of genes encoding transcription factors involved in the inflammatory process. These included pro-inflammatory factors such as Nfkb1, Nfkb2, as well as anti-inflammatory factors such as Nfkbiz (Supplementary Fig S4 C). The upregulation of gene expression for both pro-and anti- inflammatory transcription factors as early as 4 hours further supported the conclusion that pro- and anti- inflammatory responses are not mutually exclusive in these early-derived macrophages. Furthermore we found that by 4-hours post-LPS stimulation, canonical myeloid transcription factors, such as Spi1(28) and Zeb2(29) increased their expression while tissue specific transcription factors in each subset, for example Nr1h3 in Kupffer cells and Nfatc2 in cardiac macrophages, downregulated their expression (Fig 4B). All these transcription factors returned to steady state levels by 2-weeks post LPS-stimulation (Supplementary Fig S4D). Thus the changes in transcription factor expression allowed these cells to maintain myeloid identity, while temporarily renouncing their tissue identity.

During systemic stresses such as ischemia and stroke, long-lived cardiac macrophages and Kupffer cells have been shown to replenish their populations through selfrenewal(10, 11). Thus we analyzed for the expression of transcription factors that are associated with self-renewal in early-derived cardiac macrophages and Kupffer cells upon LPS stimulation. Gene expression of transcription factors that inhibit self-renewal (Mafb(30), Zbtb4(31)) were downregulated by 4 hours post-LPS, suggesting that these cells are preparing to self-renew as a consequence of the inflammatory stress (Fig 4C). Collectively, the changes in transcription factor expression during acute inflammation revealed a previously unknown feature of these early-derived long-lived macrophage subsets; i.e in LPS-induced acute systemic inflammation, these macrophages renounce their tissue identity temporarily to combat the stress and regain their tissue specific identity once the stress has resolved.

Figure 4: Transcription factor dynamics of the early-derived cardiac macrophages and Kupffer cells (eCMQ and eKC) during acute inflammation and post-resolution.

Figure 4:

Open in a new tab

A. Heat map comparing genes coding transcription factors specific to each macrophage subset at steady state in RNAseq analysis for N=2 mice. Bar graphs indicate RT-qPCR validation of selected enriched genes in each subset for N=6 independent samples. B. Heat maps indicating the changes in tissue specific and common transcription factor gene expression during acute inflammation and post-resolution due to LPS stimulation in RNAseq analysis for N=2 mice. Bar graphs indicate the RT-qPCR validation of selected genes for N=6 independent samples. A-B: Heatmaps indicate unsupervised clustering of genes >5 fold enriched in each subset. Values are Raw z-scores. Brown text: enriched in eKCs, Purple text: Enriched in eCMQs. Black text: Common TFs to both subsets C. Bar graphs indicating down regulation of self-renewal inhibitory transcription factor genes by 4 hours in each macrophage subset. Bars represent two independent experiments for total of N=6 mice. A-C: Bars indicate the mean ± SD, *p ≤0.05,**p ≤0.01. Mann-Whitney U test with p values adjusted for multiple comparisons using Holm-Bonferroni method. A-B: Heat maps indicate RNAseq analysis for N=2 mice per each condition. LPS: Lipopolysaccharide.

Discussion

Here, we sought to understand how early-derived macrophages survive acute stress, and sustain themselves for a long period of time. Using lineage tracing and translatome analysis we found that the early derived cardiac macrophages and Kupffer cells assume a temporary activation state in response to acute inflammation. Gene expression analysis of myeloid and tissue specific transcription factors during this temporary activation indicated that these macrophages renounced their tissue identity but strengthened their myeloid identity. Once the stress was resolved, they resumed steady state gene expression. The temporary nature of these gene expression changes suggests that these cells resist permanent polarization, and support a state of plasticity at baseline contributing to resilience over time.

While our study provides gene expression-based evidence for the potential of these early-derived macrophage subsets to completely revert to baseline, analysis of epigenetic modifications upon LPS stimulation and resolution are explorations for the future. Repeated LPS stimulations of microglia, which are early-derived resident macrophages of the brain, has shown long term epigenetic changes post LPS stimulation, described as innate immune memory(32). Further research will determine whether such training will also benefit the resilience of these long-lasting early-derived macrophages.

While our data provide insight into how these early-derived macrophage subsets maintain resilience, we cannot establish whether this is specific for macrophages of fetal origin. Both cardiac macrophage and Kupffer cell compartments can progressively replenish from adult monocyte-derived macrophages due to age or inflammation over the course of life (8, 11). These monocyte-derived macrophages also appear to have the capacity to self-renew, live long and prosper within the tissues (22). Whether these adult monocyte derived long-lived macrophages are also capable of such flexibility in their gene expression will be a question for the future.

During this early activation phase they further down regulated transcription factors that inhibit self-renewal, suggesting preparation for self-renewal in the recovery period. While our study revealed a temporary activation state in these macrophages in response to acute inflammation, followed by an almost complete reversion to their initial state, further studies are needed to assess whether and how the temporary renunciation of tissue specificity is mechanistically linked to imminent self-renewal.

At steady state, a third of the genes enriched in each early-derived macrophage subset were specific for their host tissue, supporting previous reports. We also found that a third of the genes were commonly expressed in both macrophage subsets arguing for common functions despite the tissue differences. Translatome analysis following LPS-induction further revealed both common and subset specific changes in gene expression, as previously documented(11, 33). The tissue specific responses support the notion that tissue macrophages reside in unique niches, from which context they respond to changes in their microenvironment. Further experiments are needed to explore whether the location independent responses are determined by ontogeny. The Cx3cr1 lineage tracing we employed labeled fetal-derived, Cx3cr1 expressing cells at birth, but we cannot confirm all Cx3cr1 expressing cells come from the same precursors.

Finally, how these resilient early-derived macrophages, which constitute at least a third of the population, respond to sterile or chronic injury or repeated acute inflammatory stimuli needs further explorations. Do they revert to a state of baseline plasticity after each stress condition or do they succumbed to stress induced death leading to irreversible tissue damage? Answering these questions will provide insights into how we can target these tissue resident macrophages in diseases.

Supplementary Material

1

Key points.

  • LPS induces a reversible non classical immune response in early-derived macrophages

  • LPS temporarily down regulated tissue specific transcription factor expression

Acknowledgements

We thank the Department of Pathology Flow Core for technical support, and the Institute of Systems Biology, Seattle,WA for the use of the Fluidigm Biomark instrument.

Foot notes

This work was supported by the American Heart Association (award 17PRE33410275 to RS), the National Institutes of Health (grant 1R21AI114827 to INC).

Footnotes

Disclosures

The authors have no financial conflicts of interest.

References

  • 1.Swirski FK, and Nahrendorf M. 2018. Cardioimmunology: the immune system in cardiac homeostasis and disease. Nat Rev Immunol 18: 733–744. [DOI] [PubMed] [Google Scholar]
  • 2.Krenkel O, and Tacke F. 2017. Liver macrophages in tissue homeostasis and disease. Nat Rev Immunol 17: 306–321. [DOI] [PubMed] [Google Scholar]
  • 3.Epelman S, Lavine KJ, Beaudin AE, Sojka DK, Carrero JA, Calderon B, Brija T, Gautier EL, Ivanov S, Satpathy AT, Schilling JD, Schwendener R, Sergin I, Razani B, Forsberg EC, Yokoyama WM, Unanue ER, Colonna M, Randolph GJ, and Mann DL. 2014. Embryonic and adult-derived resident cardiac macrophages are maintained through distinct mechanisms at steady state and during inflammation. Immunity 40: 91–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Schulz C, Gomez Perdiguero E, Chorro L, Szabo-Rogers H, Cagnard N, Kierdorf K, Prinz M, Wu B, Jacobsen SE, Pollard JW, Frampton J, Liu KJ, and Geissmann F. 2012. A lineage of myeloid cells independent of Myb and hematopoietic stem cells. Science 336: 86–90. [DOI] [PubMed] [Google Scholar]
  • 5.Gomez Perdiguero E, Klapproth K, Schulz C, Busch K, Azzoni E, Crozet L, Garner H, Trouillet C, de Bruijn MF, Geissmann F, and Rodewald HR. 2015. Tissue-resident macrophages originate from yolk-sac-derived erythro-myeloid progenitors. Nature 518: 547–551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Yona S, Kim KW, Wolf Y, Mildner A, Varol D, Breker M, Strauss-Ayali D, Viukov S, Guilliams M, Misharin A, Hume DA, Perlman H, Malissen B, Zelzer E, and Jung S. 2013. Fate mapping reveals origins and dynamics of monocytes and tissue macrophages under homeostasis. Immunity 38: 79–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Sheng J, Ruedl C, and Karjalainen K. 2015. Most Tissue-Resident Macrophages Except Microglia Are Derived from Fetal Hematopoietic Stem Cells. Immunity 43: 382–393. [DOI] [PubMed] [Google Scholar]
  • 8.Molawi K, Wolf Y, Kandalla PK, Favret J, Hagemeyer N, Frenzel K, Pinto AR, Klapproth K, Henri S, Malissen B, Rodewald HR, Rosenthal NA, Bajenoff M, Prinz M, Jung S, and Sieweke MH. 2014. Progressive replacement of embryo-derived cardiac macrophages with age. J Exp Med 211: 2151–2158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Hoeffel G, Chen J, Lavin Y, Low D, Almeida FF, See P, Beaudin AE, Lum J, Low I, Forsberg EC, Poidinger M, Zolezzi F, Larbi A, Ng LG, Chan JK, Greter M, Becher B, Samokhvalov IM, Merad M, and Ginhoux F. 2015. C-Myb(+) erythro-myeloid progenitor-derived fetal monocytes give rise to adult tissue-resident macrophages. Immunity 42: 665–678. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Dick SA, Macklin JA, Nejat S, Momen A, Clemente-Casares X, Althagafi MG, Chen J, Kantores C, Hosseinzadeh S, Aronoff L, Wong A, Zaman R, Barbu I, Besla R, Lavine KJ, Razani B, Ginhoux F, Husain M, Cybulsky MI, Robbins CS, and Epelman S. 2019. Self-renewing resident cardiac macrophages limit adverse remodeling following myocardial infarction. Nat Immunol 20: 29–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Hoyer FF, Naxerova K, Schloss MJ, Hulsmans M, Nair AV, Dutta P, Calcagno DM, Herisson F, Anzai A, Sun Y, Wojtkiewicz G, Rohde D, Frodermann V, Vandoorne K, Courties G, Iwamoto Y, Garris CS, Williams DL, Breton S, Brown D, Whalen M, Libby P, Pittet MJ, King KR, Weissleder R, Swirski FK, and Nahrendorf M. 2019. Tissue-Specific Macrophage Responses to Remote Injury Impact the Outcome of Subsequent Local Immune Challenge. Immunity 51: 899–914 e897. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Lavine KJ, Epelman S, Uchida K, Weber KJ, Nichols CG, Schilling JD, Ornitz DM, Randolph GJ, and Mann DL. 2014. Distinct macrophage lineages contribute to disparate patterns of cardiac recovery and remodeling in the neonatal and adult heart. Proc Natl Acad Sci U S A 111: 16029–16034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Soysa R, Lampert S, Yuen S, Douglass AN, Li W, Pfeffer K, and Crispe IN. 2019. Fetal origin confers radio-resistance on liver macrophages via p21(cip1/WAF1). J Hepatol. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Sanz E, Yang L, Su T, Morris DR, McKnight GS, and Amieux PS. 2009. Cell-type-specific isolation of ribosome-associated mRNA from complex tissues. Proc Natl Acad Sci U S A 106: 13939–13944. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Zhou Y, Zhou B, Pache L, Chang M, Khodabakhshi AH, Tanaseichuk O, Benner C, and Chanda SK. 2019. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun 10: 1523. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Babicki S, Arndt D, Marcu A, Liang Y, Grant JR, Maciejewski A, and Wishart DS. 2016. Heatmapper: web-enabled heat mapping for all. Nucleic Acids Res 44: W147–153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Eisen MB, Spellman PT, Brown PO, and Botstein D. 1998. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 95: 14863–14868. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Hagemeyer N, Kierdorf K, Frenzel K, Xue J, Ringelhan M, Abdullah Z, Godin I, Wieghofer P, Costa Jordao MJ, Ulas T, Yorgancioglu G, Rosenbauer F, Knolle PA, Heikenwalder M, Schultze JL, and Prinz M. 2016. Transcriptome-based profiling of yolk sac-derived macrophages reveals a role for Irf8 in macrophage maturation. EMBO J 35: 1730–1744. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Haimon Z, Volaski A, Orthgiess J, Boura-Halfon S, Varol D, Shemer A, Yona S, Zuckerman B, David E, Chappell-Maor L, Bechmann I, Gericke M, Ulitsky I, and Jung S. 2018. Re-evaluating microglia expression profiles using RiboTag and cell isolation strategies. Nat Immunol 19: 636–644. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Parkhurst CN, Yang G, Ninan I, Savas JN, Yates JR 3rd, Lafaille JJ, Hempstead BL, Littman DR, and Gan WB. 2013. Microglia promote learning-dependent synapse formation through brain-derived neurotrophic factor. Cell 155: 1596–1609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Bain CC, Bravo-Blas A, Scott CL, Perdiguero EG, Geissmann F, Henri S, Malissen B, Osborne LC, Artis D, and Mowat AM. 2014. Constant replenishment from circulating monocytes maintains the macrophage pool in the intestine of adult mice. Nat Immunol 15: 929–937. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Scott CL, Zheng F, De Baetselier P, Martens L, Saeys Y, De Prijck S, Lippens S, Abels C, Schoonooghe S, Raes G, Devoogdt N, Lambrecht BN, Beschin A, and Guilliams M. 2016. Bone marrow-derived monocytes give rise to self-renewing and fully differentiated Kupffer cells. Nat Commun 7: 10321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Lavin Y, Winter D, Blecher-Gonen R, David E, Keren-Shaul H, Merad M, Jung S, and Amit I. 2014. Tissue-resident macrophage enhancer landscapes are shaped by the local microenvironment. Cell 159: 1312–1326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Gosselin D, Link VM, Romanoski CE, Fonseca GJ, Eichenfield DZ, Spann NJ, Stender JD, Chun HB, Garner H, Geissmann F, and Glass CK. 2014. Environment drives selection and function of enhancers controlling tissue-specific macrophage identities. Cell 159: 1327–1340. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Ansel KM, Harris RB, and Cyster JG. 2002. CXCL13 is required for B1 cell homing, natural antibody production, and body cavity immunity. Immunity 16: 67–76. [DOI] [PubMed] [Google Scholar]
  • 26.Das A, Yang CS, Arifuzzaman S, Kim S, Kim SY, Jung KH, Lee YS, and Chai YG. 2018. High-Resolution Mapping and Dynamics of the Transcriptome, Transcription Factors, and Transcription Co-Factor Networks in Classically and Alternatively Activated Macrophages. Front Immunol 9: 22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Walter W, Alonso-Herranz L, Trappetti V, Crespo I, Ibberson M, Cedenilla M, Karaszewska A, Nunez V, Xenarios I, Arroyo AG, Sanchez-Cabo F, and Ricote M. 2018. Deciphering the Dynamic Transcriptional and Post-transcriptional Networks of Macrophages in the Healthy Heart and after Myocardial Injury. Cell Rep 23: 622–636. [DOI] [PubMed] [Google Scholar]
  • 28.Kueh HY, Champhekar A, Nutt SL, Elowitz MB, and Rothenberg EV. 2013. Positive feedback between PU.1 and the cell cycle controls myeloid differentiation. Science 341: 670–673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Scott CL, T’Jonck W, Martens L, Todorov H, Sichien D, Soen B, Bonnardel J, De Prijck S, Vandamme N, Cannoodt R, Saelens W, Vanneste B, Toussaint W, De Bleser P, Takahashi N, Vandenabeele P, Henri S, Pridans C, Hume DA, Lambrecht BN, De Baetselier P, Milling SWF, Van Ginderachter JA, Malissen B, Berx G, Beschin A, Saeys Y, and Guilliams M. 2018. The Transcription Factor ZEB2 Is Required to Maintain the Tissue-Specific Identities of Macrophages. Immunity 49: 312–325 e315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Soucie EL, Weng Z, Geirsdottir L, Molawi K, Maurizio J, Fenouil R, Mossadegh-Keller N, Gimenez G, VanHille L, Beniazza M, Favret J, Berruyer C, Perrin P, Hacohen N, Andrau JC, Ferrier P, Dubreuil P, Sidow A, and Sieweke MH. 2016. Lineage-specific enhancers activate self-renewal genes in macrophages and embryonic stem cells. Science 351: aad5510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Yu Y, Shang R, Chen Y, Li J, Liang Z, Hu J, Liu K, and Chen C. 2018. Tumor suppressive ZBTB4 inhibits cell growth by regulating cell cycle progression and apoptosis in Ewing sarcoma. Biomed Pharmacother 100: 108–115. [DOI] [PubMed] [Google Scholar]
  • 32.Wendeln AC, Degenhardt K, Kaurani L, Gertig M, Ulas T, Jain G, Wagner J, Hasler LM, Wild K, Skodras A, Blank T, Staszewski O, Datta M, Centeno TP, Capece V, Islam MR, Kerimoglu C, Staufenbiel M, Schultze JL, Beyer M, Prinz M, Jucker M, Fischer A, and Neher JJ. 2018. Innate immune memory in the brain shapes neurological disease hallmarks. Nature 556: 332–338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Sajti E, Link VM, Ouyang Z, Spann NJ, Westin E, Romanoski CE, Fonseca GJ, Prince LS, and Glass CK. 2020. Transcriptomic and epigenetic mechanisms underlying myeloid diversity in the lung. Nat Immunol 21: 221–231. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1704536-supplement-1.pdf (847KB, pdf)

RESOURCES