Identification and Characterization of Alveolar and Recruited Lung Macrophages during Acute Lung Inflammation

Pulmonary macrophages represent a critical component of the innate immune system and form a heterogeneous population of immune cells that patrol the lungs, phagocytose particulates and microbes, and activate the inflammatory cascade when necessary. In addition to playing key roles in host defense and initiating the innate immune response, macrophages also participate in resolving inflammation and repairing the lung after injury (1, 2). Pulmonary macrophages are highly plastic and respond to the local microenvironment by adopting dynamic, multidimensional phenotypic profiles (3).

During homeostasis, two main subsets of macrophages coexist in the lungs. Alveolar macrophages (AMs) are considered tissue-resident macrophages that arise from fetal progenitors and are a long-lived self-renewing population with little or no replenishment from circulating bone marrow (BM)–derived monocytes in the steady state (4, 5). In comparison, BM-derived monocytes are recruited to the lungs in large numbers and differentiate into BM-derived macrophages that share a cell surface marker profile with interstitial macrophages (IMs). Although these BM-derived macrophages can amplify inflammation, they also participate in resolving inflammation and restoring homeostasis (6–8). Multiple studies have demonstrated phenotypic characterization of macrophages in normal lungs by flow cytometry (9, 10); however, specific identification of macrophage populations is limited during lung inflammation due to altered expression of cell surface markers (11). To date, identification and characterization of macrophage subpopulations at different stages of lung inflammation remain challenging.

In this study, we report a new method using flow cytometric strategies coupled with in vivo labeling of macrophages with PKH lipophilic dyes, including PKH26 (red fluorescence dye) and PKH67 (green fluorescence dye), to identify resident AMs and BM-derived macrophages. These dyes label cell membranes by intercalating their aliphatic portion into the exposed lipid bilayer (12). After intratracheal (i.t.) delivery of Escherichia coli LPS, we sorted macrophage populations based on the labeling of PKH dyes and determined transcriptional signatures during distinct time points, including initiation of inflammation (day 1), peak neutrophilic inflammation (day 3), and resolution of inflammation (day 7). Our results indicate that this method accurately distinguishes and phenotypes macrophage subsets during the course of LPS-induced inflammation and should prove helpful for investigating the roles of macrophage subsets during lung inflammation, infection, and injury.

C57BL/6 mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Animals were maintained under pathogen-free conditions and used for experiments between 8 and 12 wk of age. All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committee.

LPS (serotype 055: B5, Sigma-Aldrich, catalog no. L2880) was diluted in sterile PBS and injected (3 μg/g body weight) i.t. using an intubation procedure as previously described (13). A 5 μM solution of PKH67 fluorescent cell linker (catalog no. P7333, Sigma-Aldrich) in sterile PBS was injected (100 µl) by the i.t. route using an intubation procedure.

Mice were anesthetized with isoflurane by inhalation and prepared for surgery by sterilizing the surgical sites. A small incision of ∼0.5 cm was made in the area to expose the tibia. A 28G needle was inserted through the joint surface of the tibia and into the BM cavity, followed by injection of 20 μl (5 μM solution) of PKH26 fluorescent cell linker (catalog no. P9691, Sigma-Aldrich) into the tibia BM cavity using a syringe with a 28G needle. The incision was closed with sterile 5-0 sutures.

RNA was isolated using an RNeasy mini kit (catalog no. 741014, Qiagen), digested with DNase (catalog no. AM1906, Thermo Fisher Scientific), and reverse transcribed into cDNA by an iScript cDNA synthesis kit (catalog no. 1708891, Bio-Rad). Real-time PCR was performed using iQ SYBR Green supermix (catalog no. 4367659, Applied Biosystems). The relative amount of each mRNA species was calculated based on its threshold cycle (Ct) and normalized to GAPDH expression. The primer sequences were as follows: GAPDH (5′-TGCACCACCAACTGCTTAGC-3′ and 5′-GGCATGCACTGTGGTCATGAG-3′); CCR2 (5′-GTTACCTCAGTTCATCCA-3′ and 5′-CAAGGCTCACCATCATGGTAGTC-3′); IL-1β (5′-GTTACCTCAGTTCATCCA-3′ and 5′-CAAGGCTCACCATCATCGTAGTC-3′); CXCL1 (5′-CCGAAGTCATAGCCACACTCAA-3′ and 5′-GCAGTCTGTCTTCTTTCTCCGTTAC-3′); CXCL3 (5′-AAGTGTGAATGTAAGGTCCCC-3′ and 5′- GTGCTCCCCTTGTTCAGTATC-3′); IL-6 (5′-CCACTCACCTCTTCAGAACG-3′ and 5′-CATCTTTGGAAGGTTCAGGTTG-3′); MerTK (5′-CGCCAAGGCCGCATT-3′ and 5′-TCGGTCCGCCAGGCT-3′); and Ax1 (5′-TCATGTGAAGCCCACAATGC-3′ and 5′-GGAGCACTGTGATGGTGGCT-3′).

Bronchoalveolar lavage (BAL) was performed by instilling five 1-ml aliquots of sterile PBS through a tracheotomy tube. To obtain single-cell suspensions from lung tissue, lungs were perfused with sterile PBS, removed en bloc, and perfused lungs were digested in RPMI 1640 medium containing collagenase XI (0.7 mg/ml; Sigma-Aldrich) and type IV bovine pancreatic DNase (30 µg/ml; Sigma-Aldrich). RBCs were lysed with RBC lysis buffer (BioLegend) as described elsewhere (14).

Single-cell suspensions were incubated with a Fc receptor block (catalog no. 553141, BD Biosciences) to reduce nonspecific Ab binding. The panel of Abs used in these experiments included CD45-Brilliant Violet 650 (catalog no. 103151), CD11b-allophycocyanin (catalog no. 101212), CD11c-PE/Cy7 (catalog no. 117317), Ly6G-allophycocyanin/Cy7 (catalog no. 127623), F4/80-PE/Cy5 (catalog no. 123111), Siglec-F-AF647 (catalog no. 142407), MerTK-Brilliant Violet 605 (catalog no. 151517), CD11c-AF700 (catalog no. 117320), CD11b-PE/Cy7 (catalog no. 101216), MHC class II-FITC (catalog no. 107605), and CD64-allophycocyanin (catalog no. 139306) from BioLegend; and Siglec-F-PE (catalog no. 552126) from BD Biosciences. Dead cells were excluded using DAPI (catalog no. MBD0015, Sigma-Aldrich). Flow cytometry was performed using BD LSR II and BD FACSAria III flow cytometers (BD Biosciences), and data were analyzed with FlowJo software.

Total RNA was isolated from PKH26⁺ and PKH67⁺ macrophages using an RNeasy mini kit (catalog no. 741014, Qiagen), followed by DNA digestion (catalog no. AM1906, Thermo Fisher Scientific). Quality checks of RNA, mRNA enrichment, and cDNA library preparation utilizing stranded mRNA (poly(A) selected) were conducted at Vanderbilt Technologies for Advanced Genomics. RNA sequencing (RNA-seq) was performed on an Illumina NovaSeq 6000 system with a paired-end mRNA library prep, PE-150, with 30 million reads. Fastq files were checked for quality issues using FastQC (v0.11.9) and trimmed with Trimmomatic (v0.39) to remove low-quality bases with default options. The quality of reads was reassessed with FastQC after this step to confirm quality improvements. Reads from fastq files were aligned to the mouse reference genome (GRCm38) using HISAT2 (v2.1.0) with default parameters. The number of reads per gene was performed using featureCounts (v2.0.0). Differentially expressed genes (DEGs) were identified using DEseq2 (v1.34) between groups with the absolute value of log₂ fold change >1 and adjusted p value <0.1. Upregulated and downregulated DEGs were separately tested for enrichment in KEGG pathways (adjusted p value <0.1) using the version of DAVID Knowledgebase (v6.8). The STRING Database (version 11.5) was used to analyze potential interactions among DEGs corresponding to peroxisome activity. All datasets have been deposited in the National Center for Biotechnology Information/Gene Expression Omnibus under accession number GSE225406.

Mouse neutrophils were isolated by density-gradient Histopaque-1077 and Histopaque-1119 from mouse peripheral blood. Neutrophils were then labeled with bisBenzimide H 33342 (Sigma-Aldrich), a fluorescent nuclear dye (10 μg/ml, 30 min, 37°C), and cultured overnight (5 × 10⁶ cells/ml in PBS). PKH67-labeled resident AMs and PKH26-labeled BM-derived macrophages were collect by FACS. Resident and BM-derived macrophages were plated onto 96-well plates at 5 × 10⁴ cells/well followed by adding labeled apoptotic neutrophils (3:1 ratio, neutrophils to macrophages) for 1 h at 37°C as described (15). After incubation, the cells were washed with PBS to remove nonefferocytosed neutrophils, and extracellular fluorescence was quenched using trypan blue (1:50 dilution). The fluorescent signal was assessed using a Molecular Devices Spectra M5 plate reader.

Data are presented as means ± SEM. Statistical analyses were performed with GraphPad Prism software version 5.04 for Windows (GraphPad Software, La Jolla, CA) using an unpaired t test for comparisons between two groups and two-way ANOVA with a Bonferroni posttest for comparisons that assess interactions between several treatment variables. Correlations between DEGs across time points were calculated by a Fisher’s exact test. The criterion for statistical significance was p < 0.05.

We first demonstrated the kinetics of myeloid cell populations in the lungs after i.t. instillation of E. coli LPS by flow cytometry on single-cell suspensions using standard immunophenotyping markers (9–11, 16, 17) (Fig. 1A). After gating on CD45⁺ leukocytes, neutrophils were selected based on Ly6G expression, and other myeloid cells were identified based on CD11b and/or CD11c expression. CD45⁺ leukocytes, neutrophils, and other myeloid cells were increased after i.t. LPS. Consistent with previous data (11), inflammatory cells peaked in the lungs at day 3 and subsequently returned to baseline cell number by day 7 for neutrophils and by day 14 for myeloid cells (Fig. 1B).

Next, we did further analysis based on a standard gating strategy used to identify macrophage populations, where resident AMs are identified as CD11b^−/low/CD11c⁺/Siglec-F⁺/F4/80⁺ cells, and both IMs/BM-derived macrophages are identified as CD11b⁺/CD11c⁺/Siglec-F⁻/F4/80⁺ cells (16, 18). Using this gating strategy, we found that the number of interstitial/BM-derived macrophages was significantly increased after i.t. LPS, with a peak at day 3. In the AM gate, however, there was a dramatic reduction in cells at day 3 (Fig. 1C, 1D). We then confirmed these findings using additional published gating strategies (9, 10) used to identify AMs (Supplemental Fig. 1). These data suggested either a loss of AMs during acute inflammation or a shift in cell surface marker expression.

To better distinguish resident AMs from BM-derived macrophages, we developed a (to our knowledge) novel technique to label these two cell populations using in vivo delivery of fluorescent dye. We instilled PKH67 green fluorescent dye into the airways by i.t. injection to label resident cells in the alveolar compartment and injected PKH26 red fluorescent dye directly into the tibia to label cells in the BM. Five days after injection of both fluorescent dyes, we performed flow cytometry on cells obtained from the lungs. As shown in Fig. 2A and 2B, PKH67 (green) labeled 5.22 ± 0.69% of all lung cells whereas PKH26 (red) labeled a very small population of lung cells that derive from BM origin (0.2 ± 0.06% of total lung cells). Further characterization of PKH67⁺ (green) cells using cell surface markers showed that 97.5 ± 0.52% of PKH67⁺ cells were immune (CD45⁺) cells, of which 98.1 ± 0.57% were identified as AMs by surface markers (Fig. 2C, 2D). Separately, we collected BAL fluid and showed that most BAL cells were PKH67⁺ and these cells were almost exclusively AMs (Supplemental Fig. 2A, 2B). In comparison, at 5 d after intratibial dye injection, 97.63 ± 1.67% of PKH26⁺ (red, BM origin) cells in the lungs were CD45⁺, of which 91.23 ± 1.70% were identified as monocytes/macrophages (Fig. 2E, 2F). In the blood, <1% of all leukocytes were labeled by PKH26; however, the vast majority of PKH26⁺ cells were identified as monocytes (89.87 ± 2.34%) (Supplemental Fig. 2C). Of note, i.t. PKH67 dye did not induce lung inflammation as measured by differential BAL cell counts between naive mice and mice that received PKH dye injections (data not shown).

Next, we used this PKH dye labeling technique to track resident AMs and BM-derived macrophages following i.t. LPS. As described above, resident AMs are characterized as CD11c⁺CD11b^−/low during homeostasis, and PKH67⁺ cells are included within this population (Fig. 3A, top panels). After i.t. LPS, however, dynamic changes in CD11c and CD11b expression by resident AMs result in substantial population shifts on flow cytometry plots, thus explaining the apparent loss of AMs based on cell surface marker identification (Fig. 3A). PKH67⁺ AMs upregulated expression of CD11b by day 1 after LPS treatment and showed peak expression on day 3 (Fig. 3A). At day 7 and day 14 after i.t. LPS, most PKH67⁺ AMs displayed reduced CD11b expression (toward baseline levels) along with persistently high CD11c expression. However, a small proportion of PKH67⁺ AMs showed high CD11b expression (with or without reduced CD11c expression) characteristic of interstitial or BM-derived macrophages at days 3, 7, and 14 after i.t. LPS (Fig. 3A). In addition to changes in CD11b and CD11c expression, PKH67⁺ AMs expressed increased levels of F4/80 at day 1 and 3 after i.t. LPS, which was reduced to baseline levels by day 7 (Supplemental Fig. 3A). In contrast, Siglec-F expression was not significantly altered during the course of i.t. LPS (Supplemental Fig. 3A).

We also tracked BM-derived macrophages based on PKH26-PE fluorescence. Although only 0.39 ± 0.03% of BM-derived macrophages were PKH26⁺ in the lungs of untreated mice, PKH26⁺ macrophages were markedly increased to 2.50 ± 0.38% of CD45⁺/CD11b⁺/CD11c⁺ cells at day 1 after LPS (Fig. 3A). At day 7 after LPS, a population of PKH26⁺ BM-derived macrophages showed reduced/absent CD11b expression similar to AMs (Fig. 3A), and this population was increased at day 14 after i.t. LPS (Fig. 3a, bottom rows of panels). Similar to AMs, PKH26⁺ BM-derived macrophages expressed higher levels of F4/80 at day 1 and 3 after i.t. LPS compared with day 7 (Supplemental Fig. 3B).

Quantification of resident AMs and BM-derived macrophages based on PKH dye fluorescence showed that BM-derived macrophages peaked in the lungs at day 1–3 after LPS, whereas AMs did not show statistically significant differences during the 14-d time course (Fig. 3B, 3C). Importantly, quantification of dye-labeled AMs and BM-derived macrophages yielded a dramatically different pattern of cellular changes in the lungs after LPS compared with myeloid cell subtype identification based on surface marker identification (see Fig. 1D). Taken together, our results show that cell surface markers used to distinguish macrophage subsets are unreliable in separating resident AMs and BM-derived cells during the course of LPS-induced inflammation. Our new methodology, however, facilitates dissection of macrophage subpopulations and avoids the problem of shifting surface marker expression.

We next sought to characterize the individual phenotypes of resident AMs and BM-derived macrophages present in the lungs after i.t. LPS at initiation (day 1), peak (day 3), and resolution (day 7) phases of lung inflammation using RNA-seq. After sorting PKH dye–labeled resident AMs and BM-derived macrophages (as shown in Supplemental Fig. 3C), we performed bulk RNA-seq on both cell populations. We included AMs from untreated mice in this analysis, but lungs of untreated mice contained too few BM-derived macrophages to analyze. Using principal component analysis and Pearson’s correlation (from >19,000 unique transcripts), we found a prominent distinction between resident AMs and BM-derived macrophages, with recruited BM-derived macrophages at day 1 and day 3 after LPS being the most divergent populations (Fig. 4A, Supplemental Fig. 4A, 4B).

Next, we performed a pairwise analysis of gene expression between resident AMs and recruited BM-derived macrophages at days 1, 3, and 7 after i.t. LPS (Fig. 4B). We observed the greatest transcriptomic difference between cells on day 1 after i.t. LPS with 4574 DEGs, including 2136 enriched in resident AMs and 2438 enriched in BM-derived macrophages. In comparison, 1505 genes were differentially expressed between cell types on day 3, and 469 genes were differentially expressed on day 7 after i.t. LPS. In addition, we investigated the relationships between DEGs across time points and found that 44.5% of day 3 DEGs (n = 647) were present at day 1 and 46.7% of DEGs (n = 191) at day 7 were present at day 3 (p < 0.1 × 10⁻⁴ using a Fisher’s exact test). Taken together, these findings indicate persistent phenotypic differences between resident AMs and BM-derived macrophages.

To identify differences in pathway activation between AMs and BM-derived macrophages, we performed pathway analysis of DEGs using DAVID Knowledgebase. More than 130 unique pathways were involved at one or more time points, and the most enriched pathways were present at day 1 (Fig. 5A). Compared to resident AMs, BM-derived macrophages showed enrichment in pathways related to the “signal transduction,” “immune system,” and “cellular processes” throughout the time course of lung inflammation. Within these categories, we identified upregulation of a variety of proinflammatory pathways in BM-derived macrophages, including MAPK signaling, HIF-1 signaling, JAK-STAT signaling, PI3K-AKT signaling, TNF signaling, NF-κB signaling, NOD-like receptor signaling, TLR signaling, and chemokine signaling. Other processes indicating cellular activation were upregulated in BM-derived macrophages, including cell adhesion molecules, extracellular matrix–receptor interactions, transendothelial migration, FcγR-mediated phagocytosis, focal adhesion, and regulation of actin cytoskeleton. In contrast, AMs showed enrichment in cellular processes and “metabolism”-related pathways, including PPAR signaling, peroxisome, and fatty acid degradation and metabolism throughout the course of LPS-induced inflammation (Fig. 5B). Taken together, these data indicate that recruited BM-derived macrophages express a more proinflammatory phenotype during LPS-induced inflammation compared with resident AMs. In addition, differences in metabolic pathways suggest distinct bioenergetic profiles between these macrophage subsets.

To further investigate inflammatory signaling in resident AMs following LPS treatment, we compared AMs from untreated mice (no LPS) to AMs obtained 1, 3, and 7 d after i.t. LPS (Supplemental Fig. 4C, 4D). LPS treatment resulted in upregulation of a number of immune system and signal transduction pathways, particularly at days 1 and 3 after i.t. LPS, thereby indicating a proinflammatory phenotype in AMs after LPS treatment. Differential gene expression between AMs from untreated mice and LPS-treated mice was largely resolved by day 7. These data show that, although BM-derived macrophages exhibited a more inflammatory gene expression profile compared with resident AMs, both macrophage subsets developed proinflammatory phenotypes at 1 and 3 d after LPS treatment.

Because pathway analysis indicated differential inflammatory responses between resident AMs and BM-derived macrophages, we examined expression levels of mediators considered to be important during lung inflammation, focusing on genes regulated through the NF-κB pathway. We evaluated NF-κB target gene expression profiles in resident AMs and BM-derived macrophages at day 1 after i.t. LPS and AMs from untreated mice based on NF-κB target genes databases (https://www.ncbi.nlm.nih.gov/gene and https://www.bu.edu/nf-kb/gene-resources/target-genes/). Compared to AMs from untreated mice, resident AMs and BM-derived macrophages from LPS-treated mice both showed increased expression of a number of NF-κB–dependent mediators, including IL-12, IL-1β, and CXCL3 (Fig. 6A). Direct comparison between resident AMs and BM-derived macrophages at day 1 after i.t. LPS revealed that BM-derived macrophages expressed higher levels of NF-κB pathway genes and immunoreceptor genes, whereas resident AMs expressed higher levels of several chemokines (Fig. 6B). In addition to RNA-seq data, we performed RT-PCR for selected gene products and confirmed that IL-1β expression and CCR2 expression were higher in BM-derived macrophages and that CXCL1, CXCL3, and IL-6 expression levels were higher in resident AMs (Fig. 6C). Taken together, these data indicate that lung macrophages develop differential inflammatory phenotypes between resident AMs and BM-derived macrophages.

Efferocytosis of apoptotic neutrophils is fundamental for resolution of neutrophilic inflammation and regulates macrophage phenotypes (19). Therefore, we assessed efferocytosis-associated transcripts in resident AMs and BM-derived macrophages after i.t. LPS and observed that resident AMs expressed higher levels of MerTK and Axl compared with BM-derived macrophages (Fig. 7A–C). RT-PCR studies validated these results (Fig. 7D, 7E). To investigate the functional impact of these gene expression changes, we performed efferocytosis assays using PKH dye–labeled macrophages isolated from the lungs by flow cytometry at day 7 after i.t. LPS. As predicted by gene expression analysis, resident AMs showed a greater capacity for efferocytosis of apoptotic neutrophils ex vivo compared with BM-derived macrophages (Fig. 7F). These data suggest that differences in efferocytosis could impact the immune phenotype of AMs and BM-derived macrophages in the lungs.

Peroxisomes are organelles that regulate intracellular lipid metabolism and reactive oxygen species and have been shown to influence innate immune pathways (20, 21). Based on pathway enrichment for peroxisome and fatty acid metabolism pathways in resident AMs following i.t. LPS, we indicated peroxisome, fatty acid metabolism, and reactive oxygen species metabolism genes that were upregulated in AMs (Fig. 8A). Network analysis of potential interactions (Fig. 8B) showed functional interconnection of these gene products and suggested increased peroxisome activity coordinated by PPARγ signaling. Consistent with this conclusion, expression of PPARγ was significantly increased in resident AMs compared with BM-derived macrophages after i.t. LPS (Fig. 8C, 8D). Taken together, these findings indicate differential NF-κB signaling, efferocytosis, and PPARγ signaling in resident AMs and BM-derived macrophages after i.t. LPS.

Although both resident AMs and recruited BM-derived macrophages play important roles in lung inflammation (22, 23), the plasticity and complexity of pulmonary macrophages has made full characterization of these two macrophage subsets challenging (24–26). In our study, we found that CD11b expression was induced in AMs after i.t. LPS, peaking at day 3 of inflammation. Although CD11b expression was subsequently reduced toward basal levels in most AMs by day 7, a minority of AMs had persistently high CD11b expression and variable CD11c levels at day 7 and 14 after i.t. LPS. In contrast, BM-derived macrophages initially express both CD11c and CD11b, but over time a portion of these cells reduce CD11b and increase CD11c, giving them characteristics of resident AMs. These dynamic changes of key cell surface markers make it impossible to distinguish AMs and BM-derived macrophages solely by cell surface marker expression during nonhomeostatic conditions. We developed a dual PKH dye–labeling technique to overcome these issues and validated this method during the course of LPS-induced inflammation. By specific identification of AMs and BM-derived macrophages in the lungs, further analysis of the transcriptome identified the greatest number of DEGs between these two cell types at day 1 after i.t. LPS. Specifically, BM-derived macrophages were enriched in the expression of genes related to pathways involving signal transduction and immune system activation after i.t. LPS, indicating increased inflammatory signaling by these cells. In contrast, resident AMs were enriched in cellular and metabolic processes such as PPAR signaling, peroxisome pathways, and fatty acid degradation and metabolism.

During homeostasis, myeloid cells in the lungs are generally categorized as AMs, IMs, neutrophils, and monocytes (27). As the name implies, AMs are primarily localized in alveoli and patrol the distal lung to clear bacteria and particulates that enter the distal lungs (28). In contrast, IMs appear to reside primarily in the interstitium surrounding bronchioles and can be separated into different subsets that self-renew or are replenished from circulating progenitors (10, 29). During inflammation, another type of macrophage, known as recruited BM-derived macrophages, monocyte-derived macrophages, or exudative macrophages, arise from monocytes recruited to the lungs via the CCL2/CCR2 axis (30–32). These recruited BM-derived macrophages play roles in upregulation of the initial inflammatory response and resolution after the peak of inflammation (31, 32). Although there have been important advances in understanding the separate roles of AMs and BM-derived macrophages during lung infection and inflammation, study design has remained challenging because of difficulties in implementing convenient methodologies to distinguish these cellular subsets. Prior studies have used BM transplantation to separate resident AMs and BM-derived macrophages for functional studies (11, 33); however, this process is labor-intensive and may impact the inflammatory response (even when thoracic shielding is employed).

To develop a more efficient system for distinguishing resident AMs and BM-derived macrophages, we employed a dual PKH dye–labeling technique. PKH dyes are cationic fluorochromes with long saturated aliphatic tails and long half-lives (12), making them ideal for long-term cell labeling. Although prior studies (11, 34) have used i.t. delivery of PKH dyes to identify resident AMs, intratibial labeling with PKH26 dye represents a (to our knowledge) novel approach to specifically identify BM-derived macrophages during inflammation. For unclear reasons, the vast majority of cells that are labeled with these dyes in vivo are CD45⁺ immune cells, predominantly monocyte/macrophages, with few structural cells taking up the PKH dye. Thus, tracking these fluorescently labeled monocytes/macrophages is straightforward. Importantly, we found very few dual-labeled cells in our model using this technique, indicating that the cell labeling is distinct for the two macrophage populations and suggesting that these dyes are infrequently transferred among cells by release of exosomes or efferocytosis. Although the labeling of BM-derived cells by this method is relatively inefficient (≤2.5% of CD45⁺/CD11b⁺/CD11c⁺ cells at each time point), specificity is high (>90% of labeled cells are monocytes/macrophages). Of note, i.v. injection of PKH dyes can also label monocytes/macrophages, but compared with the intratibial technique the efficiency of labeling is not enhanced and the specificity for BM-derived macrophages is reduced (data not shown).

The functions of lung macrophages during LPS-induced lung inflammation have been extensively studied with the development of multicolor flow cytometry and next-generation sequencing techniques (9–11, 35–37). An important study by Janssen et al. (11) used BM transplantation combined with a PKH26-PCL–labeling AM method to show that recruited macrophages peak in the lungs at day 3 after LPS treatment. Subsequent studies by this group using CD11b and CD11c Abs for sorting AMs and recruited macrophages showed increased inflammatory signaling by recruited macrophages (36). Most recently, single-cell RNA-seq of macrophages obtained by BAL from mice at baseline and day 3 and 6 after i.t. LPS indicated that five clusters of macrophages were separated into recruited and resident AMs based on comparison with cells subsets present at baseline. In BAL, three subsets of myeloid cells arose after i.t. LPS and were classified as “recruited” macrophages (37). Among these cells, TLR signaling, NF-κB signaling, and IFN signaling was increased in recruited macrophages, similar to our findings in labeled BM-derived macrophages. In addition, resident AMs in this study showed increased PPAR signaling and “lipid digestion, mobilization, and transport,” also consistent with our findings (37). Our studies support and extend these observations, emphasizing the notion that resident AMs and BM-derived macrophages have different characteristics associated with immune response and metabolism changes during lung inflammation.

In summary, our studies support the dual PKH dye–labeling technique as a valid approach to distinguish macrophages subsets. This approach allowed accurate tracking of resident AMs and BM-derived macrophages during the course of LPS-induced inflammation. We believe this (to our knowledge) novel method will facilitate future studies to identify and characterize AMs and BM-derived macrophages during different types of inflammatory conditions in the lungs.

The authors have no financial conflicts of interest.