Anti‐inflammatory mechanisms in cancer research: Characterization of a distinct M2‐like macrophage model derived from the THP‐1 cell line

Abstract Aims Macrophages play an essential role in cancer development. Tumor‐associated macrophages (TAMs) have predominantly M2‐like attributes that are associated with tumor progression and poor patient survival. Numerous methods have been reported for differentiating and polarizing macrophages in vitro, but there is no standardized and validated model for creating TAMs. Primary cells show varying cytokine responses depending on their origin and functional studies utilizing these cells may lack generalization and validity. A distinct cell line‐derived TAM‐like M2 subtype is required to investigate the mechanisms mediated by anti‐inflammatory TAMs in vitro. Our previous work demonstrated a standardized protocol for creating an M2 subtype derived from a human THP‐1 cell line. The cell expression profile, however, has not been validated. The aim of this study was to characterize and validate the TAM‐like M2 subtype macrophage created based on our protocol to introduce them as a standardized model for cancer research. Methods and results Using qRT‐PCR and ELISA, we demonstrated that proinflammatory, anti‐inflammatory, and tumor‐associated marker expression changed during THP‐1‐derived marcrophage development in vitro, mimicking a TAM‐related profile (e.g., TNFα, IL‐1β). The anti‐inflammatory marker IL‐8/CXCL8, however, is most highly expressed in young M0 macrophages. Flow cytometry showed increased expression of CD206 in the final TAM‐like M2 macrophage. Single‐cell RNA‐sequencing analysis of primary human monocytes and colon cancer tissue macrophages demonstrated that cell line‐derived M2 macrophages resembled a TAM‐related gene profile. Conclusions The THP‐1‐derived M2 macrophage based on a standardized cell line model represents a distinct anti‐inflammatory TAM‐like phenotype with an M2a subtype profile. This model may provide a basis for in vitro investigation of functional mechanisms in a variety of anti‐inflammatory settings, particularly colon cancer development.


| INTRODUCTION
Macrophages have various roles in cancer development, including inflammation, proliferation, and tissue remodeling. 1,2Tissue macrophages induce responses that can have either local or systemic effects, ranging from the regulation of wound healing to the induction of acute phase protein synthesis through hepatocytes, thereby affecting systemic inflammation and metabolism. 2,35][6][7][8] These models show varying cytokine responses depending on their origin and the respective cell treatment, which makes comparing findings of different studies difficult.A standardized approach using a cell line model with reproducible cellular responses mimicking primary human cells is needed to gain insight into the complexity of macrophage-mediated cancer development.
Two major subtypes of macrophages can be defined depending on their origin.Contrary to the erroneous assumption that tissue macrophages mostly originate from circulating bone marrow monocytes that invade tissues, most local macrophage populations are derived from embryonic progenitors that are seeded before birth. 1,9,10hese fetal macrophages initially evolve from yolk sac blood islands that develop during the first 2 weeks of gestation, and in adults they are believed to maintain their population by self-renewal. 11Whether they can actually persist through adulthood and to which extent they replenish their populations is currently still discussed. 12n contrast, bone marrow-derived tissue macrophages are part of the well-established mononuclear phagocytic system (MPS) and evolve from circulating peripheral monocytes. 9Among the mononuclear phagocytic lineage, macrophages represent the final cell type after differentiation of monocytes that originate in the bone marrow. 9he monocyte-like THP-1 cell line was derived from peripheral monocytes of a 1-year-old infant with acute monocytic leukemia and has been widely used as an in vitro model for demonstrating mechanisms in human monocytes and macrophages. 13,14THP-1 cells can be differentiated into macrophage-like cells and then polarized into either proinflammatory M1 or anti-inflammatory M2 phenotypes.Due to their high plasticity, macrophages can continuously switch between a predominantly proinflammatory or mostly anti-inflammatory state. 15][18] Depending on the desired outcome of a study, different reagents are used to induce differentiation and polarization. 6,17,19While differentiation of primary monocytes is induced using granulocyte macrophage colony-stimulating factor (GM-CSF) or macrophage colony-stimulating factor (M-CSF), THP-1 cells require protein kinase C activators, such as phorbol 12-myristate 13-acetate (PMA) or byrostatin for differentiation into macrophages. 20Either interleukin 4 (IL-4) or a combination of IL-4 and interleukin 13 (IL-13), tumor growth factor-β1 (TGF-β1), or interleukin 10 (IL-10) can be used for polarization of THP-1 monocyte-derived macrophages (MDMs). 7,18Comparison of studies using a variety of methods, however, is problematic due to the lack of standardized and established procedures that use well-defined macrophage subtypes.
Macrophages are characterized by their high plasticity, changing their cellular metabolism to switch between M1and M2-like phenotypes. 21This dichotomous model cannot fully describe their characteristics, as macrophages, specifically tumor-associated macrophages (TAM), do not exclusively express either M1-or M2-like markers.Macrophage characterization based on this paradigm is, however, widely used and the fundamental basis for further functional analyses. 5,22Proinflammatory M1-like markers, such as CD80, TNFα, CXCL10, tend to be associated with an antitumor immune response, while macrophages expressing other proinflammatory cytokines (IL-1β, IL-6) can mediate tumor cell migration and cancer progression. 23,24Predominantly anti-inflammatory markers, such as CD206, IL-10, IL-8, CCL8, and CCL22 are known to be associated with advanced tumor stage and poor survival. 25urthermore, TAM express tumor-promoting mediators that are involved in tumor neovascularization, such as matrix metalloproteinases (MMP) and ADAM metallopeptidase with thrombospondin type 1 motif 1 (ADAMTS1). 26,27The transcription factor activator protein 1 (AP-1) is closely related to cell proliferation differentiation and apoptosis and mediates proinflammatory cytokine expression in macrophages. 28Alpha/beta-hydrolase domain containing 5 (ABHD5), a key enzyme in of functional mechanisms in a variety of anti-inflammatory settings, particularly colon cancer development.

K E Y W O R D S
colon cancer, colonic neoplasms, immunology, inflammation, RNA-Seq, THP-1 cells, tumorassociated macrophages lipolysis regulating tumor biology 29 and toll-like receptor 4 (TLR4), induced by fatty acids and mediating nuclear factor-κB (NFκB) pathways, both also play an important role in TAM-mediated tumor progression. 30Another cancer-related gene involved in lipid metabolism and macrophage activation is monoacylglycerol lipase (MGLL) contributing to lipid accumulation in macrophages and regulating tumor progression. 302][33] In colon cancer, the M2a subtype with a STAT6-stimulated pathway plays a particular role. 34,35n order to investigate the mechanisms mediated by anti-inflammatory TAMs in a cell culture model, a distinct TAM-like M2 macrophage subtype is required.Therefore, in vitro inflammatory stimuli must be overcome, including the mechanical stress to the cells as a consequence of media changes, and/or cell treatment with proinflammatory compounds, such as PMA to induce monocyte differentiation. 36,37he aim of this study was characterize THP-1 MDMs after differentiation and polarization into a distinct M2a subtype.The protocol for creating these M2-like macrophages has been previously described. 38Herein, cytokine and tumor-associated marker gene expression, protein secretion, and cell surface marker expression of cells within this model were investigated and validated as a distinct TAM-like M2 phenotype.

| MATERIALS AND METHODS
Macrophages were created based on our 14-day cell line protocol 38 and characterized according to their marker profile using quantitative real-time PCR (qRT-PCR), enzyme-linked immunosorbent assay (ELISA), flow cytometry, and RNA-sequencing analysis.

| Cell supernatant preparation and enzyme-linked immunosorbent assay
At baseline and after cell differentiation and polarization, respectively, growth medium was changed to RPMI-1640 without supplements.Cell supernatants were collected and spun down at 1600 rpm for 7 min and separated from remaining cells.

| Flow cytometry
Macrophages were lifted from 24-well plates using the cold shock method by incubating cells with ice-cold PBS/5% FBS and placing them on ice for 45 min.Following this step, cells were mechanically detached using cell scrapers and washed with cold PBS/5% FBS.
THP-1 cells (n = 5), young M0 (n = 5), and M2-like cells (n = 5) were investigated.Nonspecific binding of staining antibodies was inhibited by incubation with Fcγ-receptor block (BD Pharmingen, San Diego, USA) at room temperature for 10 min.Cells were then stained according to the manufacturer's instructions (BD Parmingen, San Diego, USA) with FITC-conjugated mouse anti-human CD14 and CD80 antibodies, with PE-conjugated mouse anti-human CD11b antibodies and with PE-Cy5-conjugated mouse anti-human CD206 antibodies and their isotype-matched IgG for 30 min at 4°C.Cells were then fixed with 1% formaldehyde.20,000 cells were acquired for each measurement.
Four-color flow cytometric analysis was performed and fluorescence quantitated using a BD FACSCalibur flow cytometer with CellQuest software (BD Biosciences, San Diego, USA).Gating of cells was carried out to exclude cell debris according to forward and side scatter.Cell viability after processing cells for flow cytometry with detachment of macrophages from cell culture plates was determined in a representative set of samples (n = 6) after incubating the cells with 7-aminoactinomycin-D (7-AAD) for 5 min.

ELISA analysis
For qRT-PCR analysis, all delta CT values between cell types were compared using a one-way analysis of variance (ANOVA) model with a post hoc Benjamini-Hochberg correction to control the false discovery rate at 5%. 39 ELISA cytokine concentrations were compared using a one-way ANOVA with a Bonferroni test. 40The heat map for qRT-PCR was created using R software and pirate plots for ELISA results were generated using the software R package "yarrr." 41,422.5.2 | RNA-sequencing analysis

Sample collection
For the cell line analysis, three THP-1-derived M2-like macrophage samples were created according to our protocol and sequenced by the HudsonAlpha Genome Sequencing Center (Huntsville, AL).Two corresponding THP-1-derived peripheral blood monocyte samples were downloaded in fastq format from Gene Expression Omnibus (GEO) 43 (data accessible at NCBI GEO database, Phanstiel et al. 2017, accession GSE96800 [samples GSM2599707 and GSM2599708]).Raw read counts from a third THP-1 peripheral blood monocyte sample were downloaded from the Dependency Map Portal (DepMap) 44 with the corresponding ID ACH-000146.
For the primary tissue analysis, raw single-cell read counts for human TAMs from colon cancer core tissue were obtained from the Single Cell Atlas accession E-MTAB-8410. 45E-MTAB-8410 samples were gathered from nine patients and covered several tissue types: colon, caecum, ascending colon, sigmoid colon, and rectosigmoid colon.Three samples were collected from each patient: tumor cells, tumor border cells, and adjacent normal cells.Raw single-cell read counts for human peripheral blood monocytes derived from colon cancer patients were obtained from GEO (data accessible at NCBI GEO database, accession GSE178318). 46n overview of the data is shown in Table S1.

Cell line RNA-sequencing analysis
The THP-1-derived M2-like TAM cell lines sequenced by HudsonAlpha and the two THP-1 peripheral blood monocyte cell lines retrieved from GEO were aligned to reference assembly hg38 using RSEM. 47Expected counts were generated with Gencode annotations V34.
The expected counts for a third monocyte cell line retrieved from DepMap were previously generated by RSEM using Gencode V34.Raw counts for the three TAM cell lines and the three monocyte cell lines were input to DESeq2 48 for differential expression analysis which uses relative log expression as its default normalization method.

Primary tissue RNA-sequencing analysis
Raw counts for peripheral blood mononuclear cells (PBMCs) from GSM2599708 were integrated with raw counts for all cells from E-MTAB-8410 using Seurat. 49vailable metadata were also input to Seurat.Cells with greater than 6000 features (possible doublets), fewer than 200 features (poor capture), and greater than 20% mitochondrial reads (possible dead and dying cells) were removed prior to integration.As part of the integration process, scaling and principal component analysis were performed on the individual samples.Integration anchors were identified using RPCA reduction with up to 30 dimensions specified.Scaling, dimension reduction (PCA/UMAP), and clustering at a resolution of 0.5 was performed on the integrated data.The integrated Seurat object was uploaded to BioTuring's BBrowserX 50 for cell type prediction.Version 3 of BioTuring's machine learning algorithm, which uses 84 million cells as a reference, was selected for cell type prediction.TAMs were identified by considering the annotations provided by Lee et al. in the original study (anti-inflammatory myeloid cell, SPP1 + B myeloid cell, SPP1 + A myeloid cell, and pro-inflammatory myeloid cell) along with BioTuring's BBrowserX annotations (macrophage).One set of TAMs includes the 103 cell annotations in agreement across the two methods, which was analyzed and is presented in the results section.A second set includes the 103 TAM annotations from Set 1 and all additional macrophage annotations by Lee et al., which resulted in analyzing 656 TAMs (Data S1).Differential expression between TAMs and PBMC monocytes was performed for each set of TAMs using Seurat's findMarkers function and the Wilcoxon rank sum test.

| Expression of pro-and anti-inflammatory markers increases during development of THP-1 monocytes into TAM-like M2 macrophages
Pro-and anti-inflammatory marker expression was significantly different among all cell types for CD80, TNFα, IL-1β, CXCL10, CD206, IL-10, IL-8/CXCL8, CCL18, and CCL22; each with p < 0.001, and for IL-6 p = 0.033 (Tables 1 and 2).Expression patterns (mean ΔCT values) of all markers for the respective cell types are shown in Figure 2 and p-values of pairwise comparisons are given in Table S2.

| M1-associated marker expression increases, but not interleukin-1β expression
Overall p-values with means and SDs of PCR ΔCT values for all cell types are shown in Table 1.
The pro-inflammatory marker TNFα was significantly downregulated in "young M0" and "old M0" compared to M2 macrophages and compared to THP-1 monocytes.The difference between M2-macrophages and THP-1 cells at baseline was not statistically significant (Figure 2, Table 1).
The M1-associated marker IL-1β was significantly upregulated in "young M0" and "aged M0" macrophages compared to THP-1 cells as well as compared to M2 subtype macrophages."Young M0" macrophages showed the highest level of IL-1β expression (Figure 2, Table 1).
CXCL10 showed a significantly higher expression in "young M0," "aged M0," and M2 macrophages compared to THP-1 monocytes.The highest level of CXCL10 expression was demonstrated in both M0 cell groups (Figure 2, Table 1).

| M2-associated marker expression increases, but not interleukin-8 expression
Means, SDs, and overall p-values of PCR ΔCT values for all cell types are shown in Table 2.
The highest M2-associated CCL18 expression was seen in M2 macrophages with significant upregulation compared to all other cell types."Aged M0" cells showed a significantly higher expression compared to "young M0" and THP-1 cells (Figure 2, Table 2).

| Expression of tumor-associated markers increases in transition from THP-1 monocyte to TAM-like M2 macrophage
Tumor-associated marker expression was significantly different among all cell types for MMP2, MMP7, MMP9, MMP12, ABHD5, ADAMTS1, AP-1, and MGLL each with p < 0.001 (Table 3).TLR4 expression showed no significant difference.The corresponding p-values of pairwise comparisons are given in Table S2.
All investigated tumor-promoting MMPs were significantly increased in M2-like macrophages, with MMP2 showing the lowest upregulation in M2-like macrophages compared to THP-1 monocytes (Table 3).The metallopeptidase ADAMTS1 was significantly downregulated in M2like, "young M0" and "aged M0" macrophages compared to THP-1 cells (Table 3).ABHD5, a lipolytic factor either potentiating tumor growth or functioning as a tumor suppressor in certain types of cancer, was significantly downregulated in M2-like macrophages compared to all other cell types (Table 3).Upregulation of MGLL, a key enzyme in lipid metabolism that also has both tumor promoting and suppressing effects, was most evident in both M0 phenotypes compared to THP-1 monocytes, but also significant in M2-like macrophages (Table 3).
The immune response receptor TLR4 showed no significantly different expression among cell types.The transcription factor AP-1 exerting a dual role among different types of cancers, showed a slight upregulation in "young M0" macrophages compared to THP-1 cells, but no difference was found in M2-like macrophages versus THP-1 (Table 3).

| Expression of M2a-specific subtype markers in cell line-derived TAM-like M2 macrophages
Expression of M2a-specific subtype markers was significantly upregulated in cell line-derived TAM-like M2 macrophages compared to THP-1 monocytes (Table S3).
F I G U R E 2 Marker gene expression using qRT-PCR among cell types.Heat map of marker gene expression (ΔCT values) of pro-and anti-inflammatory markers among all cell types.Lowest gene expression is depicted by red, highest expression by green.CCL, C-C motif chemokine ligand; CD, cluster of differentiation; CXCL, C-X-C motif chemokine ligand; TNFα, tumor necrosis factor α; IL, interleukin.

| Cell line-derived M2 macrophages express anti-inflammatory cell surface markers
Cell viability of differentiated macrophages after incubating cells with 7-AAD was >95% throughout all samples.

|
The RNA-sequencing profile of cell line-derived TAM-like M2 macrophages resembles the profile of primary human colon cancer TAMs 3.5.1 | Inflammatory and tumor-associated genes change significantly in TAM-like M2 macrophages compared to THP-1 monocytes RNA-Seq data were used to investigate the expression level of a respective gene in the baseline cell line monocyte and to compare changes in cell line-derived M2-like macrophages (Table 4).
Single-cell RNA-Seq data were analyzed to demonstrate the proportion of cells expressing the respective genes in primary PBMCs compared to primary human colon cancer tissue macrophages with a mean log 2 fold change (Table 5).
A high proportion of primary peripheral blood monocytes (>20%) expressed CD86, TNFα, IL-1β, IL-8, and AP-1.The proportion of TAMs expressing these genes was decreased and showed significantly lower expression, except for CD86.Proinflammatory CD80, anti-inflammatory CD206, CCL18, and CCL22, and the tumor-associated markers including MMPs, MGLL, and TLR4 were expressed by a low proportion of monocytes (<1%), but significantly upregulated in TAM.ADAMTS1 was neither expressed in monocytes nor TAM.
A Spearman correlation was used to investigate the association of log2 fold changes for genes between the cell line-derived TAM-like M2 macrophages and primary human TAM.Compared to primary cells from colon cancer patients, cell line-derived monocyte-macrophage development showed a similar change in cell gene expression profiles (Figure 4).While more profound changes in gene expression could be observed with the cell line as expected, primary cell changes correlated significantly (R s = 0.81, p < 0.001).Taken together, these data indicate that the gene expression of cell line-derived TAM-like M2 macrophages resembles that of primary human colon cancer TAMs.
An additional analysis with the subset of 656 primary colon cancer TAMs labeled by Liu et al. compared to primary human peripheral blood monocytes (N = 2099) showed similar confirmatory results (Table S4, Figure S3).Macrophages mediate central mechanisms in wound healing, fibrosis, and cancer growth.Specifically, the proportion of anti-inflammatory M2 macrophages is associated with advanced tumor stage and decreased overall survival in certain types of cancers. 32,51,52Currently, there is no distinct M2 macrophage subtype based on an established reproducible protocol that makes studies investigating functional analyses in cancer research comparable.7][18] Cell line-derived macrophage models are based on immortalized cells derived from only a single patient and increase reproducibility by showing more consistent and enhanced cellular responses compared to primary cells. 53While models using primary cells have been described to resemble a more complex, and fore in vivo macrophage response, data obtained using such cells often vary and are difficult to reproduce.Data reproducibility is a key component of funding mandates. 54e use of primary human cells is associated with various other disadvantages, such as impaired cell growth, limited replicative capacity, and cell senescence, varying cellular responses due to donor variability, potential ethical concerns, and high cost. 55,56On the other hand, cell lines are genetically manipulated, which may alter their responsiveness to stimuli. 53Cross-contamination, misidentification, and mycoplasma contamination of cell lines can alter cellular behavior. 57,58Therefore, cell lines should not be used exclusively to investigate functional cellular responses.After identifying a cellular mechanism within cell lines, experiments utilizing primary cells may help in contributing to confirm these mechanisms as a basis for planning future in vivo models.

Mean
We have provided a protocol of a cell line model for differentiating and polarizing THP-1 monocytes within 14 days and herein characterize the distinct TAM-like M2a macrophage subtype that is thereby created.This provides the basis for investigating mechanisms mediated by anti-inflammatory macrophages in vitro. 38Macrophages can be divided into several subsets from M2a-d. 59These subsets have been investigated in association with the development of different cancers, and their gene expression profiles are overlapping.Macrophage subsets can only be identified by focusing on specific key genes.In our distinct TAM-like macrophage subtype, an M2a profile could be identified.These macrophages show high STAT6-related gene expression, induced by IL4 and IL13 exposure, 34 which is a central mechanism in colon cancer development and progression. 34,35Key genes of M2a macrophages play a particular role in colon cancers.ITGAM or CD11b, a marker of macrophage activation, has been shown to be upregulated in colon cancer. 60STAB1 is a scavenger receptor on anti-inflammatory macrophages and was demonstrated to facilitate lymphatic metastasis.Its upregulation is associated with shorter disease-specific survival. 61CCL17 is involved in colon cancer cell migration as well but has not shown to be significantly upregulated in our TAM-like macrophages. 62LEC4A and TREM2 were significantly upregulated in the created cell line-derived TAM-like macrophages described herein.These tumor suppressor genes are linked to colon cancer cell proliferation and development. 63,64LEC4A has been demonstrated to have an impact on colon inflammation and the onset of colitis in particular. 64herefore, our TAM-like macrophage subtype represents an attractive model to study the underlying inflammatory mechanisms impacting tumor development within functional experiments.
One of the limitations of this study is the use of a human cell line model rather than primary cells as a basis to investigate human macrophage mechanisms in vivo.In contrast, primary human macrophages are characterized by great variation in marker expression patterns and have different functional roles according to their source of origin. 65Previous data showed that PMA-treated THP-1 macrophages did not completely mimic the complexity of primary MDM activation. 66The use of a cell line, however, can provide a reproducible and standardized model to recreate functional responses in vitro.Furthermore, F I G U R E 4 Correlation of changes in gene expression between primary human peripheral blood monocytes (N = 2099) and primary human colon cancer tissue TAMs (N = 103) and cell line-derived THP-1 monocytes (N = 3) and cell line-derived TAM-like M2 macrophages (N = 3).Primary cell data represents single-cell RNA-sequencing data, while cell line-derived results are based on bulk RNA-sequencing data.Red: proinflammatory gene.Blue: anti-inflammatory gene.Black: tumor-associated gene.ABHD5, alpha/beta-hydrolase domain containing 5; ADAMTS1, ADAM metallopeptidase with thrombospondin type 1 motif 1; AP-1, activator protein 1; CCL, C-C motif chemokine ligand; CD, cluster of differentiation; CXCL, C-X-C motif chemokine ligand; IL, interleukin; MGLL, monoglyceride lipase; MMP, matrix metalloproteinase; TLR4, toll-like receptor 4; TNFα, tumor necrosis factor α.
characteristics of THP-1-derived cells used for mimicking macrophage mechanisms in vivo can be altered to achieve similar marker profiles to primary macrophages.Another limitation is the complexity of myeloid cell characterization due to cell plasticity and the wide variety of functional abilities of these cells.This leads to limited characterization of M2 macrophages created by this protocol.The cell treatment scheme of this study, however, demonstrates the reproducible creation of a distinct M2-like phenotype of macrophages, that can be further characterized depending on the intended experimental setting. 38Cytokine mRNA and protein expression and flow cytometry used in this study demonstrated a clear anti-inflammatory shift of differentiated macrophages and differences between resting cells versus cells receiving polarization treatment.Further analyses concerning pro-and anti-inflammatory cytokine expression, protein production, and cell surface marker expression as well as functional analyses of these M2-like macrophages should follow on the basis of the underlying treatment protocol.
The differentiation of THP-1 monocytes is widely performed by incubating cells with the proinflammatory compound PMA. 17,36,37,67This method of differentiation is especially suitable for mimicking primary human MDM, showing similar characteristics concerning cell morphology, macrophage surface markers, and cytokine profiles. 68ro-inflammatory responses of the macrophages that are created by this approach must decrease in order to obtain non-polarized M0 macrophages and to assure an adequate M2-like polarization process.For this reason, cells are reported to rest in medium only after PMA treatment for various periods of time. 6,16,67This model is based on a 5-day rest period after PMA incubation, creating macrophages that mimic primary human MDM in terms of cytoplasmatic volume and cell surface adherence. 67,69he cytokine profile of young M0 macrophages that are described in this study demonstrated significantly increased expression of the proinflammatory markers CD80, IL-1β, IL-6, and CXCL10 compared to THP-1 monocytes, while no difference in TNFα expression was shown.Primary colon cancer cells showed downregulation of TNFα.Even though the term "M0 macrophages" defines differentiated macrophages in a resting state in vitro, characteristics of cellular activation (M1-like state) are usually identified in the real-world setting. 70Therefore, the standardized model presented herein clearly describes cytokine expression as a baseline threshold for future experiments.The anti-inflammatory marker IL-8, however, is most highly expressed in young M0 macrophages and significantly upregulated in THP-1-derived M2-like macrophages compared to all the other cell stages.This was confirmed by protein expression analysis, showing high variability of IL-8/CXCL8 concentrations among "young M0" macrophages.RNA-Seq data of colon cancer TAMs showed IL-8 expression in a high proportion of cells with downregulation compared to PBMCs.Variation in expression levels of IL-8/CXCL8 is a previously described characteristic of this cytokine in several different cell types, such as pulmonary epithelial cells or cervical cancer cells. 71HP-1-derived macrophages have also been reported to show highly variable IL-8/CXCL8 protein expression after treatment with proinflammatory stimuli. 69,71This observation is partially understood to result from the functional role of IL-8/CXCL8, which involves synchronization of several regulatory pathways, including NFκ-B, c-Jun N-terminal kinase (JNK), and p38 mitogen-activated protein kinase (p38 MAPK). 69he M2-like macrophages created by this model demonstrate a significant upregulation of the anti-inflammatory markers CD206, IL-10, and CCL18 compared to both "young M0" and "aged M0" macrophages, with CCL18 upregulation showing the greatest difference in M2-like cells as compared to either type of M0 cells.These results could be confirmed by primary colon cancer TAM scRNA-Seq data.This demonstrates that the cell line model of this study is mimicking primary human MDM, as CCL18 is also abundantly produced by these cells following M2-like polarization. 72][75] Both groups of M0 macrophages did not differ from M2like cells with respect to CCL22 expression.Production of IL-10 in M2-like macrophages was confirmed by a significantly increased IL-10 protein expression in these cells compared to that observed in THP-1 monocytes and young M0 macrophages.IL-10 production in "aged M0" macrophages did not differ from M2 macrophages on the protein level, however, the variability in IL-10 production was high among "aged M0" cells.This was confirmed by our scRNA-Seq data analysis of primary TAMs.IL-10 is reported to be upregulated in polarized primary human MDM as well, mostly as part of an autocrine mechanism using the signal transducer and activator of transcription 3 (STAT3) pathway. 8low cytometry analysis for cell surface expression markers revealed that the distinct M2-like subtype that is created by this protocol not only shows high expression of the anti-inflammatory marker CD206, but also of both myeloid markers CD14 and CD11b.The macrophage differentiation marker CD14 is also highly expressed by primary human MDM as a response to IL-10 and distinguishes THP-1-derived macrophages from THP-1 cells that are differentiated into a dendritic cell type with low CD14 expression. 76,77Co-expression of the myeloid CD11b has been reported in primary human muscle tissue macrophages. 78Furthermore, the anti-inflammatory M2-like markers CD206 and IL-10 are expressed by resident macrophages of the colon lamina propria. 79he tumor-associated attributes of created macrophages were additionally assessed focusing on marker genes that play a role in neovascularization as well as cellular metabolism and therefore macrophage polarization.
MMPs are important cancer-related targets promoting vascularization and tumor development. 26,27All investigated subtypes of MMPs were upregulated in the M2 macrophage phenotype, which was confirmed in our primary colon cancer TAM data.The metallopeptidase ADAMTS1 was significantly downregulated in all cell types compared to THP-1 cells and showed the most evident decrease in M2-like macrophages.Primary TAM scRNA-Seq data showed no expression of ADAMTS1 among TAMs, which confirms ADAMTS1 playing a minor part in the final macrophage phenotype.
][82][83] All of these markers are expressed in the M2 macrophages created by this model and expression was also demonstrated to be present in primary TAMs.This suggests that changes in gene expression due to cell treatment or cell signaling can be assessed.The demonstrated tumor-associated gene expression patterns are the basis for expression changes in different experimental settings.
A clear similarity of cells created by this model to primary human macrophages and tumor-associated characteristics of the M2-like phenotype are to be confirmed by further functional experiments comparing THP-1 derived macrophages versus primary cells.
In conclusion, M2-like macrophages created using this 14-day cell line model of differentiating and polarizing THP-1 monocytes represent a distinct anti-inflammatory TAM-like phenotype.Pro-and anti-inflammatory gene expression patterns, protein expression, and cell surface marker analyses revealed an M2-like expression profile of macrophages similar to reported gene expression patterns in human primary MDM, with tumor-associated attributes.The distinct M2-like macrophage M2a subtype provides the basis to investigate M2-like macrophage-mediated mechanisms in vitro in various experimental settings, particularly colon cancer progression.

F I G U R E 3
Pirate plots of protein expression of anti-inflammatory markers among cell types (A) interleukin 8 (IL-8), (B) interleukin 10 (IL-10).IL, interleukin.T A B L E 3 Tumor-associated marker expression (ΔCT values) among cell types.
Anti-inflammatory marker expression (ΔCT values) among cell types.Proinflammatory markers expression (ΔCT values) among cell types.Marker expression was investigated using quantitative real-time PCR.Results are shown as mean ΔCT values and standard deviations (SD) for all cell types.Significant results of pairwise comparisons between cell types are listed.Abbreviations: CD, cluster of differentiation; CXCL, C-X-c motif chemokine ligand; IL, interleukin; TNFα, tumor necrosis factor α. p < 0.05 indicates a significant pairwise comparison.*p < 0.05, **p ≤ 0.01, ***p ≤ 0.001.
T A B L E 2Note: Marker expression was investigated using quantitative real-time PCR.Results are shown as mean ΔCT values and standard deviations (SD) for all cell types.Significant results of pairwise comparisons between cell types are listed.Abbreviations: CCL, C-C motif chemokine ligand; CD, cluster of differentiation; CXCL, C-X-C motif chemokine ligand; IL, interleukin.p<0.05 indicates a significant pairwise comparison.*p<0.05, **p ≤ 0.01, ***p ≤ 0.001.T A B L E 1Note: