Autophagic reprogramming of bone marrow–derived macrophages

Macro-autophagy is a highly conserved catabolic process among eukaryotes affecting macrophages. This work studies the genetic regulatory network involving the interplay between autophagy and macrophage polarization (activation). Autophagy-related genes (Atgs) and differentially expressed genes (DEGs) of macrophage polarization (M1–M2) were predicted, and their regulatory networks constructed. Naïve (M0) mouse bone marrow–derived monocytes were differentiated into M1 and M2a. Validation of the targets of Smad1, LC3A and LC3B, Atg16L1, Atg7, IL-6, CD68, Arg-1, and Vamp7 was performed in vitro. Immunophenotyping by flow cytometry revealed three macrophage phenotypes: M0 (IL-6 + /CD68 +), M1 (IL-6 + /CD68 + /Arg-1 +), and M2a (CD68 + /Arg-1). Confocal microscopy revealed increased autophagy in both M1 and M2a and a significant increase in the pre-autophagosomes size and number. Bafilomycin A increased the expression of CD68 and Arg-1 in all cell lineages. In conclusion, our approach predicted the protein targets mediating the interplay between autophagy and macrophage polarization. We suggest that autophagy reprograms macrophage polarization via CD68, arginase 1, Atg16L1-1, and Atg16L1-3. The current findings provide a foundation for the future use of macrophages in immunotherapy of different autoimmune disorders. Supplementary Information The online version contains supplementary material available at 10.1007/s12026-022-09344-2.


Introduction
Macrophages are major players in the immune system, and their phagocytic function contributes to host-pathogen defense mechanisms [1]. The activation of macrophages affects the quality of phagocytosis [2]. Autophagy is a highly conserved cellular catabolic process essential for cellular recycling that regulates phagocytosis in macrophages through modulation of the surface receptors [3][4][5][6][7]. Autophagy is also involved in the pathophysiology of many diseases such as neurodegenerative disorders [8], tumorigenesis [9], diabetes [10], and the immune response to infections [11]. The process starts with phagophore formation and elongation, autophagosome formation, and lysosomal fusion, followed by autolysosomal degradation [12].
According to their polarization state, inflammatory macrophages are classified into pro-inflammatory M1 macrophages and anti-inflammatory M2 macrophages [32][33][34]. Previous studies reported in vitro polarization of macrophages with IFN-γ, lipopolysaccharide, and interleukin 4 (IL-4) or IL-13 and showed high levels of IL-6 in the M2 phase [35]. The inhibition of the IL-6/ STAT3 pathway with anti-IL-6 treatment caused M2 to change into M1 type.
Here, we investigate how autophagy reprograms macrophage polarization, as the interplay between autophagy and macrophage polarization is poorly understood. Finding the targeted proteins that mediate the interplay between autophagy and macrophage polarization among a pool of autophagy-related proteins and hundreds of growth factors and proteins that regulate macrophage polarization is quite challenging. Therefore, we implemented a systems biology approach to narrow down the protein targets that mediate the interplay between the two processes. These target proteins were validated in vitro using bone marrow-isolated macrophages.

Ethical disclosure
All procedures were performed in compliance with the National Institutes of Health (NIH) guidelines for the Care and Use of Laboratory Animals (NIH Publications No. 8023, revised 1978), and according to Directive 2010/63/ EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes. All methods are reported in accordance with ARRIVE guidelines.
Isolation and characterization of bone marrow-derived monocytes Female C57B/6 J mice were euthanized by an overdose of ketamine xylazine followed by cervical dislocation. The femur and tibia were removed and rinsed in ethanol 70% for 5 min, followed by 1 × phosphatebuffered saline (PBS), 6.7 mM PO 4 , without calcium and magnesium. The tibia and femur were rinsed in Dulbecco's modified Eagle's medium: F12, DMEM:F12 with HEPES (25 mM), 1:1 mixture with 3.151 g/l glucose, with l-glutamine (Lonza, Basil, Switzerland) for 10 min.
The cell suspension was lysed with 1 × ammoniumchloride-potassium lysing buffer saline (Lonza, Basil, Switzerland) for 5 min to eliminate red blood cell and thrombocyte contamination. Following the lysis, the cell suspension was centrifuged for 5 min at 500 g. The cells were resuspended in lymphocyte separation medium (Lonza Basil, Switzerland) combined with DMEM/ F12 Complete Medium (DMEM/F12 + 10% FBS + 1% penicillin and streptomycin) and centrifuged at 500 g for 10 min. The cell suspension was collected, counted, and seeded at a density of 300,000 cells/well in 12-well plates (Greiner, Kremsmünster, Austria) and incubated for 72 h, at 37 °C and 5% CO 2 .
Cell viability and cytotoxicity assay Macrophages were seeded in 96-well plates (10,000 cells/well). MTT tetrazolium reduction assay was performed as previously reported [53]. In summary, following a 3-h incubation with MTT reagent, the media were removed, and DMSO was added to dissolve the formazan crystals. The cells were examined using an inverted microscope (Olympus 1X70, Tokyo, Japan), and absorbance was measured at 570 nm using a microplate reader (Ultrospec 3100 pro). Cell viability (%) was calculated based on the following equation: where Ab sample is the sample absorbance, Ab blank is the absorbance of blank, and Ab control is the absorbance of the control.
Autophagy assay On day 5, naïve macrophages (M0) were seeded at 96-well plates at a seeding density of Survival rate(%) = (Ab sample − Ab blank)∕(Ab control − Ab blank) × 100 10,000 cells/well for 48 h. Autophagy assay was performed according to the manufacturer's instruction (MAK138 f luorometric assay kit; Sigma-Aldrich, Saint Louis, MO, USA). The media was removed, and autophagosome detection reagent was added and incubated in the dark for 1 h at 37 °C and 5% CO 2 . Cells were washed gently by adding 100 µl of washing buffer, and the fluorescence intensity was measured (λ ex = 360/λ em = 520 nm).
Phagocytosis assay Naïve macrophages were seeded at day 5 into 96-well plates at a seeding density of 10,000 cells/well to contain a final volume 100 µl/well primed for 48 h to M1 and M2a lineages as previously mentioned. Cells were stained with MAK138 autophagosome detection reagent as mentioned earlier. E. coli top 10 bacteria were grown in LB broth liquid (purchased from Thermo Fisher Scientific, Waltham, MA, USA) and were added to the cells. Cells were stained with 1 µg/ ml 4′,6-diamidino-2-phenylindole·2HCl (DAPI) stain (Lonza, Basil, Switzerland) and examined under fluorescent microscopy (inverted fluorescent microscope; Leica Microsystems, Germany). Phagocytic events were counted for each condition.
Early apoptosis detection Macrophages were primed to M1 and M2a as previously described. SH-SY5Y neuroblastoma cells (ATCC CRL-2266) were cultured in conditioned media from naïve macrophages (M1 and M2a) for 24 h. Cells were fixed with 4% PFA and permeabilized for 10 min with 0.3% triton X-100. Cells were washed and stained with DAPI and mounted on slides. Cells were examined under the microscope (inverted fluorescent microscope; Leica Microsystems, Germany). Cells treated with 20 ng/ml cisplatin were used as a positive control.
To detect intracellular trafficking of Atg7, Atg16l1, and LC3B inside the cytoplasmic or nuclear compartment, an automated spot detector plug-in SICE was used as described by Bayle et al. [55]. Images were taken by a fluorescent microscope (Leica Microsystems, Germany) and imported to ImageJ® software. A minimum of 8 images was counted for each condition.
Flow cytometry Macrophages were collected and washed with 0.5% FBS in 1 × PBS and centrifuged at 350 g for 5 min. Cells were stained with mouse-specific antibody conjugate eFlour660 CD68 and Alexa Flour 488 conjugated arginase 1 (eBioscience, USA) for 30 min and washed with 1 × PBS at 500 g for 10 min. Unstained samples were used as a negative control. Samples were measured and gated on a flow cytometer (CytoFLEX, Beckman Coulter, USA, using two lasers: red laser (with an excitation wavelength of 660 nm) for allophycocyanin (APC) and blue laser (with an excitation wavelength of 488 nm) for fluorescein isothiocyanate (FITC).

Statistical analysis
Statistical analyses were carried out using GraphPad Prism® software. Data was expressed as mean ± standard deviation, or median and range were used for data expression. All tests were two-tailed. Post hoc tests and one-way ANOVA were used to compare the differences of mean values between different groups. p values that were less than 0.05 were considered statistically significant.

In silico analysis of autophagy-related genes
We used a network-based systems biology approach to model the interplay between the complex signaling pathways of autophagy and macrophage polarization. The analysis of the different databases identified common significantly enriched pathways and common regulatory transcription factors that co-regulate both transcription factors and Atgs and M1-M2-DEGS (Supplementary Data).

In vitro isolation and polarization of bone marrow macrophages
Murine bone marrow monocytes were isolated and differentiated to M0 using 20% L929 conditioned media. On day 5, type II interferon-γ was used (1250 IU/ml) combined with For M2a, IL-4 was used (2500 IU/ml) in combination with LPS (100 ng/ml) for 48 h. The three lineages were characterized using flow cytometry with three markers: interleukin-6, CD68, and arginase 1 (Fig. 1, Supp. Data Fig. 2S).

Interferon-γ promotes IL-6 expression in M1 lineage, while IL-4 inhibits the IL-6 expression in M2a lineage
Flow cytometry studies show that interferon-γ stimulated M1 lineage expressing the phagocytosis marker IL-6 significantly (56%) compared to M2a lineage (37%) (Fig. 3). However, IL-6 expression was also high in the control M0 lineage. Also, the fluorescence intensity for IL-6 protein showed that M1 lineage had the most significant increase in IL-6 protein expression. Surprisingly, the conditioned media of M2a 7-day macrophages showed cytotoxic activity on neuroblastoma cell line SH-SY5Y (Supp. Data Fig. 11S).
Immunolocalization using Atg16L1 in M0, M1, M2a, and M0 + Earle's balanced salt revealed that no autophagosomes were observed in the M0 control and in M1 macrophages. However, we observed pre-autophagosomes in M2a lineage and in the positive autophagy M0 cells treated with EBS. M2a lineage showed a significant increase in both nuclear pre-autophagosome numbers (p value = < 0.05) and cytoplasmic Atg16L1 size. This supports our gene expression data that showed increased fold change of Atg16L1-1 gene variation M2a at 7-day polarization ( Fig. 5).

INF-γ increased Atg7 expression in M1 cells and increased pre-autophagosome size
Immune co-localization studies show Atg7 expression as pre-autophagosomes distributed in the cytoplasmic compartment in M1 and M2a lineages. Confocal images revealed a significant number of pre-autophagosomes formed in M1 and M2a lineages. However, M0 control showed the largest size of pre-autophagosomes (Fig. 6). Statistical analysis of Atg7 and the pre-autophagosome number per cell show that there was no significant difference between M1 and M2a (300 spots/cell). M1 cells showed an increased preautophagosome size to more than 1 µm in diameter (n = 6 images, p value = < 0.05) (Fig. 6).
Relative fold change of gene expression normalized to GAPDH as endogenous control shows that the fold changes relative to GAPDH in M0 were as follows: mean = 3.7 and 2.36 folds, ± 0.46 and ± 0.56 for M1 and M2a, respectively (Fig. 6).
Therefore, INF-γ promoted the expression of the Atg7 protein and mediated upregulation of Atg7 gene expression in M1 and M2a cells. In contrast, INF-γ and lipopolysaccharide increased Atg7 protein and messenger RNA (mRNA) in M1 lineage.

Autophagy-associated protein complex LC3A and LC3B expression increased in M1 and M2a macrophages
The MAP1-LC3s or LC3A and LC3B quantification showed that the distribution of autophagosomes inside the nuclear and cytoplasmic compartments is not uniformly distributed. However, autophagosomes were not localized in the nucleus in M1 and M0 control. The average number of basal autophagosomes in M0 was 1800. In M1, it was 2436 spot and in M2a increased to 2471 (Fig. 12S). Remarkably autophagosome aggregations were also observed. Flow cytometry single-cell quantification showed a significant increase in M1 and M2a cells (p value = 0.01-< 0.05, Fig. 7).
Finally, mRNA levels of LC3B but not LC3A increased in M1 to 4 folds and in M2a to 3 folds, respectively. Collectively, INFG + LPS induced macro-autophagy inside M1 and IL-4 + LPS induced macro-autophagy in M2a cells.
High autophagosome aggregates were formed in both M1 and M2a lineages compared to M0 control (Fig. 8). Relative fluorescence intensity of LC3A and LC3B showed a significant increase in M1 and M2a lineages compared to M0 control (Supp. Data Fig. 12S).

Increased Smad1 gene expression in M1 and M2a lineages
We report Smad1 as one of our predicted transcription factors and its downstream targets, IL-6, and MAPLC3A genes. Smad1 was downregulated in M1 and M2a at 7-day polarization compared with 14-day polarization results. Flow cytometry showed a significant difference in Smad1 protein expression at 7-day polarization in M2a compared to M0 and M1. Fold change in M2a cells at 7-day polarization was only 1.5 folds (p values = < 0.05, Fig. 9). However, there was significant overexpression of Smad1 in both M1 and M2a lineages at 14-day polarization (p value = < 0.05, Fig. 9).

Autophagy inhibitor bafilomycin A significantly increased CD68 and arginase 1 expression in M0, M1, and M2a Lineages
Autophagy inhibition using bafilomycin A (200 nM) in M0, M1, and M2a cell lineages showed a high expression pattern for both arginase 1 and CD68 in M0, M1, and M2a cells. Also, bafilomycin A increased the percentage of double-positive (CD68/arginase1) cells in both M1 and M2a. Surprisingly, autophagy inhibition showed a great increase in arginase 1 expression more than 50% in M0 lineage at 7-day polarization compared to M0 at 7-day polarization with the normal basal autophagy activity (Figs. 10 and 14S).

Autophagy induction decreased the phagocytosis activity of M2a but not M1
The average number of phagocytic events in M2a lineage showed decreased phagocytic activity. However, no significant effect on M0 and M1 lineage was observed. Immune staining studies using Mak38 autophagy detection kit showed that autophagy decreased the phagocytic activity of M2a compared to M1 and M0 (Supp. Data  Fig. 15S).

Discussion
Autophagy depends on the formation of double-membrane autophagosomes that fuse with the lysosome to degrade pathogens, proteins, and organelles. Both phagocytosis and autophagy are interdependent processes. The interplay between autophagy, macrophage activation, and phagocytosis is still poorly understood [60,61].
In this study, we dissected both the autophagy and macrophage activation process to understand the nature of this interplay. We were able to identify a list of common pathways, transcription factors, and target proteins that mediate this interplay (Supplementary Data). We further validated these targets in an in vitro study.
The predicted target proteins, Atg7, and Atg16L1 serve as central proteins for several signaling pathways in autophagy, macrophage polarization, and phagocytosis. Nevertheless, more experimental validation is needed for other predicted targets. Atg16L1 mediates the preautophagosome formation, which is essential for interaction with the Atg5-Atg12 complex that mediates the conjugation with PE [70].
Bone marrow-derived macrophages are a heterogeneous population. To characterize the phenotypes of the isolated bone marrow-derived macrophages and the activated macrophages in vitro, we investigated the expression of phagocytic markers CD68, IL-6, and arginase 1 among various macrophage populations. CD68 is a cell surface heavily glycosylated glycoprotein localized near the endosomal/lysosomes compartment, that is commonly used as a phagocytic marker in dendritic cells and strongly expressed in total macrophages, including M1 and M2 [78,79]. It is also a marker of tumor-associated macrophages [80]. M0 and M1 macrophages were confirmed by the high expression of CD68 (more than 60%).
IL-6 is a pro-inflammatory cytokine that we predicted to mediate the interplay between autophagy and macrophage activation. Interferon-γ and lipopolysaccharide combination promoted the expression of IL-6 in M1 lineage. Besides phagocytosis, cytotoxic activity is one of the characteristics of bone marrow-derived macrophages [84].
In the current study, INF-γ and IL-4 in combination with LPS significantly induced macro-autophagy in both M1 and M2a lineages at 7-day polarization. Previous reports [85,86] showed that INF-γ induced autophagy in hepatocellular carcinoma through increased LC3A and LC3B expression. Increased autophagy activity was found to increase the phagocytosis of Mycobacterium tuberculosis by the INF-γ signaling pathway [87]. IL-4 induced macro-autophagy in antigen-presenting B cells and is linked to asthma pathophysiology [88]. Finally, it is noteworthy to mention that IL-4 boosted autophagy induction to form LC3A and LC3B aggregates.
Our results show that the 14-day polarization resulted in the loss of arginase expression and increased autophagy-related gene expression Atg16L1-1. Since arginase 1 is a phagocytic marker for M1 and M2a, loss of expression of Arg-1 indicates loss of activation in M1 and M2a lineages at 14-day polarization [89]. Interestingly, Atg16L1-1 alpha and Atg16L1-3 gamma variant showed an increase in M2a 7-day poloarization than in M2a 14-day polarization. These studies suggest that high autophagy activity at 14-day polarization can attenuate arginase 1 expression. However, previous reports [90] Fig. 6 Immune co-localization of Atg7 using confocal microscopy. Three-dimensional reconstruction of Z-stack confocal microscopy images revealed a significant number of pre-autophagosomes formed at M1 and M2a lineages. M0 control showed the largest size of preautophagosomes. Atg7 expression is seen as green or yellow dots (pre-autophagosomes) distributed in the cytoplasmic compartment (red arrows). J shows the count of Atg7 + pre-autophagosomes per cell (n = 6, p value = < 0.05). No significant difference between M1 and M2a was seen. K shows a violin plot of Atg7 pre-autophagosome size in µm 2 measured using ImageJ.®. M1 showed an increased spot size (n = 6, p value = < 0.05). L shows Atg7 gene expression was significantly upregulated in M1 cells (n = 4, *p value = < 0.05, **p value = <0.01, *** p value = < 0.001) ◂ show that autophagy is required for arginase 1 expression in alternatively activated M2a at 7-day polarization.
Phagocytosis assay was performed to test the ability of M0, M1, and M2a cells to engulf heat-killed E. coli bacteria.
As mentioned earlier, Atg16L1 is the most important hub protein in macro-autophagy and macrophage polarization. Overexpression of Atg16L1-3 and Vamp7 in M2a at 7 days of polarization and the increased number of cytoplasmic pre-autophagosomes suggest that Atg16L1 is essential for IL-4-induced macro-autophagy in M2a (Arg-1 + /CD68 +) cells. Also, autophagy induction decreased the phagocytic ability of M2a (Arg-1 + /CD68 +) cells.
To better understand the interplay between autophagy and phagocytosis, we blocked the autophagosome and lysosomal fusion with autophagy inhibitor bafilomycin A as previously described [96]. Our results indicate that bafilomycin  E, and H). C, F, and I show immune co-localization of cytoplas-mic autophagosome (yellow to green spots) in M0, M1, and M2a lineages. LC3A and LC3B nuclear autophagosomes (yellow dots inside the nuclear compartment) (J, K). The number of autophagosomes was counted (L) inside the cytoplasm and in the nuclear using ImageJ® increased CD68 and arginase 1 expression in M0, M1, and M2a lineages, while autophagy induction decreased phagocytosis of M2a but not M1. Other studies reported bafilomycin-induced autophagy inhibition and the knockdown of autophagy-related protein Atg5 promoted M2 polarization [97].
Other studies [98] reported that autophagy inhibition by 3-MA (autophagy inhibitor) increased the phagocytic ability of macrophages and rescued mice from methicillinresistant Staphylococcus aureus (MRSA) bacterial infection. Also, Atg16L1 mutation increased the phagocytosis ability of monocytes isolated from Crohn's disease patients [99]. Therefore, we suggest that Atg16L1 might serve as a therapeutic target for the treatment of altered phagocytosisrelated diseases such as bacterial infection, inflammation, lupus nephritis, and cancer. Further studies for this target protein are needed.

Conclusion
Our findings suggest that autophagy induction decreased the phagocytosis activity of M2a but not M1 macrophages. We also suggest that autophagy reprograms macrophage Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.