Simple propagation method for resident macrophages by co-culture and subculture, and their isolation from various organs

Mø derived from the mouse liver, spleen, lung, and brain showed high propagation when co-cultured with stromal cells of the respective organs in standard culture media [Dulbecco’s modified Eagle’s medium (DMEM) for the liver, spleen, and lung, and DMEM/F12 for the brain] including 10% foetal bovine serum (FBS) without any additional growth factors for Mø such as CSF-1 and CSF-2. By changing the culture media every 4–6 days, primary stromal cells, including Mø, reached over-confluence usually within 2–3 weeks for liver, spleen, and lung cells, and within 3–4 weeks for brain cells. The over-confluent cells formed a multi-layered structure on a standard tissue culture dish. The cells were then subcultured until reaching over-confluence again, which occurred within a similar period of time. The cells were considered to be Mø (designated as cMø) according to observations of their propagation behaviours by microscopy, which were similar among the cells cultured from the four organs (Fig. 1a–d). Within ten days after seeding, there were three morphological types of cMø apparent on the dish: flat cells with a few long processes adhering to large stromal cells, elongated cells with a few long processes adhering directly to the dish surface in a relatively low cell density area, and round or fusiform cells in a densely populated area on the dish surface (Fig. 1b, c). Two to three weeks after the subculture, when stromal cells became over-confluent, there were two morphological types of cMø observed in multi-layered cells: round, small cells located in the top layer, adhering to stromal cells forming the middle cell layer; and fusiform cells with a few long processes adhering directly to the dish surface, in which the stromal cells forming the middle cell layer unexpectedly detached from the dish surface (Fig. 1a, d).

Mø and stromal cells in co-culture derived from the liver, spleen, and lung that were subcultured for more than eight passages propagated and became over-confluent as observed for the primary cells, whereas cells from the brain showed a remarkable decrease in propagation after more than three passages. The over-confluent co-cultured cells were subcultured or frozen at a cell dilution ratio of 1:3 for the liver, spleen, and lung and at 1:2 for the brain. The thawed and cultured frozen cells were treated to the same cultivating condition. Frozen cells from the liver, spleen, and lung propagated similar to the unfrozen cells, whereas those from the brain propagated slowly and were therefore not suitable for Mø propagation.

The Mø were separated from the other stromal cells in co-culture according to their adhesive property to bacteriological Petri dishes, in which only the Mø should adhere to the dish. Within a few days of seeding co-cultured over-confluent cells, small round cells with a few processes, i.e. Mø adhering to the dish surface, were observed; cells of other shapes rarely adhered to the dish, and cell aggregates floating in the media were also evident (Fig. 1e–h). These cell aggregates were easily removed by washing with conditioned media. The cell density of Mø was almost unchanged, with or without cell aggregates, in the dishes when culture continued for several days. We usually collected more than 1.5 × 10⁶ adherent cells per 10-cmø bacteriological Petri dish.

Phagocytosis of fluorescent beads was evaluated to precisely determine the percentage of Mø in the collected segregated cells. During incubation, almost all of the cells segregated from the liver, spleen, lung, and brain stromal cells phagocytosed the fluorescent beads (Fig. 2a). Most of the cells had numerous beads in their cytoplasm, and the cytoplasm of some cells was completely filled with beads. This demonstrated that the Mø propagated in co-culture possess a high phagocytic property. The bead-positive and negative cells were counted to estimate the percentage of Mø in the segregated cells. We defined cells phagocytosing more than two beads as bead-positive cells and counted more than 700 cells per sample. Overall, these cells comprised more than 98.8% Mø from the liver, spleen, and brain as well as more than 97.6% Mø from the lung (Fig. 2b). Thus, Mø segregation according to their property of adhesion to the bacteriological Petri dish represents a simple method to purify Mø from stromal cells from various organs in co-culture.

The identity of the Mø segregated from subcultured liver, spleen, lung, and brain stromal cells was further confirmed based on the expression of Mø markers using flow cytometry: integrin αM subunit (CD11b), integrin αX subunit (CD11c), scavenger receptor class D (CD68), CD86 (B7–2), CSF-1R (CD115), CSF-2R (CD116), Siglec-1 (CD169), C-X-C chemokine receptor type 4 (CD184), and C-type mannose receptor 1 (CD206), EGF-like module-containing mucin-like hormone receptor-like 1 (F4/80), and major histocompatibility complex class II (MHC II). The liver, spleen, lung, and brain Mø consisted of a single population based on histograms of the marker expression distribution (Figs. 3, 4, 5, 6): all histograms showed a single peak except for those of CD11c, which included both CD11c-positive and -negative fractions. Moreover, the respective Mø showed similar expression patterns of these molecules, and the patterns were quite similar between the liver and spleen Mø. As a whole, all Mø showed high expression of CD11b, CD68, CD169, and CD206; substantial or high expression of CD86, CD115, CD184, and F4/80, except for faint expression of CD86, CD115, and CD184 in the brain Mø; and faint or almost negative expression of CD116 and MHC II. These expression analyses clearly confirmed that the cells segregated from the co-culture of all four organs were Mø.

We further examined whether polarisation of the segregated Mø to classical M1 and alternative M2 Mø occurred in response to stimulation with cytokines and the Toll-like receptor ligand. We used the spleen Mø as a representative of the four Mø. The combination of LPS and interferon-γ (IFN-γ) was used as an inducer for M1 polarisation, and IL-4 was used as an inducer for M2 polarisation as previously reported for monocyte-derived Mø [8, 17]. We also employed CD11c and CD86 as M1 polarisation markers, and CD206 as an M2 polarisation marker as well as CD11b as a control pan-Mø marker according to previous reports [18, 19].

Flattened and/or elongated Mø frequently appeared on the bacteriological Petri dish after treatment with LPS plus IFN-γ for 24 h, while thin elongated Mø (with few long processes) occupied the dish after treatment with IL-4 (Fig. 7a). Flow cytometry revealed that LPS plus IFN-γ clearly upregulated CD86 expression, while IL-4 clearly upregulated CD11c and CD206 expression (Fig. 7b). These findings indicated that the propagated spleen Mø responded to the typical M1 and M2 polarisation inducers. Moreover, the polarisation properties likely differ between spleen and monocyte-derived Mø because CD11c is used as an M1 polarisation marker [18, 19]. Further, the cell adhesion capacity of the spleen Mø most likely increased in response to the M2 inducers because CD11c is an integrin αX subunit, which together with the ß2 subunit binds to integrin ligands such as ICAM-1, fibrinogen, and collagen [20].

Stromal cells from the respective organs were harvested from specific-pathogen-free ICR male mice. The mice obtained from Japan SLC, Inc. (Hamamatsu, Japan) had been maintained under a standard housing condition in clean-grade environment on a 12-h light-dark cycle, and fed with standard diet and water ad libitum. In total, 18 mice were used. The animal experimentation protocol was approved by the Animal Research Committee of the Osaka Prefecture University. All experiments were performed in accordance with relevant guidelines and regulations of the Osaka Prefecture University.

Mice were sacrificed with an overdose of pentobarbital injected intraperitoneally (150 mg/kg body weight; Somnopentyl, Kyoritsu Seiyaku, Tokyo, Japan), and then intracardially perfused with Ca/Mg-free HBSS (Sigma-Aldrich, St Louis, MO, USA) supplemented with 50 U/mL heparin (Mochida Pharmaceutical, Tokyo, Japan) to remove the blood. The liver, lung, and spleen from 8-week-old mice and the brain from 4-week-old mice were aseptically dissected and immediately dipped in ice-cold HBSS. The gallbladder from the liver, adipose tissues around the splenic and pulmonary hilum from the spleen and lung, respectively, and the meninges, brainstem, and cerebellum from the brain were then removed. The lung was also cleared of alveolar cells, including alveolar Mø, by bronchoalveolar lavage. A 23-gauge intravenous catheter was inserted into the trachea, and 1.6 mL HBSS was injected and immediately withdrawn a few times. Half to about one-third of the whole liver as well as the whole spleen, lung, and cerebrum was minced with a razor blade into approximately 1-mm³ pieces and transferred to 15-mL conical tubes containing cell dispersion enzyme solution: 12 mL of 0.5 mg/mL Collagenase Type IV (Sigma-Aldrich) for liver tissues; 7.5 mL and 10 mL of 0.5 mg/mL Collagenase Type IA (Sigma-Aldrich) for spleen and lung tissues, respectively; and 10 mL of 1.0 mg/mL dispase (Thermo Fisher Scientific, Waltham, MA, USA) for brain tissues in 20 mM HEPES (Dojindo, Kumamoto, Japan)-buffered HBSS containing 1 mM CaCl₂. The tissues were then digested at 37 °C for 40–60 min under gentle stirring at 120 rpm with one change of the digestion solution until tissue pieces were no longer visible. After washing with HBSS, cell/tissue suspensions were further dispersed by pipetting. The suspensions were sedimented at 100×g for 5 min (Model 2410, Kubota, Tokyo, Japan) and resuspended in HBSS to remove cell debris. Cells/tissues from half to one third of the whole liver, spleen, lung, and cerebrum per mouse were plated on three, one, three, and two 10 cm ø tissue culture dishes (AGC Techno Glass, Haibara, Japan), respectively. The dishes were coated for 2.5 h at 37 °C with or without collagen (Nitta Gelatin, Yao, Japan) in HBSS at 1.6 μg protein/cm² for the liver, spleen, and lung cells and with poly-l-ornithine (Sigma-Aldrich) in HBSS at 1.9 μg protein/cm² for the brain cells. Coated dishes were used for primary brain cell culture and in some cases for primary liver, spleen, and lung cell culture, but not for subculture. Cells were cultured in 12 mL DMEM (Sigma-Aldrich) containing 10% FBS (Sigma-Aldrich), 100 U/mL penicillin, and 100 μg/mL streptomycin (pen/strep; Sigma-Aldrich) (DMEM-FBS) for the liver, spleen, and lung as well as 12 mL DMEM/Ham’s nutrient mixture F-12 containing 10% FBS and pen/strep (DMEM/F12-FBS) for the brain. Cells were maintained in a humidified 5% CO₂/95% air incubator at 37 °C. The medium was changed after a few hours and again after 1 day to remove non-adherent cells and cell debris, and thereafter every 4–6 days until dishes were covered by multi-layered cells composed of Mø and other stromal cells such as fibroblasts or astrocytes. Over-confluent cells were detached by 0.1% trypsin/1 mM EDTA in HBSS at 37 °C for 10–15 min followed by pipetting. Subsequently, cells at a dilution ratio of 1:3 for the liver, spleen, and lung and at 1:2 for the brain were subcultured or frozen at − 80 °C in a cell suspension with Bambanker (Nippon Genetics, Tokyo) as a cryopreservative and maintained in the same medium until they became over-confluent again.

Co-cultured, over-confluent cells obtained from the liver and lung up to five passages, from the spleen up to eight passages, and from the brain up to two passages were used for separation of Mø. Cells harvested from a 10-cmø tissue culture dish at over-confluence were seeded in a 10-cmø bacteriological Petri dish (As One, Osaka, Japan) containing 10 mL DMEM-FBS. After one to a few days, when the Mø selectively adhered onto the dish surface and stromal cells formed aggregates floating in the medium, the cells were washed with conditioned media to remove cell aggregates. The adherent cells were then detached by 5 mL of 5 mM EDTA in 10 mM HEPES-buffered HBSS (HEPES-HBSS) at 37 °C for 10–15 min followed by pipetting. The cell suspension was passed through a cell strainer (BD Falcon, Franklin Lakes, NJ, USA) to remove cell aggregates, sedimented at 220×g for 5 min, suspended in phosphate-buffered saline (PBS) containing 1% bovine serum albumin (BSA; Sigma-Aldrich), 2 mM EDTA, and 0.01% NaN₃ (BSA/EDTA-PBS), and the number of cells was calculated and used in experiments.

Cells (2.5 × 10⁵/0.5 mL DMEM-FBS) were placed in a 5-mL tube that was siliconised (Fuji-Rika Industries, Osaka, Japan) according to the manufacturer’s protocol to prevent adhesion to the tube wall. After addition of 1.0 μL fluorescent yellow-green-conjugated latex beads (mean diameter, 1.0 μm; Sigma-Aldrich), the cells were incubated at 37 °C for 2 h with gentle shaking at 18 rpm on a seesaw-type shaker (Wave SI slim; Taitec, Koshigaya, Japan), washed three times with HBSS, and plated on a 3.5-cmø glass-bottom dish (AGC Techno Glass) with 1.5 mL DMEM-FBS for ~ 2 h until almost all cells adhered to the surface. After fixation with 10% formalin (Kanto Chemical, Tokyo, Japan) in PBS for more than 10 min at room temperature (RT: 22–28 °C), phase-contrast and green fluorescence images of the same fields were captured using a 10× and 20× objective lens (IX71; Olympus, Tokyo, Japan). Cells engulfing more than two latex beads were denoted as Mø. We counted more than 700 cells per sample, and the percent of Mø in each organ was calculated from independent experiments (four mice and four experiments for the liver, spleen, and lung cells; six mice and six experiments for the brain cells). Statistical analyses were performed with the statistical software package Statcel (OMS Publishing Inc., Tokorozawa, Japan) implemented in Excel. Data are presented as means ± SD, and differences between groups were evaluated with unpaired t-tests. P values less than 0.05 were considered significant.

Flow cytometry was used to determine the expression of Mø markers (CD11b, CD11c, CD68, CD86, CD115, CD116, CD169, CD184, CD206, F4/80, and MHC II) in cells segregated using bacteriological Petri dishes according to the method of Mukai et al. [35] with some modifications. Cells were prepared at a concentration of 1 × 10⁶ cells/mL in BSA/EDTA-PBS and fixed in 5% formalin in BSA/EDTA-PBS for 20 min at RT. After washing with BSA/EDTA-PBS, the cells were permeabilised in 0.2% saponin (Nacalai Tesque, Kyoto, Japan) in BSA/EDTA-PBS for 5 min at RT. To avoid non-specific Fc-gamma receptor-mediated binding of fluorochrome-conjugated antibodies, cell suspensions (~ 2.0 × 10⁵ cells/50 μL) were pre-treated with 0.5 μg of anti-mouse CD16/32 antibody (rat IgG2b; Tonbo Biosciences, San Diego, CA, USA) for 10 min at RT. To the 50 μL cell suspension, we added 0.5 μg FITC-conjugated anti-CD11b antibody (rat IgG2b; Tonbo), 0.25 μg APC-conjugated anti-CD11c antibody (hamster IgG; Tonbo), 0.15 μg FITC-conjugated anti-CD68 antibody (rat IgG2a; Miltenyi Biotec, Bergisch Gladbach, Germany), 0.125 μg FITC-conjugated anti-CD86 antibody (rat IgG2a; Tonbo), 0.5 μg FITC-conjugated anti-CD115 antibody (rat IgG2a; Tonbo), 0.1 μg APC-conjugated anti-CD116 antibody (rat IgG2a; R&D Systems, Minneapolis, MN, USA), 0.15 μg APC-conjugated anti-CD169 antibody (recombinant human IgG1; Miltenyi), 0.15 μg FITC-conjugated anti-CD184 antibody (recombinant human IgG1; Miltenyi), 0.25 μg APC-conjugated anti-CD206 antibody (rat IgG2b; Thermo Fisher Scientific, Waltham, MA, USA), 0.5 μg APC-conjugated anti-F4/80 antibody (rat IgG2a; Tonbo), and 0.25 μg APC-conjugated anti-MHC II antibody (rat IgG2b; Tonbo) according to the manufacturer’s instructions, followed by incubation for 10 min at RT. After washing, 20,000 or 30,000 cells were analysed for their expression characteristics using a flow cytometer (S3 Cell Sorter; Bio-Rad Laboratories, Hercules, CA, USA). As controls, we used cell suspensions that were pre-treated with the anti-mouse CD16/32 antibody and then treated with the same fluorochrome-labelled isotype control antibody of the same amount as the test antibody. Expression of marker molecules was determined from more than three independent experiments in cells propagated from each organ derived from more than three mice (four mice and four experiments for the liver, spleen, and brain Mø; three mice and three or four experiments for the lung Mø).

Flow cytometry was also used to examine M1 and M2 polarisation in the spleen Mø as described above. The cells (8 × 10⁵) were plated on a 10-cmø bacteriological Petri dish with 10 mL DMEM-FBS. After 1 day of seeding, we added 20 ng/mL LPS (Sigma-Aldrich) plus 50 ng/mL IFN-γ (PeproTech, Rocky Hill, NJ, USA), 10 ng/mL IL-4 (Tonbo), or vehicle (HBSS) to the dish; 24 h later, the cells were detached by 5 mM EDTA. Cell suspensions (~ 2.0 × 10⁵ cells/50 μL) pre-treated with 0.5 μg of anti-mouse CD16/32 antibody were incubated with a mixture of 0.5 μg of FITC-conjugated anti-CD11b antibody and 0.25 μg of APC-conjugated anti-CD11c antibody, or 0.125 μg of FITC-conjugated anti-CD86 antibody and 0.25 μg of APC-conjugated anti-CD206 antibody.