Materials

Mice C57BL/6J mice (6-8 week) National Laboratory Animal Center (NLAC) Antibodies PE anti-mouse/human CD11b Antibody BioLegend, San Diego, USA

PE anti-rat IgG2b, κ Antibody BioLegend, San Diego, USA APC/Cy7 anti-mouse F4/80 Antibody BioLegend, San Diego, USA APC/Cy7 anti-rat IgG2a, κ Antibody BioLegend, San Diego, USA FITC anti-mouse CD80 Antibody BioLegend, San Diego, USA FITC anti-Armenian Hamster IgG Antibody BioLegend, San Diego, USA APC anti-mouse CD206 Antibody eBioscience, San Diego, USA APC anti-rat IgG2a, κ Antibody eBioscience, San Diego, USA Cytokines Recombinant Murine IL-4 Peprotech, Rocky Hill, USA

Recombinant Murine IL-13 Peprotech, Rocky Hill, USA Recombinant Murine IFN-γ Peprotech, Rocky Hill, USA

Lipopolysaccharide Sigma Aldrich, Saint Louis, USA Primers

(qPCR)

All primers are described in Table 1 Purgo, Taipei, Taiwan

siRNA 10 µM stock Santa Cruz Biotechnology, Texas, USA 10 mL Penicillin/ Streptomycin (100X) 10 mL Sodium pyruvate (100 mM, 100X)

Chemicals

and

Reagents

Name Company

2-Mercaptoethano (2-ME) Sigma Aldrich, Saint Louis, USA

FBS Life Technologies,California, USA

DMEM Gibco,Waltham, USA

NA2 EDTA J.T, Baker,New Jersey, USA

NH4Cl Mallinckrodt, Dublin, Ireland

Sodium azide Sigma Aldrich, Saint Louis, USA

L-Glutamine HyClone, Texas, USA

Penicillin/Streptomycin HyClone, Texas, USA Non-essential amino acid (NEAA) HyClone, Texas, USA

Sodium pyruvate HyClone, Texas, USA

Polybrene Thermo, Massachusetts, USA

isol-RNA lysis reagent 5-PRIME, California, USA direct-zol RNA miniPrep Zymo Research, Irvine, USA Foxp3/Transcription Factor Staining Buffer Set Thermo, Massachusetts, USA TransIT-TKO® Transfection Reagent Mirus Bio, Madison, USA

2. Methods

2.1 Isolation and development of bone marrow-derived macrophage (BMDMs)

Female C57BL/6 mice at 6-8 week old were purchased from National Laboratory

Animal Center. Bone marrow cells were isolated from femur and tibia bones of mice with incomplete DMEM (Dulbecco’s modification of Eagle’s medium, Gibco) by using

23G needles, and centrifuged at 4°C with 800 xg for 5 mins. Red blood cells were lysed by 900 μL RBC lysis buffer (10 mM Tris-HCl and 0.83％ NH4Cl in ddH2O, pH 7.2) for 3 mins on ice, and 100 μL 10X dPBS (Gibco) was added to cells. Cells were then centrifuged at 4°C with 800 xg for 5 mins, and re-suspended with complete DMEM containing 10% fetal bovine serum (FBS), 1X penicillin/streptomycin (HyClone), 1X

L-Glutamine (HyClone), 1X NEAA (non-essential amino acid, HyClone), 1X sodium pyruvate (HyClone) and 5 μM 2-ME (2-Mercaptoethanol, Thermo). Cells (5x10⁶) were

seeded in 10-cm² petri dishes (α-Plus) at 37°C in 5% CO2 incubator for 7 days with 20% L929 cell supernatant in complete DMEM. The cells were stained with

PE-conjugated CD11b (Biolegend) and APC/Cy7-conjugated F4/80 (Biolegend) in 100 μL FACS buffer on ice for 30 mins in the dark, and analyzed with FACSCanto (BD)

after washed twice with FACS buffer (PBS, 1% sodium azide, 1% FBS, and 0.05 M EDTA). PE-conjugated rat IgG 2b, κ (Biolegend) and APC/Cy7-conjugated rat IgG 2a, κ (Biolegend) were used as isotype controls. Data were quantified by software FlowJo

7.6.1 (Tree Star, Oregon, USA). Concentration of all conjugated antibodies was 0.1 μg /100 μL FACS buffer.

2.2 M1 and M2 polarization in vitro

After cultured in 20% L929 cell supernatant for 7 days, bone marrow-derived macrophages (BMDMs) were collected and further seeded in 12-well and 6-well plates (Thermo) at a density of 1x10⁶/mL per well for flow cytometry and 2.5x10⁶/mL per well for real time PCR, respectively. For macrophage differentiation, BMDMs were treated with LPS (1000 ng/mL) and IFN-γ (20 ng/mL, Peprotech) for M1 polarization, IL-4 (20 ng/mL, Peprotech) and IL-13 (20 ng/mL, Peprotech) for M2 polarization, and complete medium only as M0 macrophages for 24 hours.

2.3 Characterization of M1 and M2 macrophages by flow cytometry

To characterize M1 and M2 subsets, the differentiated macrophages were stained with

PE-conjugated CD11b, FITC-conjugated CD80 and APC-conjugated CD206 (Biolegend) in 100 μL FACS buffer at 4°C for 30 mins, and washed twice with FACS buffer. Cells

were then resuspended in FACS buffer and analyzed with FACSCanto (BD).

PE-conjugated rat IgG 2b, κ (Biolegend), FITC-conjugated rat Armenian Hamster IgG

(Biolegend) and APC-conjugated rat IgG 2a, κ (eBioscience) were used as isotype controls.

2.4 Characterization of M1 and M2 macrophage by real-time PCR

In brief, total RNAs were extracted from M0, M1 and M2 macrophages by using isol-RNA lysis reagent (S-PRIME) and direct-zol RNA miniPrep (Zymo Research). The quality and quantity of extracted RNA were measured by DS-II⁺ spectrophotometer (DeNoyix). Moreover, complementary DNAs (cDNAs) were synthesized from 700 ng RNA with 0.5 nM Oligo dT primer (Purigo Biotech, Inc.) by MMLV high performance

reverse transcriptase (MMLV HP RT, Epicenter Biotechnologies). cDNAs were added in a total volume of 10 μL SYBR green mixture (Bioline) and measured by PikoReal 96

Real-Time PCR System (Thermo Scientific). Primers for genes are described in Table 1.

The Ct (cycle threshold) values were determined by PikoReal Software 2.0 (Thermo Scientific). Relative mRNA expression were normalized with β-actin, a housekeeping gene and an internal control.

2.5 Small interfering RNA (siRNA) transfection

BMDM cells were transfected using TransIT-TKO transfection reagent (Mirus Bio, Madison, WI). In brief, cells were plated into six-well plates 24 hrs to obtain 60-80%

confluence before transfection. siRNA complex including 10 μL TransIT-TKO transfection reagent and 100 nM siRNA per sample was incubated in serum-free DMEM (SFM) containing 1X L-Glutamine (HyClone), 1X NEAA (non-essential amino acid, HyClone), 1X sodium pyruvate (HyClone) and 5 μM 2-ME (2-Mercaptoethanol, Thermo) for 25 mins. TransIT-TKO Reagent/siRNA complex mixture was then added to the cells and the dish was gently rock back and forth to distribute the complexes evenly.

Cells were incubation at 37°C in 5% CO2 for 48 hrs for further experiments.

2.6 Enzyme-linked immunosorbent assay (ELISA) for IL-1β and TNF-α

IL-1β or TNF-α capture antibodies (1X, Biolegend) was coated with 50 μL coating buffer (0.05 M Carbonate-Bicarbonate, pH 9.6) each well on Nunc™ 442404 Nunc-Immuno 96-Well plate (Thermo Scientific) at 4°C overnight. After the plate was washed by 1X PBS-Tween-20 (0.05％) for three times, 50 μL 1X ELISA/ELISPOT Diluent (eBioscience) was added each well to bind non-specific molecules at RT for 1hr.

The plate was then washed as mentioned for three times followed by adding supernatant with 50 μL 1X ELISA/ELISPOT Diluent at 4°C overnight. After the plate was washed for three times, 1X IL-1β or TNF-α detection antibodies (Biolegend) was added with 50 μL 1X ELISA/ELISPOT Diluent at RT for 1hr. After washed for three times, the plate

in the dark at RT for 30 mins. And then, the plate was washed for seven times, and 100 μL 1X TMB substrate (Invitrogen) was added for detection and the reaction was

stopped by adding 50 μL stop solution (2N H2SO4). The fluorescence was detected at 450 nm in an ELISA reader (Biorad).

2.7 BMDM transduction by Ak4 shRNA and scramble lentivirus

Bone marrow cells were transduced at day 3 by Ak4 shRNA and scramble lentivirus at MOI of 30 with serum free medium containing 8 μg/mL polybrene (Sigma) and 20％

L929 supernatant at the speed of 1100g for 90 mins. After centrifuge, 2X complete culture medium containing 20 ％ L929 supernatant, 20% FBS, 1X penicillin/streptomycin, 1X L-Glutamine, 1X NEAA, 1X sodium pyruvate and 5 μM 2-ME was added with 1X volume of transduction medium overnight. Finally, the cells were washed with dPBS and cultured in fresh 20％ L929 culture medium up to day 7.

2.8 Statistical analysis

Statistical analysis was performed by Graphpad Prism (GraphPad Soft ware, La Jolla, CA, USA). All mRNA data acquired from at least three independent experiments are mean ± SD and its significances were determined by nonparametric tests of Student’s t

tests. All protein data acquired from at least three independent experiments are mean ± SEM and its significances were determined by parametric tests of Student’s t tests.

III. Results

1. Bone marrow (BM) cells were able to polarize into M1 and M2 macrophages

in vitro

M1 and M2 macrophage have distinct roles in innate immunity. To perform macrophage polarization, bone marrow cells were cultured in 20% L929 cell supernatant to become bone marrow-derived macrophages (BMDM) for 7 days, and further stimulated in M1-induced or M2-induced conditions for 24 hrs. After 7 day culture, there were around 90％ of CD11b⁺F4/80⁺ BMDM (Figure 1A). After cytokine stimulation for 24 hours, CD80 expression was enhanced in M1 macrophages (MFI=39.2) than M2 macrophages (MFI=24.7). Conversely, CD206 expression was elevated in M2 macrophages (MFI=129.3) than M1 macrophages (MFI=86.6) (Figure 1B). Our data demonstrated that bone marrow cells can differentiate into BMDM and be further stimulated into M1 and M2 macrophages.

2. Nos2 and Tnfa were highly expressed in M1 macrophages while Arg1 and

Egr2 were highly expressed in M2 macrophages

For further confirmation of M1 and M2 subsets, Nos2, Tnfa, Arg1 and Egr2 were analyzed by real-time PCR. The expressions of Nos2 and Tnfa, M1 markers, were upregulated in M1 compared to M0 and M2 (Figures 2A-2B). On the other hand, the

expressions of Arg1 and Egr2, M2 markers, were upregulated in M2 compared to M0 and M1 (Figures 2C-2D). Our data demonstrated that BMDM can be skewed into M1 and M2 with distinct expression of signature genes.

3. ATP level was increased by short hairpin RNA (shRNA) knockdown of Ak4 in

M1 subset

Adenylate kinase 4, located in mitochondria matrix, maintains homeostasis of cellular nucleotide (Panayiotou et al., 2014). However, the role of Ak4 in macrophages remains unclear. Here, the ratio of relative Ak4 mRNA expression in M1 to M0 was 100 and the ratio of relative Ak4 mRNA expression in M1 to M2 was 47 (Fig. 3A). Moreover, Ak4 protein expression was higher in M1 than M0 (5.9 folds) and M2 (2 folds) (Fig. 3B). To identify the role of Ak4 in M1, BMDMs were transduced with Ak4 or scramble shRNA lentivirus, further stimulated by LPS and IFN-γ, and GFP⁺ cells were sorted out for in vitro analysis (Fig. 3C and Supplemental Fig. S1). The titers of Ak4 and scramble

shRNA lentivirus were 6.2 x 10^6 TU/mL and 8.3 x 10^6 TU/mL, respectively (Supplemental Fig. S2). KD efficiency by Ak4 shRNA in M1 macrophages was 98.2％

(Fig. 3D). ATP level was elevated significantly (p=0.048) by Ak4 shRNA knockdown (Fig. 3E). Moreover, ADP/ATP ratio was reduced but not significant (p=0.148) in Ak4

4. Ak4 shRNA knockdown resulted in less cell numbers but not BMDM

maturation

Bone marrow cells were transduced with Ak4 shRNA or scramble shRNA lentivirus in the early stage of development and further cultured in 20％ L929 complete medium.

After 7 day culture in 20％ L929 culture medium, Ak4 KD group had less cell numbers of BMDM than scramble groups (Fig. 4A). However, the percentages of scramble and Ak4 KD mature macrophages, represented as CD11b+F4/80+, were comparable (91.8％

and 95％, respectively) (Fig. 4B).

5. The expressions of Nos2, Hif1a and Il1b mRNAs were downregulated and

Cybb mRNA was upregulated in Ak4 KD M1

Proinflammatory genes such as Nos2, Hif1a, Il1b, Tnfα and Cybb are upregulated in M1 macrophages (Gensel et al., 2017). We showed that the expressions of Nos2, Hif1a and Il1b were declined in Ak4 shRNA KD M1 macrophages (Fig. 5A-5C). In addition, Tnfa

and Egr2 expressions were decreased in Ak4 siRNA KD M1 macrophages (Supplemental Fig. S3). However, Cybb expression was increased in Ak4 shRNA KD M1 (Fig. 5D).

6. Protein expressions of iNOS, HIF-1α, IL-1β and TNF-α were reduced in Ak4

shRNA KD M1 macrophages.

iNOS and HIF-1α were expressed in M1 rather than M0. However, NOX2 highly expressed in M0 and M1 (Fig. 6A). Ak4 expression was decreased (24％) in Ak4 shRNA KD M1 (Fig. 6B) Furthermore, NOX2 showed no difference between scramble M1 and Ak4 shRNA KD M1(Fig 6C). Surprisely, the protein expressions of iNOS and HIF-1α were reduced (p=0.017 for iNOS; p=0.010 for HIF-1α) in Ak4 shRNA KD M1 compared to scramble M1 (Fig. 6D-6E). Moreover, IL-1β and TNF-α were also decreased significantly (p=0.005 for IL-1β; p=0.014 for TNF-α) in Ak4 shRNA KD M1 by ELISA analysis (Fig. 6F-6G).

7. Ak4 siRNA KD M1 exhibited increased ATP level and decreased ADP/ATP

ratio

To further confirm, mature BMDM were transfected with Ak4 siRNA for 3 hrs, and further stimulated by LPS and IFN-γ (Fig. 7A-7B). With around 68％ of Ak4 siRNA KD efficiency, ATP level and ADP/ATP ratio were significantly increased and decreased in Ak4 shRNA KD M1, respectively (Fig. 7C-7E).

8. Expressions of Nos2, Cybb, Hif1a and Il1b mRNAs were reduced, but not

In Ak4 siRNA KD groups, the expressions of Nos2, Cybb, Hif1a and Il1b mRNAs were decreased slightly compared to scramble groups (Fig. 8A to 8D).

9. The expressions of MHC class II, costimulatory molecules CD80 and CD86,

and mannose receptor CD206 were similar between Ak4 siRNA KD and

scramble siRNA M1 macrophages

MHC II, CD80, CD86 and CD206 are upregulated in activated BMDM. In Ak4 siRNA KD M1 macrophages, expression of MHC II showed no difference between Ak4 KD M1 and scramble M1 (Fig. 9A). Moreover, CD80, CD86 were comparable between Ak4 and scramble siRNA KD M1 (Fig. 9B-9C). Furthermore, the expression of CD206 in Ak4 KD M1 was same as that in scramble M1 (Fig. 9D).