Mouse model

Animal experiments were approved by the National Animal Research Authorities, Norwegian Food and Safety Authority (FOTS: 17895 and FOTS: 26021) and carried out according to the European Convention for the Protection of Vertebrates used for Scientific Purposes. Eight- to twelve-week-old female BALB/cJ mice were purchased from Janvier Labs. France, and housed (3–6 mice/cage) under controlled light/temperature/humidity conditions with ad libitum access to food and water. Tumour growth and mouse weight were monitored and followed the pattern we reported previously [17]. The sample size for each experiment was decided based on our previous experience with these tumours [17, 18].

Cell culture and tumour induction

The 67NR and 66cl4 cell lines were obtained from the Karmanos Cancer Institute, Detroit, MI, USA. The cells were cultured in Dulbecco’s modified Eagle’s medium (Lonza, BioWhittaker, Cat #BE12-604F) supplemented with 10% fetal calf serum (FCS, Thermo Fisher Scientific, Gibco, Cat #10272-106), 2 mM L-glutamine (Lonza, Cat #BEBP17-605E), and 50 U/mL penicillin–streptomycin (Thermo Fischer Scientific, Gibco, Cat #15070063). Cells were incubated at 37°C in 5% CO2 and allowed to grow to 75% confluence, after which they were used to prepare a cell suspension of 20 × 106 cells/mL in sterile PBS (Sigma-Aldrich, Cat #D8537). For orthotopic tumour induction, mice were selected randomly and 1 × 106 viable cells were injected into the inguinal (fourth) mammary fat pad of mice under anesthesia (1.5% isoflurane with 67.5% N2 and 32.5% O2). Tumour growth and mouse weight were monitored weekly, and the experiments were performed 21–26 days after injection. Control mice were healthy mice without injections. The total number mice used: controls N = 21; 67NR N = 25; and 66cl4 N = 25; none of the mice showed unexpected or adverse effects.

Tissue processing

Blood was collected from mice under anaesthesia (4% isoflurane with 67.5% N2 and 32.5% O2) by cardiac puncture in heparin-containing tubes. Then, it was centrifuged at 1500 × g for 10 min, and plasma was collected and stored at −80°C until analysis (ELISA and LC–MS/MS). If the collected plasma showed haemolysis, it was excluded from the analysis. Leukocytes were obtained after red cell lysis using RBC lysis buffer (Thermofisher Scientific, 1X RBC Lysis Buffer, Cat #00-4333-57).

After blood collection, mice were euthanised by cervical dislocation and organs were harvested. Bone marrow cells were obtained by flushing the marrow cavities of the femurs with PBS using a syringe with a 26-gauge needle, followed by red cell lysis using RBC lysis buffer. Splenic cells were obtained by gently dissociating small spleen pieces through a 70 μm sterile cell strainer. Tumour and lung single-cell suspensions were prepared by mincing 0.5 mm pieces in an uncoated Petri dish containing 2 mL of serum-free RPMI using a scalpel. The minced tissue was placed in a tube containing a mix of Liberase DL (0.835 U/mL, Roche, Cat #5466202001), Liberase TL (0.835 U/mL, Roche, Cat #5401020001) and DNase I (13 U/mL, Qiagen, Cat #79254) in DMEM for 30 min at 37°C on a shaking incubator. The tissue single-cell suspensions were washed with serum-free RPMI, and one red-cell lysis step was performed. Samples were centrifuged at 400 × g for 8 min, and single cells were suspended in cold FACS buffer (PBS, supplemented with 2% fetal calf serum and 2 mM EDTA). Tissues showing > 50% necrosis were excluded.

Flow cytometry

The single cell suspensions obtained by tissue processing explained above were resuspended in FACS buffer and incubated with 0.5 μg of Fc block antibody (anti-mouse CD16/32, Affymetrix, eBioscience, Cat #14-0161-85) for 15 min on ice. Surface staining of cell suspensions was performed by addition of anti-mouse CD45-FITC (Biolegend, Cat #103108, 0.5 μg/test), CD11b-BV421 (Biolegend, Cat #101251, 1 μg/test) and Ly6G-PE/Cy7 (Biolegend, Cat #127618, 0.5 μg/test) antibodies and incubation for 20 min on ice protected from light. After incubation, the cell suspensions were washed twice, and samples were stained with Zombie Aqua (BioLegend, Cat #423101) for 10 min on ice, according to the manufacturer’s instructions. After surface staining, cells were washed twice with cold FACS buffer and fixed using the Fixation Buffer (R&D Systems, Cat #FC009) for 10 min at RT according to the manufacturer’s instructions. Following fixation, the cells were washed twice and resuspended in 200 μL of Permeabilization Buffer (R&D Systems, Cat #FC009) to perform the intracellular staining. The cells were incubated with 0.5 μg of Fc block antibody (anti-mouse CD16/32) for 15 min, followed by 30 min with ARG1-APC antibody (Invitrogen, Cat #17-3697-82, 1 μg/test) at RT. Finally, the cells were washed twice, resuspended in FACS buffer and analysed on a BD LSR II flow cytometer.

Myeloid cell subpopulations were identified by surface staining with CD45-FITC, CD11b-BV421, Ly6G-PE/Cy7, Ly6C-PerCP/Cy5.5 (Biolegend, Cat #128012, 0.5 μg/test), F4/80-PE-Texas Red (ThermoScientific, Cat #MF48017, 1 μg/test) and ARG1-APC antibodies.

Single staining and fluorescence-minus-one (FMO) staining were performed. Antibody volumes were used according to manufacturer recommendations. The generated data was analysed using FlowJo software v10. Gating strategies for the analyses are described in supplementary data.

Cell sorting

The single-cell suspensions obtained by tissue processing explained above were resuspended in FACS buffer and incubated with 0.5 μg of Fc block antibody (anti-mouse CD16/32) for 15 min on ice. Then, the cells were incubated with CD45-FITC antibody for 30 min on ice, followed by staining with the viability dye Zombie Aqua, as mentioned above. Cells were washed, resuspended in FACS buffer and sorted based on CD45 expression levels using a BD FACSAria™ Fusion flow cytometer.

RNA sequencing

RNA sequencing was performed on the RNA extracted from the sorted populations. RNA quantity, quality, integrity, and purity were evaluated using Qubit, bioanalyser and NanoDrop. Sequencing libraries were generated using the SENSE mRNA-Seq library prep kit V2, according to the manufacturer’s instructions (Lexogen GmbH), with 350 ng of total RNA as input.

Libraries were normalised to 2.6 pM, subjected to clustering and single-read sequencing. According to the manufacturer's instructions, sequencing was performed for 86 cycles on a NextSeq500 HO flow cell (Illumina). Base calling was performed using NextSeq500 instrument software (RTA 2.4.6). FASTQ files were generated using the bcl2fastq2 Conversion Software v2.20.0.422. The sequences are available from February 01, 2024 under the GEO NCBI Accession code: GSE211223.

RNA sequencing data analysis

FASTQ files were quality controlled with fastqc (v0.11.9), filtered and trimmed using fastp (v0.20.0). Trimmed sequences were aligned to the reference genome using STAR (v2.7.3), and quality metrics were extracted with Picard CollectRNASeqMetrics (v2.21.5). Transcript counts were generated using quasi-alignment (Salmon v1.3.0) to GRCh38 transcriptome reference sequences. Transcript counts were imported into R statistical software and aggregated to gene counts using the tximport (v1.14.0) Bioconductor package for downstream statistical analysis. Gene counts were normalised and analysed for differential expression using the DESeq2 Bioconductor package. Differential expression was defined as genes with a false discovery adjusted p-value < 0.05 and log fold change difference > 1. Pathway enrichment analysis was performed using the enriched Bioconductor (v3.0) package.

Quantitative real-time PCR

Total RNA from the whole tumour was obtained by first placing small tumour pieces in lysis buffer (RLT, provided in RNeasy® Mini Quiagen Kit, Cat #74106) and then disrupted with 2.8 mm ceramic beads using an MP Biomedical FastPrep-24™ tissue homogeniser (SKU 116004500), with two rounds of 4.0 m/s for 20 s each. Samples were vortexed, and RNA was isolated following manufacturer instructions (RNeasy® Mini Kit, Qiagen, Cat #74106). Total RNA from sorted CD45+ and CD45- cells was obtained using the Quick-RNA™ MicroPrep Kit (Zymo Research, Cat #R1051), following manufacturer instructions. RNA was quantified by spectrophotometry (ND1000 Spectrophotometer, NanoDrop, Thermo Scientific) and stored at −80˚C until further analysis. cDNA was synthesised from 500 ng (whole tumour) and 100 ng (sorted cells) total RNA using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems™, Cat #4387406). Quantitative real-time PCR (RT-qPCR) was performed in 20 µl reactions containing 10 µl of 2X QuantiTect SYBR Green PCR master mix (Qiagen), 2 µl 10X QuantiTect Primer Assay and 8 µl of the sample containing 2 ng of cDNA. QuantiTect Primers for Mm_Hprt_1_SG (Cat #QT00166768) -as a reference gene- and Mm_Arg_1_SG (Cat #QT00134288) were used. RT-qPCR was performed on the StepOne Plus system (Applied Biosystems) using the following cycling conditions: 95°C for 10 min, followed by 40 cycles at 95°C for 15 s, 60°C for 30 s and 72°C for 30 s. Relative gene expression levels were calculated using the 2(-ΔΔCT) method.

Immunoblotting

Primary tumour pieces were homogenised in lysis buffer containing 8 M urea, 4% CHAPS (Cat #C9426), 1 M DTT (Cat #646563), Complete® protease inhibitor (Cat #1187350001) and phosphatase inhibitor cocktail II and III (Cat #P5726 and #P0044). Homogenised tissues were placed on a shaker for 20 min at 4°C and centrifuged at 16,000 × g for 20 min at 4°C. Protein concentration was measured using BioRad protein assay (Cat #5000006). Equal amounts of proteins were separated using NuPAGE® Novex® Bis–Tris gels (4–12%) (Invitrogen), transferred onto nitrocellulose membranes and stained with Revert™ 700 Total Protein Stain (P/N: 926–11016). After destaining and washing, the membrane was further stained with ARG1 (GeneTex, Cat #GTX109242, 1:2000) and ERK1/2 (Cell Signaling, Cat #9107S, 1:2000) antibodies. Near-infrared fluorescent secondary antibodies were used and imaged on an Odyssey Near Infrared Scanner (Li-Cor Biosciences). Images were processed using Li-Cor Odyssey software Image Studio v5.2.5.

Fluorescence in situ hybridisation (FISH)

Formalin-fixed paraffin-embedded 66cl4 and 67NR tumours were cut into 4 µm sections using a microtome, mounted on Superfrost Plus slides, dried and used for FISH.

The RNAScope Pretreatment Kit (ACD Bio, Cat #310020) was used according to the manufacturer’s instructions. Following pretreatment, the sections were stained using the RNAScope 2.5 Detection Kit- Brown (ACD Bio, Cat #322310) and washed with RNAscope Wash Buffer (ACD Bio, Cat #310091) following the manufacturer’s instructions. The probe Mm-ARG1 (ACD Bio, Cat #403431) was used. Images were captured using an IX71 inverted fluorescence microscope (Olympus).

Enzyme-linked immunosorbent assay (ELISA)

Levels of ARG 1 protein in plasma samples, obtained from tumour-bearing and healthy mice, were determined using a Mouse Arginase 1 ELISA kit (Abcam, Cat #ab269541). Only samples with no visual evidence of haemolysis were analyzed. Plasma samples were diluted at 1:40, analyzed according to the manufacturer’s protocol and absorbance at 450 nm was read on the Bio-Rad iMark Microplate Reader Bio-Rad (Cat #168-1135).

Immunohistochemistry (IHC)

Formalin-fixed paraffin-embedded 66cl4 and 67NR tumours were cut into 4 µm sections using a microtome, mounted on Superfrost Plus slides and dried. Antigen retrieval was performed using citrate buffer (pH 6; Cat #C9999). Slide preparation and staining were performed using a Rabbit-specific HRP/DAB (ABC) Detection IHC Kit (Cat #ab64261) and VECTASTAIN Elite ABC-HRP Kit, Peroxidase (Rat IgG) (PK-6104), respectively, following the manufacturer’s instructions for all steps except treatment after primary staining. After primary staining, the slides were treated with hydrogen peroxide, and the manufacturer’s protocol was resumed. The slides were incubated with an anti-ARG1 antibody (Cell Signaling, Cat #93668, 1:100) for 1 h at room temperature.

Images were captured using an IX71 inverted fluorescence microscope (Olympus).

LC–MS/MS

Metabolites were extracted, measured and analysed using a previously described method [19].

Statistics

GraphPad Prism 9.2.0 (332) was used for statistical analysis. For the analysis of 2 groups, we used Mann–Whitney t-tests (Unpaired) and Kruskal–Wallis test (Unpaired) for more than two groups, followed by a multiple comparisons test.

Data availability

The data are scheduled to be publicly available from February 01, 2024, in the Gene Expression Omnibus (GEO) at GSE211223. The data are available earlier upon request.