Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive cancers, with a majority of patients presenting with metastatic disease at diagnosis. Extracellular vesicles (EVs) released by cancer cells have emerged as key mediators of intracellular signaling, communication, and immune modulation within the tumor microenvironment (TME).  Tumor-associated macrophages (TAMs) are crucial components of TME that influence tumor growth and metastasis. The present study is focused on determining the role of PDAC-derived EVs in altering macrophage phenotype, function, and metabolism, and their impact on tumor cell proliferation and invasion. EVs were isolated from the conditioned medium that was used to culture one established pancreatic cancer cell (PANC-1), one PDX cell line (PPCL68), and a normal (non-tumorigenic) established pancreatic epithelial cell line (hTERT-HPNE), using size exclusion chromatography. The EVs were extensively characterized using nanoparticle tracking (NTA), cryo-electron microscopy, and immunoblot analysis as per ISEV guidelines. Macrophages were generated from the bone marrow cells of C57/BL6 mice cultured in vitro using macrophage colony-stimulating factor (M-CSF). Macrophages were co-cultured with isolated PDAC or normal cell-derived EVs. Flow cytometry and multi-omic analyses were performed to determine the effect of EVs on macrophage phenotype, function, and behavior. Our results show that PDACcell-derived EVs were taken up by the macrophage in a time-dependent manner that resulted in an immunosuppressive phenotype characterized by higher expression of CD206 and PD-L1 and higher secretion of immunosuppressive cytokines, including TGF-beta, IL-10, and GM-CSF.  Functionally, PDAC-EV-treated macrophages were able to suppress the proliferation of CD8 T cells under in vitro and in vivo conditions. Metabolically, cancer EV-treated macrophages showed accumulation of immunosuppressive metabolites.

Targeted Metabolomics for cell samples

1.       Sample preparation for cells:

To cell samples in a 24-well plate, 300 µL of methanol/water (1:1) containing internal standards (200 ng/mL of debrisoquine for positive mode and 200 ng/mL of 4-nitrobenzoic acid for negative mode) was added, followed by gentle scraping for 20 sec. The above samples were transferred to 2mL Eppendorf. To the 24-well plate, 200 µL of PBS was added and mixed well, followed by transferring the PBS wash to the previously collected cell lysate.  The sample collected above was freeze-dried (also known as lyophilization). The lyophilized powdered samples were suspended in 25 μL of PBS and vortexed gently for 30 sec. Samples were plunged into dry ice for 30 s and heat shocked by plunging into a 37◦C water bath for 90 s. This was repeated a total of three times. Samples were then sonicated for 30 s. Next, 100 μL of methanol/water (1:1) was added to the samples. Tubes were vortexed for 30 s, incubated on ice for 20 min, followed by the addition of 100 μL of chilled ACN. The samples were incubated at −20◦C for 20 min. Finally, samples were centrifuged at 13,000 x g for 20 min at 4◦C. Supernatant was transferred to MS vials for LC-MS analysis.

2.       Sample preparation for media:

500 µL of media collected from each sample was freeze-dried (lyophilized) and suspended in 50 µL of PBS, followed by vortexing for 30 sec. To the above mixture, 100 μL of extraction buffer (methanol/water 50/50) containing 200 ng/mL of debrisoquine (DBQ) as an internal standard for positive mode and 200 ng/mL of 4-nitrobenzoic acid as an internal standard for negative mode, was added. The sample was vortexed for 30 seconds and incubated on ice for 20 min, followed by the addition of 100 μL of acetonitrile and incubation at -20 ℃ for 20 min. Samples were centrifuged at 13,000 rpm for 20 min at 4 ℃. The supernatant was transferred to an MS vial for LC-MS analysis.

3.       Sample preparation for NIST Plasma:

25 µL of NIST plasma sample was dissolved in 100 μL of extraction buffer (methanol/water 50/50) containing 200 ng/mL of debrisoquine (DBQ) as internal standard for positive mode and 200 ng/mL of 4-nitrobenzoic acid as internal standard for negative mode. The sample was vortexed for 30 seconds and incubated on ice for 20 min, followed by the addition of 100 μL of acetonitrile and incubation at -20 ℃ for 20 min. Samples were centrifuged at 13,000 rpm for 20 min at 4 ℃. The supernatant was transferred to an MS vial for LC-MS analysis.

4.       Data acquisition

Five microliters of the prepared sample was injected onto a Kinetex F5, 2.6 μm 100 Å 150 × 2.1 mm (Phenomenex, CA, USA) using SIL-30 AC auto sampler (Shimadzu) connected with a high flow LC-30AD solvent delivery unit (Shimadzu) and Exion 30AD communication bus module (Shimadzu) online with QTRAP 7500 (Sciex, MA, USA) operating in positive and negative ion mode. A binary solvent comprising water with 0.1% formic acid (solvent A) and acetonitrile with 0.1% formic acid (solvent B) was used. The extracted metabolites were resolved at a 0.2 mL/min flow rate. The LC gradient conditions were as follows: Initial – 100% A, 0% B for 2.1 minutes; 14 minutes – 5% A, 95% B till 15 minutes; 15.1 minutes – 100% A, 0% B till 20 minutes. The auto sampler and oven were kept at 15 °C and 30 °C, respectively. Source and gas settings for the method were as follows: curtain gas = 45, CAD gas = 10, ion spray voltage = 2000 V in positive mode and ion spray voltage = 4500 V in negative mode, temperature = 500 °C, ion source gas 1 = 4,5, and ion source gas 2 = 70. The data were normalized to internal standard area and processed using Sciex OS software.

5.       Data Processing

The data were normalized to the respective internal standard area and processed using Sciex OS. The quality and reproducibility of LC-MS data were ensured using several measures. The column was conditioned using the pooled QC samples initially and was also injected periodically to monitor shifts in signal intensities and retention time as measures of reproducibility and data quality of the LC-MS data. We also ran NIST plasma, periodically, prepared in the same manner to check the instrumental variance. We also have blank solvent runs between sets of samples to minimize carry-over effects.

6.       Data analysis

The abundance measurement for metabolites was expressed as intensity units that were initially normalized to internal standards and processed using MultiQuant 3.0.3 (Sciex). The data were pre-processed using a signal/noise ratio >20:1 and retention time (RT) tolerance of 5 seconds, after manual checking of metabolite peaks by experts to find the reliable features. We also use 20% of missing values in each feature as a filter-out criterion. Missing values were imputed by half of the minimum positive value in the original data. Thereafter, we used 20% of the coefficient of variation (CV) as our filter criterion to remove any possible noise before data normalization. Analytical drifts (if any) were corrected by quality control-based robust LOESS signal correction (QC-RLSC) (Dieterle et al., 2006; Li et al., 2016). All the analyses were performed in MetaboAnalyst Online (V5.0). The normalized LC-MS data were Log10 transformed and Pareto scaled in MetaboAnalyst. For binary comparisons, the data in each of the normalized csv files were subset by groups for comparisons of interest before input into the online tool.