MSCs were derived from discarded and de-identified bone marrow filters at the end of bone marrow harvest from normal healthy donors (Age range 18-45) following IRB protocol#2016-0298 at University of Wisconsin Madison. Mononuclear cells (MNCs) were isolated from the bone marrow through a Ficoll density centrifugation gradient. Isolated MNCs were plated onto tissue culture plates containing 1x α-Minimum Essential Medium (Corning, NY, USA) with human platelet lysate (Millcreek Life Sciences, MN, USA) and penicillin/streptomycin/amphotericin B (Hyclone, UT, USA). Non-adherent suspension cells were aspirated and adherent cells washed with phosphate-buffered saline. Cells were maintained by replacing fresh media every 48-72 hours. After the observation of colonies, cells were harvested and reseeded with the density of 3000-5000 cells per centimeter square. Cells were passaged consistently with the maximum confluence of 70% to 80% and cryopreserved until needed for the experiments. MSC identity was confirmed based on cell marker expression as defined in earlier studies (CD45-CD105+CD44+CD90+CD73+) (32, 52). 48-72 hours prior to the experiment, cryopreserved MSCs are thawed and culture rescued. Experiments were performed with the culture media 1x α-Minimum Essential Medium containing fetal bovine serum and penicillin/streptomycin/amphotericin B.

Atrazine (TCI America, OR, USA) was prepared with a two-step process. In the first step, Atrazine was dissolved in a high concentration (100mg/ml) in Dimethyl Sulfoxide (DMSO). Subsequently, DMSO containing Atrazine was further diluted in the 1x α-Minimum Essential Medium to obtain the concentration of 1000 µg/mL and were sonicated for 30-45 minutes to increase the solubility. After sonication, Atrazine stocks were stored in -20C until needed for experiments. A similar procedure was done for DMSO without Atrazine, which served as a vehicle control. Atrazine or DMSO were thawed as needed and used with the appropriate dilutions as indicated in the experiments. MSCs seeded onto 96 well plates were exposed with the appropriate concentrations of Atrazine or vehicle. After the incubation period, cells were harvested and supernatants collected for individual assays. For the experiments with IFNγ stimulations, MSCs were exposed to appropriate concentrations of Atrazine. After the exposure period, cells were harvested, counted and reseeded into 96 well plate with the normalized cell density. Subsequently, cells were stimulated with and without human recombinant IFNγ (Gibco Thermofisher, MA, USA) for 48 hours, and the supernatants were stored, and cells were subjected to flow cytometry.

MSCs exposed to Atrazine or vehicle were incubated with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (TCI America, OR, USA) (0.5mg/ml final concentration) for five hours. Supernatants were removed and formazan crystals were dissolved with organic solvent. Optical densities were measured at 560 nm, and background at 670nm was subtracted.

Atrazine or vehicle exposed MSCs were harvested and stained with BD Viaprobe (BD Biosciences, USA) and acquired in BD FACS Aria II flow cytometer to quantitate percentage forward scatter high and side scatter high populations. +/-IFNγ stimulated MSC populations were stained with antibodies HLA-ABC APC (Clone W6/32) (Biolegend, USA), HLA-DR PE (Clone G46-6) (BD Biosciences, USA) along with BD Viaprobe and acquired in BD FACS Aria II flow cytometer. Mean fluorescent intensity and histogram analysis for the marker expression was performed with Flow Jo software. +/- Atrazine exposed MSCs were subjected to BD Phosflow analysis. Cells were stimulated with appropriate concentrations of IFNγ for 15 minutes and were fixed with BD Cytofix buffer (BD Biosciences, USA) for 10 minutes. After fixation, cells were permeabilized with BD Phosflow Perm III Buffer (BD Biosciences, USA) for 30 minutes and then were subjected to staining with Alexa Fluor 647 Anti-Stat1 (pY701) antibody (BD Biosciences, USA). Stained cells were acquired in BD FACS Aria II flow cytometer and the results were analyzed in Flow Jo software. For intracellular IL-8 detection, Atrazine or vehicle exposed MSCs were incubated overnight with Golgiplug (BD Biosciences, USA). Intracellular flow cytometry staining was performed with BD Cytofix and Cytoperm procedure with IL-8 APC antibody (Clone E8N1) (BD Biosciences, USA) and acquired in BD FACS Aria II flow cytometer.

MSCs were seeded at a density of 10,000 cells per well in 96-well plates. Next day, cells were transfected with either non-targeting control siRNA or IL-8 SMART Pool siRNA (Horizon Discovery, CO, USA) with 1x α-Minimum Essential Medium containing 10 mM HEPES and Dharmafect transfection reagent 1. Transfection procedure was followed as defined in earlier studies and according to the manufacturer instructions (32) (Horizon Discovery, CO, USA). After five hours of transfection, cells were replaced with fresh 1x α-Minimum Essential Medium containing fetal bovine serum and penicillin/streptomycin/amphotericin B. After overnight resting, appropriate concentrations of Atrazine or vehicle was added and the cells were incubated for seven days. After seven days, supernatants were collected and stored. Cells were subjected to metabolic assays.

Supernatants derived from appropriate experimental conditions that were stored at -80°C were thawed and centrifuged at 500 x g for 5 minutes to remove debris. Subsequently they were subjected to magnetic bead-based multiplex according to the manufacturer instructions. Cytokine 30-Plex Human Panel (Thermofisher, USA) includes G-CSF(Granulocyte-Colony Stimulating Factor), GM-CSF (granulocyte-macrophage colony-stimulating factor), IFN-α, IFN-γ, IL-1β, IL-1RA, IL-2, IL-2R, IL-4, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12 (p40/p70), IL-13, IL-15, IL-17, TNF-α, Eotaxin, IP-10, MCP-1 (Monocyte chemoattractant protein-1), MIG (Monokine induced by gamma), MIP-1α (Macrophage inflammatory protein-1 alpha), MIP-1β (Macrophage Inflammatory Protein-1 beta), RANTES, EGF (Epidermal Growth Factor), FGF-basic (Fibroblast growth factor 2), HGF (Hepatocyte Growth Factor), VEGF (Vascular Endothelial Growth Factor). Angiogenesis 16-plex panel includes Angiopoietin-1, BMP-9, CD31 (PECAM-1), EGF, EMMPRIN, Follistatin, HB-EGF, HGF, IL-8 (CXCL8), Leptin, LYVE-1, PDGF-BB, Syndecan, TIE-2, VEGF-A, VEGF-D. Results were analyzed using Luminex xMAP software to obtain the concentrations as picogramsper milliliter.

Data were analyzed with the GraphPad Prism 9.0 software. Linear regression analysis was performed to obtain correlation coefficient (r) and P values. Negative Logarithmic P (NLP) values were generated from the p values for conditions with wide variations. Paired t test was performed to compare the conditions from the same donor MSCs. Area Under Curve (AUC) values were obtained from each curve of the experimental conditions and were subjected to paired t test analysis. Statistical significance was confirmed with a P-value of <0.05.

We have investigated dose dependent and longitudinal effect of Atrazine on human MSCs derived from the bone marrow of healthy human individuals (n=8 donors). MSC identity was confirmed based on appropriate marker expression as defined in our earlier studies (CD45−CD105+ CD44+ CD90+ CD73+) (32). MSCs derived from the passages between 2 and 5 were used on the experiments. We evaluated metabolic fitness of MSCs in the presence of various concentrations of Atrazine or vehicle (DMSO with respective dilutions) on days 1, 4, 7 and 10 utilizing MTT assay. In this assay, MSCs’ fitness in reducing the tetrazolium dye MTT to its insoluble formazan is evaluated, which is dependent on cellular energetics and NAD(P)H-dependent oxidoreductase enzymes and thus represents its metabolic activity (32). Our results demonstrated that Atrazine dose dependently inhibits the metabolic activity of MSCs derived from all the donors ( Figures 1A, B ). In addition, further statistical comparative analysis has identified that lower concentrations of Atrazine (15.625 and 31.25 μg/ml) do not reduce the metabolic activity of MSCs during a short-term (Day 1 and 4) exposure ( Figure 1C ). However, longer exposure (Day 7 and 10) with the same concentrations (15.625 and 31.25 μg/ml) have significant impact ( Figure 1C ). Comparative Area under Curve (AUC) analysis between vehicle control dilutions and Atrazine identified that vehicle controls have no effect ( Figure 1D ). Altogether this analysis has identified that Atrazine’s effect on MSCs is dependent on dose and duration of the exposure. We also tested the effect of in vitro culture passage (early vs late passage) on MSC’s metabolic sensitivity to Atrazine. Our results demonstrate that there is no significant difference between early and late passage MSC’s sensitivity to Atrazine ( Figure S1 ).

We investigated phenotypical changes induced by Atrazine on MSCs. MSCs derived from 5 independent donors were exposed with various concentrations of Atrazine or vehicle. Cells were harvested longitudinally at regular intervals and stained for viability dye 7AAD. Subsequently forward scatter and side scatter profiles were analyzed under flow cytometry which indicate their size and granularity. To exclude dead/apoptotic cells, we specifically gated on 7AAD negative populations which includes only live cells in the analysis ( Figure 1E ). Thus, this assay system evaluates the effect of Atrazine or vehicle exposure in modulating the size and granularity of live but not dead/apoptotic cells. Our results indicate that Atrazine dose dependently induces high scatter profile (Size and Granularity) on live MSCs ( Figures 1F, G ). We also identified that induction of high scatter profile on MSCs is also dependent on duration of the exposure ( Figures 1F, G ). These effects are absent in vehicle exposed conditions ( Figures 1F, G ). These results identify Atrazine induced non-toxic cellular phenotypic alterations of MSCs. In addition, these results further demonstrate that Atrazine induced high scatter properties on MSCs are dependent on dose and duration of the exposure.

We investigated the effect of Atrazine in modulating MSC’s innate fitness by analyzing their secretome panel consisting of immunomodulatory and regenerative secretory factors ( Figures 2A, B ). Supernatants collected from MSC cultures (derived from 8 independent donors) were analyzed for comprehensive 30-plex human analyte panel which includes cytokines, chemokines and growth factors. Our results demonstrate that MSCs innately secrete at least six soluble analytes (IL-6, MCP-1, VEGF, IL-8, HGF, IL-13) that are significantly detected in MSC culture supernatants over the control media ( Figures 2A, B ). Next, we investigated the effect of Atrazine or vehicle (7 days exposure) in modulating these six innate secretory molecules of MSCs. To determine Atrazine’s dose dependent effect, AUC values of individual secretory molecules were determined with Atrazine and vehicle conditions for each MSC donor. Three patterns of innate secretome are identified. (1) AUC values of VEGF and IL-6 are not significantly different between Atrazine and vehicle exposure which indicates that both of these secretory molecules are not affected by Atrazine’s exposure ( Figures 2C, F ). (2) Atrazine significantly (P<0.05) inhibits MCP-1 secretion by MSCs as identified by their reduction in AUC values compared to the vehicle ( Figures 2D, F ). (3) Secretion of IL-13, HGF and IL-8 by MSCs are significantly (P<0.05) upregulated by Atrazine as identified with their increased AUC values compared to the vehicle control ( Figures 2E, F ). Specifically, upregulation of IL-8 is robust and substantial with high statistical significance (P<0.0001) ( Figures 2E, F ). The other secretory molecules (24 molecules) that are not innately secreted by MSCs are not modulated by Atrazine exposure ( Figure S2 ). These results have demonstrated that Atrazine exposure not only downregulate but also upregulate and unmodulate the secretome of MSCs.

To further confirm Atrazine-induced IL-8 secretion in human MSCs, we performed intracellular cytokine capturing analysis using flow cytometry (n=4 independent donors, Atrazine (250μg/ml) exposure for 7 days). Our results demonstrate that Atrazine but not vehicle exposure induces IL-8 expression in human MSCs as identified by their intracellular accumulation captured with Golgi transport blockage ( Figures 3A, B ). Next to identify the relationship between Atrazine induced IL-8 secretion and MSC’s metabolic activity, we performed a linear regression analysis between MSC’s MTT activity and their six innate secretory molecules ( Figure 3C ). These analyses identified the strongest inverse correlation (r value closer to -1.0) between MTT activity and IL-8 secretion with Atrazine but not vehicle exposure ( Figures 3C, D ). Next, we investigated the relationship between Atrazine induced IL-8 secretion and other innate secretory molecules ( Figure 3E ). Linear regression analysis between the secretion levels of IL-8 and other secretory molecules (IL-6, IL-13, HGF, VEGF, MCP-1) identified that IL-8 exhibits the strongest direct correlation (r value closer to 1.0) with MCP-1 which is abolished upon Atrazine exposure ( Figures 3E, F ). These results suggest that Atrazine induced IL-8 display a unique relation with MSC’s metabolic activity and MCP-1 secretion.

MSCs’ responsive fitness to exogenous stimulation, such as IFNγ, is considered as the surrogate measure of their immune functionality (53, 54). Hence, we aimed to investigate the effect of Atrazine exposure on IFNγ induced upstream signaling activation and downstream key immune effector molecules on MSCs. IFNγ induces JAK-STAT signaling pathway on MSCs through the phosphorylation of STAT1 (pSTAT1 at Tyr701) (33). We deployed Phosflow™ technology to capture STAT1 phosphorylation on +/- Atrazine exposed MSCs. +/-Atrazine exposed MSCs (n=2 independent donors, +/- Atrazine exposure with various concentrations around 10 days) were stimulated with IFNγ for 15 minutes and STAT1 phosphorylation at Tyr701 was analyzed using Phosflow™ technology. Our results indicate that IFNγ phosphorylates STAT1 on MSCs irrespective of Atrazine exposure at similar levels (Area under curve (AUC) of pSTAT1 MFI with each concentration of atrazine from two independent donors (cumulatively) did not show a statistical significance) which suggests that Atrazine does not attenuate IFNγ induced upstream signaling on MSCs ( Figures 4A, B ). Next, we investigated the effect of IFNγ on downstream key immune effector molecules, MHC class I and MHC class II. MSC (n=5 independent donors) populations were exposed with various concentrations of Atrazine around 10 days and subsequently they were activated without and with IFNγ (20ng/ml) for 48 hours. Consistent with previous publications, our results also demonstrate that IFNγ upregulates both MHC Class I and MHC class II molecules on MSCs (55). Atrazine exposure does not have any effect on IFNγ dependent upregulation of MHC class I molecules on MSCs ( Figures 4C–E ). However, Atrazine attenuates IFNγ induced MHC class II molecule upregulation on MSCs ( Figures 4C–E ). Altogether these results suggest that Atrazine blocks IFNγ induced upregulation of MHC Class II but not MHC Class I and STAT1 phosphorylation on MSCs.

We have investigated the effect of IFNγ induced immunomodulatory secretome on Atrazine exposed MSCs. MSC (n=4 independent donors) populations were exposed with various concentrations of Atrazine around 10 days and subsequently they were activated without and with IFNγ (20ng/ml) for 48 hours. 30-plex secretome analysis of +/- IFNγ activated MSCs (n=4 independent donors) has identified three patterns of secretome which includes IFNγ induced upregulation, downregulation and unmodulation ( Figure 5A ). IFNγ upregulates MCP-1, MIG and IP10 while GMCSF and IL-8 are downregulated ( Figure 5B ). Of these MCP-1 and IL-8 are part of MSCs’ innate secretome whereas IP10 and MIG are absent in resting MSCs which are upregulated de novo upon IFNγ stimulation. This 30-plex analysis also identified IFNγ in the supernatants of IFNγ stimulated MSCs which is the inoculum used for the stimulation, and hence, we did not further include in the analysis ( Figure 5A ). The rest of the other secretory molecules are neither upregulated nor down regulated by IFNγ ( Figure 5A ). Our results also show that Atrazine exposed MSCs display attenuated upregulation of MCP-1, IP-10 and MIG upon IFNγ stimulation ( Figure 5C ). Conversely, IFNγ mediated downregulation of IL-8 secretion is dose dependently relieved by Atrazine exposure ( Figure 5C ). We also observed that IFNγ mediated downregulation of GMCSF is not significantly relieved upon Atrazine exposure. Linear regression analysis between Atrazine and IFNγ responsive secretome of MSCs has also identified that Atrazine exposure predominantly correlates with secretory molecules in an inverse manner (MCP-1 (r=-0.9 P<0.0001) exhibits the top inverse correlation) except for IL-8 ( Figure 5D ). These results suggest that Atrazine exposure counteracts IFNγ responsive secretome of MSCs. To identify the relative degree of Atrazine’s effect on individual IFNγ responsive molecules, we have determined the percentage normalized AUC values, since the secretion levels of each molecule differ due to donor variability. These analyses identify that normalized AUC values of Atrazine downregulated MCP1(+/-IFNγ), IP-10(+IFNγ), MIG(+IFNγ) and upregulated IL-8(+/-IFNγ) are not significantly different ( Figure 5E ). These suggest that Atrazine exposure equitably affects +/- IFNγ responsive matrix secretome of MSCs with features of upregulation and downregulation.

IL-8 is a key factor that plays a major role in the angiogenesis of blood vessels (56). MSCs have the propensity to drive angiogenesis as part of their repair and regenerative properties (57, 58). Since Atrazine upregulates IL-8 secretion on MSCs, we further investigated the effect of Atrazine on other angiogenic factors of MSCs. MSCs (derived from 4 independent donors) were cultured with various concentrations of Atrazine for 7 days. Supernatants collected from MSC cultures were analyzed for comprehensive 16-plex human analyte panel focusing on angiogenic factors using Luminex xMAP™ technology. Our results demonstrate that MSCs secrete at least 11 soluble analytes (VEGF-A, Follistatin, EMMPRIN, IL-8, TIE-2, LYVE-1, HGF, Syndecan, Angiopoietin-1, HB-EGF, VEGF-D) that are significantly detected in the supernatants over the control media though their individual secretion levels vary ( Figure 6A ). Next, we investigated the effect of Atrazine exposure on these 11 angiogenic factors. Correlation analysis of individual angiogenic factors and Atrazine concentrations have identified at least three independent relationships which includes direct, indirect and no correlations ( Figure 6B ). IL-8, HGF, and EMMPRIN exhibit significant direct correlations which indicates Atrazine upregulates their secretion ( Figures 6B, C ). In contrast, Follistatin exhibits an inverse correlation which indicates its downregulation upon Atrazine exposure ( Figures 6B, C ). At least seven angiogenic molecules do not exhibit any significant correlation with Atrazine exposure ( Figure 6B ). Altogether, these results signify the dichotomy of Atrazine’s effect in modulating the angiogenic factors of MSCs.

Since IL-8 secretion is substantially upregulated on MSCs upon Atrazine, we aimed to identify if Atrazine induced IL-8 secretion also plays any role on MSC’s metabolic activity and sensitivity to Atrazine as a loop mechanism. We performed siRNA knockdown of IL-8. Control or IL-8 silenced MSCs (4 independent donors) were exposed with differential concentrations of Atrazine, and MTT activity was measured eight days post exposure. Our results demonstrate that both IL-8 and control silenced MSCs are equally susceptible to Atrazine mediated dose dependent inhibition of metabolic activity. Further analysis has identified that both IL-8 and control silenced MSCs do not significantly differ in their AUC values ( Figures 7A, B ). These results suggest that Atrazine induced IL-8 has no effect on MSC’s metabolic sensitivity to Atrazine. Previous studies have shown that human MSCs sense xenobiotics such as 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) sense through Aryl Hydrocarbon Receptor (AHR) (59). However, our results show that knockdown of AHR on MSCs do not modulate MSC’s sensitivity to Atrazine ( Figure S3 ) which suggests that AHR does not play a role in the MSC and Atrazine axis.

Next, we aimed to identify if Atrazine induced IL-8 also modulates other angiogenic factors. We evaluated the supernatants of control and IL-8 siRNA silenced MSCs that were exposed (7 Days) with various concentrations of Atrazine for 11 angiogenic factors using Luminex xMAP technology ( Figure 7C ). Our results demonstrate that IL-8 silenced MSCs secrete lower levels of IL-8 (as identified with statistically significant lower AUC values) compared to control silenced MSCs ( Figures 7C, D ). In addition, IL-8 silenced MSCs also reduce HGF secretion which suggests that Atrazine induced IL-8 regulates HGF secretion on MSCs ( Figures 7C, D ). No significant modulation is observed in other angiogenic factors when comparing IL-8 and control silenced MSCs ( Figures 7C, D ). Altogether these results suggest that Atrazine induced IL-8 enhances specific angiogenic factors such as HGF but not others ( Figures 7C, D ).