Manganese systemic distribution is modulated in vivo during tumor progression and affects tumor cell migration and invasion in vitro

Metastatic disease remains the leading cause of death in cancer and understanding the mechanisms involved in tumor progression continues to be challenging. This work investigates the role of manganese in tumor progression in an in vivo model of tumor growth. Our data revealed that manganese accumulates within primary tumors and secondary organs as manganese-rich niches. Consequences of such phenomenon were investigated, and we verified that short-term changes in manganese alter cell surface molecules syndecan-1 and β1-integrin, enhance collective cell migration and invasive behavior. Long-term increased levels of manganese do not affect cell growth and viability but enhance cell migration. We also observed that manganese is secreted from tumor cells in extracellular vesicles, rather than in soluble form. Finally, we describe exogenous glycosaminoglycans that counteract manganese effects on tumor cell behavior. In conclusion, our analyses describe manganese as a central element in tumor progression by accumulating in Mn-rich niches in vivo, as well as in vitro, affecting migration and extracellular vesicle secretion in vitro. Manganese accumulation in specific regions of the organism may not be a common ground for all cancers, nevertheless, it represents a new aspect of tumor progression that deserves special attention.


Wound healing assays for tumor cell lines HeLa, B16 and MDA-MB-231
The different cell lines HeLa, B16F1 and MDA-MB-231 were cultured in 6-well plates until 80% confluence, then culture medium was changed to control (regular culture medium) or pre-Mn (regular culture medium with the addition of manganese chloride (MnCl2) 5 µM.Cells were incubated in these conditions for 1 h.Next, medium (control or Mn-added) was removed, cells were rinsed and mechanically removed from the plate in a cross-shaped pattern using a sterile P1000 tip.Cells were rinsed again for removal of cell debris and incubated in control culture medium for 12 h.In order to evaluate cell migration and wound closure, cells were imaged, using the cross center as a reference at 0 h and 12 h.Migration was quantified using ImageJ software and expressed as the percentage of migrated distance relative to the original wound width.

X-ray microfluorescence analyses of tumor cell lines HeLa, B16 and MDA-MB-231
X-ray microfluorescence analyses were performed at the Brazilian Synchrotron Light Source Laboratory (LNLS).Analyzed samples comprised of cultured cells placed on ultralene thin film (SPEX SamplePrep, NJ, USA).Cultured cell samples were quantified, scraped from the plate, placed on ultralene film and let air dry.Samples were analyzed at D09B X-ray Fluorescence beamline at room temperature and ambient pressure.

Divalent cations cell toxicity assays
LCC cells were seeded in 24-well plates (6x10 4 cells/well) and after cultures reached 80% confluence, cells were incubated for 24 h in increasing concentrations of MgCl2 or ZnCl2 (5, 10, 25, 50, 100 and 500 μM).Cells were harvested by enzymatic treatment and viable cells were quantified in a Neubauer chamber using trypan blue.

Divalent cations wound healing assays
LLC cells were cultured in 6-well plates until 80% confluence, then culture medium was changed to control (regular culture medium) or pre-Mg or pre-Zn (regular culture medium with the addition of magnesium chloride (MgCl2) or zinc chloride (ZnCl2) at 10 µM and 25 µM.Cells were incubated in these conditions for 1 h.Next, culture medium (control or cation-added) was removed, cells were rinsed and mechanically removed from the plate in a cross-shaped pattern using a sterile P1000 tip.
Cells were rinsed again for removal of cell debris and incubated in control culture medium for 12 h.In order to evaluate cell migration and wound closure, cells were imaged, using the cross center as a reference at 0 h and 12 h.Migration was quantified using ImageJ software and expressed as the percentage of migrated distance relative to the original wound width.

Cytometry analyses
For flow cytometry analysis of syndecan-1, LLC cells were prepared according to the cell migration wound healing assay.After 8 h of migration, cells were removed from the plates using PBS/EDTA 1mM solution and fixed with 4% paraformaldehyde.Then, cells were stained with rat anti-mouse syndecan-1 (1:200 -RD systems, MN, USA) followed by anti-mouse IgG Cy5 (1:500 -Life Technologies, CA, USA).Syndecan-1 expression was evaluated in 10,000 events/cells using a FACS LSRFortessa (BD Bioscience, NJ, USA) and analyzed in the Summit software (Cytomation, CO, USA).
Negative controls were as follows: extracted mRNA as template for the RT-PCR (to detect possible genomic DNA contamination) and cDNA-free RT-PCR reactions (to detect possible contaminated reagents).

Heparin-treated animal model of spontaneous metastasis
C57BL/6 mice were inoculated with LLC cells as described for the animal model of spontaneous metastasis, however, at weeks 2 and 4 animals were treated with 100 µL single injections of unfractionated bovine heparin (UFH) 0.1 ng/mL at the tumor site.
After 5 weeks of tumor development, tissues were collected for further analyses.All procedures involving animal experimentation were approved by the Federal University of Rio de Janeiro Animal Experimentation Committee (protocol number: 015/18) and were performed in accordance with the Brazilian guidelines for scientific use of animals.

Multi-elemental inductively coupled plasma (ICP)-optical emission spectroscopy
Modified fetal bovine sera (Mn-low and Mn-high FBS) were submitted to multielemental ICP-OES analyses for the investigation of actual elemental concentrations post modification protocol.Analyses were performed using an Optima DV 4300 (Perkin Elmer, Waltham, MA, USA) ICP-OES.The following parameters were used for the analyses: peristaltic pump flow 1.5 mL/min, plasma flow 15 L/min, nebulization flow 0.6 L/min, 1400 W, radial plasma view for Ca and K, axial plasma view for Cu, Fe, Mn, P and Zn.Multi-elemental calibration curve ranged from 0.001 to 20.0 mg/L.

Fig. S2 .
Fig. S2.Manganese retention and its effects on cell migration of other tumor cell lines.Different tumor cell lines were evaluated for migration pattern using wound healing assays by (a-c) intermittent monitoring.The tested cell lines were (a) HeLa, (b) B16F1 and (c) MDA-MB-231.Mn content was evaluated by briefly treating cells with MnCl2 5 µM for 1 h and collecting cells for XRF analyses after 3 h of incubation in standard medium.The analyzed cell lines were (d) HeLa, (e) B16F1 and (f) MDA-MB-231.Wound healing assays HeLa control N=4, HeLa pre-Mn N=6; B16F1 control N=9, pre-Mn N=6; MDA-MB-231 control and pre-Mn N=3.*p<0.05,**p<0.01,***p<0.0001;Student's unpaired t-test.

Fig. S3 .
Fig. S3.Magnesium and zinc do not affect tumor cell growth, survival and migration at low concentrations.LLC cell growth and survival were evaluated after 24 h of incubation with different (a) MgCl2 and (c) ZnCl2 concentrations.Cell number represents adhered live cells only.LLC cell migration was evaluated in wound healing assays by pre-incubating cells in (b) MgCl2 or (c) ZnCl2 at 10 µM and 25 µM for 1 h, followed by rinsing and incubation in control medium.Cells were imaged at 0 h and 12 h of migration.Cell growth assay N=3, wound healing assay N=9.*p<0.05,one-way ANOVA and Bonferroni's multiple comparison post-test.

Fig. S4 .
Fig. S4.Detection of β1-integrin and syndecan-1 in LLC cells.(a) PCR analyses were performed in LLC cells cultured in standard conditions to evaluate the expression of ITGB1 and SDC1.Molecular-weight size marker 1 kb.Next, LLC cells were submitted to the same conditions as a wound healing assay and collected at 8 h of migration for (b) flow cytometry analyses.(c) Mean fluorescence intensity.UFH (bovine unfractionated heparin -0.1 ng/mL); pre-Mn (MnCl2 5 µM 1 h pretreatment prior to migration).PCR N=2; cytometry N=3.Statistical analyses: one-way ANOVA test and Bonferroni's multiple comparison post-test.

Fig. S6 .
Fig. S6.Syndecan-1 is highly expressed in small cell clusters in primary tumors and tumor-bearing mice livers.Representative images of syndecan-1 immunostainings from (a, b and c) primary tumors, (d, e and f) tumor-bearing mice livers and (h) control liver, all at week 5. (g) Negative control.(i) Mngreenand Fereddistribution map of a tumor-bearing mouse liver.(j) Syndecan-1 immunostaining of sequential liver section shown in (i).Scale bars are 25 µm (a, b, c) and 50 µm (d to h).Primary tumor N=4; Tumor-bearing mice livers N=6; Control mice livers N=4.

Fig. S7 .
Fig. S7.Manganese content in primary tumors treated with unfractionated heparin.Tumor-bearing mice received unfractionated heparin (UFH) injections at weeks 2 and 4 of tumor progression and primary tumors were collected at week 5 of tumor progression for XRF analyses.Control animals N=17, UFH-treated N=6.Student's unpaired t-test non-significant.