Disparate Response to Methotrexate in Stem Versus Non-Stem Cells

Abstract

Methotrexate (MTX) is a commonly used chemotherapeutic agent that kills cancer cells by binding dihydrofolate reductase (DHFR) as a competitive inhibitor. Due to its non-selectivity, MTX also impairs normal (non-cancerous) cell function and causes long-term damage to healthy tissue. These consequences have been investigated extensively in bone-derived cells due to their sensitivity to the drug. While DHFR likely plays a role in normal cell response to MTX, research in this area is limited. Moreover, how MTX sensitivity differs among cell types responsible for maintaining connective tissues is unknown. The goal of this study was to investigate the role of DHFR and subsequent nucleotide synthesis in normal cell response to MTX. We also sought to compare adverse effects of MTX among normal cell types to identify sensitive populations and resistant cell sources for regenerative procedures targeting patients undergoing chemotherapy. DHFR overexpression or exogenous amino acid + nucleoside delivery rescued normal cells from adverse MTX effects. Conversely, DHFR knockdown impaired MTX-treated adipose-derived stem cell (ASC) osteogenesis. Proliferation of ASCs and bone marrow stem cells was more resistant to MTX than that of terminally differentiated osteoblasts. However, stem cells became susceptible to the drug after beginning differentiation. These results suggest that the ability of stem cells to survive and to maintain their surrounding tissues likely depends on whether they are in a "stem" state when exposed to MTX. Therapeutic strategies that delay the differentiation of stem cells until clearance of the drug may produce more favorable outcomes in the long-term health of treated tissues.

Keywords: Cancer; Chemotherapy; Mesenchymal stem cell; Methotrexate; Regenerative medicine.

Figure 1

Exogenous amino acid + nucleoside (GAT) delivery regimen schematic. (A) GAT was delivered to MTX-treated NHF samples following four different regimens. (B) Either No GAT or B GAT schedules were used during the 48 h MTX treatment for ASCs, BMSCs, and OBs.

Figure 2

Effects of MTX on GFP and DHFR-GFP-transfected NHFs. (A) Protein expression for GFP and DHFR-GFP-transfected NHFs after treatment with 0.1 μM MTX. (B) Quantification of band intensity revealed that DHFR expression was significantly increased based on MTX as a factor (p < 0.005), but was not significant for individual comparisons. Quantified protein expression is presented as a percent of untreated, endogenous DHFR expression in GFP-transfected cells. (C) Cell counts of DHFR-GFP-transfected NHFs were significantly higher than GFP-transfected cells after MTX treatment (p < 0.001). While 0.1 μM MTX-treated, DHFR-GFP-transfected NHF counts were similar to untreated controls, 0.2 μM MTX-treated counts were significantly reduced (p < 0.005). Cell counts are presented as a fold difference of untreated, GFP-transfected control counts. All error bars depict standard deviation. Groups with different letters are significantly different (p < 0.05).

Figure 3

Temporal and concentration effects of nucleoside/GAT delivery for rescue of MTX-treated NHFs. (A) GAT delivery regimens rescued cell numbers to varying degrees, with B GAT providing complete recovery. (B) Successful GAT rescue depended on both the concentration of MTX and concentration of GAT. Too little GAT resulted in incomplete rescue of cell numbers, although some recovery could be observed for lower MTX concentrations. Cell counts are presented as a fold difference of untreated, No GAT counts. Error bars depict standard deviation. Groups with different letters are significantly different (p < 0.05).

Figure 4

Effect of DHFR knockdown on MTX-treated ASC proliferation. (A) Protein expression from control and DHFR siRNA-treated ASCs. (B) Quantification of band intensity showed an 87% reduction in DHFR protein levels after knockdown. (C) qPCR confirmed the effectiveness of DHFR siRNA transfection, also showing a 85% reduction in mRNA expression. (D) DHFR knockdown had no effect on ASC proliferation, regardless of MTX concentration. Protein and mRNA expression are presented as a fold difference of control siRNA samples. Cell counts are presented as a percent of untreated counts in control siRNA samples. Error bars depict standard deviation.

Figure 5

Effects of DHFR knockdown on MTX-treated, ASC differentiation. Quantification of large (A) lipid count and (B) diameter revealed no differences between siRNA types or among MTX concentrations for adipogenic differentiation. (C) For osteogenic differentiation, treatment with ≥ 0.1 μM MTX and knockdown of DHFR both reduced ALP activity of ASCs (p < 0.05 and 0.01, respectively). No individual comparisons showed statistically significant differences. Values are presented as a fold difference of untreated levels within non-targeting control siRNA conditions. Error bars depict standard deviation. Groups with different letters are significantly different (p < 0.05).

Figure 6

Proliferation of MTX-treated ASCs, BMSCs, and OBs. For ASCs, MTX had no effect on proliferation, but overall cell counts were significantly reduced based on GAT as a factor (p < 0.05). No effect was observed for individual comparisons. For BMSCs, neither MTX nor GAT affected cell counts. For OBs, proliferation was adversely affected by both MTX treatment and GAT delivery (p < 0.001), but these effects were not additive. Cell counts are presented as a fold difference of untreated/No GAT control counts. Error bars depict standard deviation. Groups with different letters are significantly different (p < 0.05).

Figure 7

Adipogenic potential of MTX-treated ASCs and BMSCs. (A, B) Imaging of oil red O staining revealed similar amounts of lipid production among groups for ASCs and BMSCs. (C) Quantification of large lipid counts showed both MTX and GAT had no effect on ASC adipogenesis. Conversely, MTX treatment significantly increased large lipid counts of BMSC samples for both No GAT and GAT conditions (p < 0.05) but was not significant for individual comparisons. (D) MTX treatment did not affect large lipid diameter in ASC or BMSC cultures. Large lipid count and diameter are presented as a fold difference of ASC or BMSC untreated/No GAT control values, respectively. Error bars depict standard deviation. (Scale bar indicates 200 μm).

Figure 8

Osteogenic capacity of MTX-treated ASCs, BMSCs, and OBs. (A) MTX treatment significantly reduced the ALP activity of ASCs and OBs (p < 0.01) in No GAT conditions but did not affect BMSCs. No significant differences were observed between untreated/No GAT and MTX-treated/GAT conditions for any cell type, indicating complete rescue. (B) MTX significantly reduced the calcified matrix production of OBs (p < 0.005) but had no significant effect on ASCs or BMSCs. As for ALP activity, no differences were observed between untreated/No GAT and MTX-treated/GAT conditions for any cell type. Values are presented as a fold difference of ASC, BMSC, or OB untreated/No GAT control levels, respectively. Error bars represent standard deviation. Groups with different letters are significantly different (p < 0.05).