Acetoacetate reduces growth and ATP concentration in cancer cell lines which over-express uncoupling protein 2

Abstract

Background: Recent evidence suggests that several human cancers are capable of uncoupling of mitochondrial ATP generation in the presence of intact tricarboxylic acid (TCA) enzymes. The goal of the current study was to test the hypothesis that ketone bodies can inhibit cell growth in aggressive cancers and that expression of uncoupling protein 2 is a contributing factor. The proposed mechanism involves inhibition of glycolytic ATP production via a Randle-like cycle while increased uncoupling renders cancers unable to produce compensatory ATP from respiration.

Methods: Seven aggressive human cancer cell lines, and three control fibroblast lines were grown in vitro in either 10 mM glucose medium (GM), or in glucose plus 10 mM acetoacetate [G+AcA]. The cells were assayed for cell growth, ATP production and expression of UCP2.

Results: There was a high correlation of cell growth with ATP concentration (r = 0.948) in a continuum across all cell lines. Controls demonstrated normal cell growth and ATP with the lowest density of mitochondrial UCP2 staining while all cancer lines demonstrated proportionally inhibited growth and ATP, and over-expression of UCP2 (p < 0.05).

Conclusion: Seven human cancer cell lines grown in glucose plus acetoacetate medium showed tightly coupled reduction of growth and ATP concentration. The findings were not observed in control fibroblasts. The observed over-expression of UCP2 in cancer lines, but not in controls, provides a plausible molecular mechanism by which acetoacetate spares normal cells but suppresses growth in cancer lines. The results bear on the hypothesized potential for ketogenic diets as therapeutic strategies.

Figures

Figure 1

Specificity of UCP2 ELISA. (A) Standard curve of UCP2 antigen is calibrated against sandwich ELISA assay (see text). (B) UCP2 is calculated from standard curve in cell lysates from RFP3, SW48 and MDA MB 231 (B, C, and D respectively). Slopes of regressions represent ng UCP2 per μg total protein.

Figure 2

Lack of non-specific protein inhibition. There is no significant correlation between expression of cellular protein and the % change in ATP due to supplemental acetoacetate in the medium. r2 = 0.08 (p > 0.2) Data are from Table 1.

Figure 3

(A): The effect of adding acetoacetate to glucose medium on ATP concentration and cell growth. ATP and cell growth are both expressed as a percent of the respective values obtained at 96 h in 10 mM glucose (only) medium. M64 and M65 represent MCH 064 and MCH 065 cell lines, respectively. (B) The relation of cell growth to ATP concentration. Normal fibroblast controls RFP3, MCH 064 and 065 (in top right oval) demonstrate neither growth inhibition nor reduced ATP. All other data points represent seven cancer cell lines of Table 1 with parallel reduction in ATP and cell growth as also see in 3A. Regression equation plots growth (g) vs. ATP concentration (a) in blue in lower right of figure (p < 0.01). Data are from Table 1.

Figure 4

Cell growth vs. UCP2 expression in fibroblast controls vs. cancer lines. Purple series represents UCP2 in cells grown in glucose medium; blue series represents cell lines grown in glucose plus acetoacetate. Cancer lines over-express UCP2 vs. control fibroblasts when grown in either medium. Asterisks reflect significant differences (p < 0.05) in UCP2 values for cells grown in G or G+AcA growth medium when compared to corresponding results from all fibroblast lines. Data are from Table 1.

Figure 5

UCP2 expression in fibroblasts vs. MDA MB 231. Peroxidase staining of RFP3 controls (top) involves fewer cells and is less intense than of MDA MB 231 (bottom). Magnification 100×.

Figure 6

ATP concentration as a function of UCP2 expression. (ATP (y) in acetoacetate medium as a function of UCP2 expression (x) [regression equation upper right; p < 0.05]. Lowest expression of UCP2 with greatest ATP preservation is present in control fibroblasts (in oval). SW48 is identified as an outlier. (Data are from Table 1).