Figure 5. Atorvastatin enhances endo-lysosomal escape of penetratin. Two different cell types were pretreated with the indicated concentrations of atorvastatin for 72 h and the uptake of AFDye-532-penetratin and NF-penetratin were followed by flow cytometry. Time-dependent intensities and their ratio were determined after gating out debris and dead cells. The error bars represent the standard error of the mean calculated from 13-15 samples from five biological replicates. Asterisks indicate significant difference of the sample treated with the highest concentration of atorvastatin in the figure compared to the control at 20 min (p<0.05).
Characterization of the effects of atorvastatin and the dipole potential modifying compounds on membrane biophysical properties
The previous experiments revealed that treatments reducing the membrane dipole potential enhance the total cellular uptake and the release of penetratin from endosomes, while increasing the dipole potential results in an opposite, weak effect. The validity of this conclusion was checked by calculating the rank-correlation between the dipole potential and penetratin uptake (Suppl. Fig. 4). A significant, negative correlation was revealed between the dipole potential and the total uptake (AFDye532 fluorescence), the concentration of penetratin in non-acidic compartments (NF fluorescence) and the fractional release of penetratin from the endo-lysosomal compartment (ratio of NF and AFDye532 intensities). Since the experimental data corresponding to the effect of phloretin on total cellular uptake of penetratin seemed to deviate from the trend established by the other conditions, the correlation was also tested by removing these outlier data points. The significant and biologically relevant negative correlation between the endo-lysosomal release of penetratin and the dipole potential was retained. However, removing the data points corresponding to phloretin treatment eliminated most of the dipole potential-dependent changes in total cellular uptake (Suppl. Fig. 4). This observation indicates that although both atorvastatin and phloretin decrease the dipole potential, their cellular effects and their influence on penetratin uptake are different. Since our measurement approach based on the pH-dependent fluorescence of NF assumes that the pH of the endo-lysosomal compartment is equally acidic in all compared samples, a possible explanation for the different results obtained with atorvastatin and phloretin is that they alter the pH of lysosomes. Therefore, we performed fluorescence ratiometric measurements of lysosomal pH, which revealed no significant effect of any of the treatments on lysosomal pH (Suppl. Fig. 5). These results imply that the ratio of the fluorescence of AFDye532-penetratin and NF-penetratin correctly reports the release of penetratin from acidic compartments. The issue of the different effect of phloretin and atorvastatin on penetratin release will be further considered in the Discussion.
Since treatments altering the dipole potential could potentially also modify other biophysical properties of the membrane, the correlations between penetratin uptake and two other fluorescent indicators were also tested. Membrane fluidity, inversely proportional to microviscosity and the fluorescence anisotropy of TMA-DPH (Kuhry, Fonteneau, Duportail, Maechling & Laustriat, 1983), was not changed by any of the treatments in the SKBR-3 cell line, and phloretin and 6-ketocholestanol were also without effect in MDA-MB-231 cells (Fig. 3B). In contrast, an unexpected decrease in the membrane fluidity of this cell line was induced by atorvastatin, an effect most likely attributable to the compensatory increase in the synthesis of certain lipid species in order to compensate for inhibited cholesterol synthesis (Dawaliby et al., 2016). Membrane fluidity and total cellular uptake of penetratin did not correlate with each other, but there was a strong correlation between membrane fluidity and the fractional escape of penetratin from acidic compartments (Suppl. Fig. 4). However, the power of membrane fluidity to predict endo-lysosomal escape of penetratin is limited by two observations: (i) the comparable effects of phloretin and atorvastatin on penetratin uptake are not correlated with their distinct effects on membrane fluidity in MDA-MB-231 cells; (ii) none of the treatments induced any significant change in the membrane fluidity of SKBR-3 cells, while phloretin increased endosomal release of penetratin.
The effect of the treatments on the generalized polarization of Laurdan, known to be proportional to membrane compactness and inversely related to membrane hydration (Parasassi, De Stasio, Ravagnan, Rusch & Gratton, 1991), was also tested. None of the conditions led to any significant change in membrane compactness in the two cell lines (Fig. 3C). The generalized polarization of Laurdan did not show appreciable correlation with total penetratin uptake or its concentration in non-acidic compartments (Suppl. Fig. 4). Given the fold-changes induced and the statistical significance of the correlations we can conclude that the dipole potential has the largest power for predicting penetratin uptake, and that release of the cell-penetrating peptide from endosomes and its concentration in the cytoplasm are both strongly correlated with the dipole potential.