Discussion
In this study, correlative measurement of the biophysical properties of the plasma membrane and the uptake of penetratin allowed us to reach the following major conclusions: (i) The physiological, positive membrane dipole potential inhibits the total cellular uptake of penetratin and its release from acidic endo-lysosomes. These conclusions are based on temporal measurement of the cellular intensity of AFDye532-labeled and NF-labeled penetratin, respectively. The effect of the dipole potential on both steps of penetratin uptake is most likely mediated by an alteration in the membrane insertion of the cell-penetrating peptide, since incorporation and penetration of peptides and small, hydrophobic molecules into the membrane are known to depend on the dipole potential (Asawakarn, Cladera & O’Shea, 2001; Cladera & O’Shea, 1998). (ii) Treatment with phloretin and atorvastatin reduced the membrane dipole potential and led to a significantly enhanced cytoplasmic concentration of penetratin. Since the applied, nanomolar concentration of atorvastatin is identical to that used in the clinical setting, the finding of statin-boosted uptake of penetratin is of potential medical significance. (iii) By analyzing the correlation between several membrane biophysical properties and penetratin uptake the dipole potential turned out to be the only characteristic of sufficient value for predicting penetratin uptake.
While both phloretin and atorvastatin decreased the dipole potential and increased penetratin uptake, their differential effects raise question about which step in the uptake process is the most sensitive to the dipole potential. In particular, although atorvastatin decreased the dipole potential more substantially than phloretin, only the latter induced an increase in the total cellular uptake of penetratin, while atorvastatin only increased the escape from acidic endo-lysosomes. These apparent contradictions can be resolved by the following three points: (i) The fact that phloretin does not increase the fraction of penetratin released from the endo-lysosomal compartment is explained by the fact that this compound is unlikely to reach sufficient concentrations in intracellular membranes during the brief, 10-minute incubation and therefore the dipole potential of these compartments remains largely unaffected. On the other hand, the three-day treatment with atorvastatin is sufficiently long so that substantial decrease in the cholesterol content and consequent reduction in the dipole potential of endo-lysosomal membranes can take place. (ii) The minuscule, phloretin-induced decrease of the dipole potential and the significantly elevated uptake of penetratin after phloretin treatment also require clarification. We assume that the physiological level of the dipole potential already limits penetratin uptake as much as potentially achievable by the dipole potential. This conclusion is supported by the fact that the significantly enhanced dipole potential after 6-ketocholestanol treatment had hardly any effect on the characteristics of penetratin uptake. On the other hand, even the relatively minor decrease in the dipole potential, achieved by phloretin, may be sufficient to facilitate penetratin uptake. However, the magnitude of this change in the dipole potential is not sufficiently large to reach statistical significance given the measurement errors. (iii) Although the reduction in the dipole potential achieved by atorvastatin is significantly larger than the phloretin-induced change, atorvastatin had no effect on the total cellular concentration of penetratin. The sudden rise in AFDye532-penetratin intensity and the protracted increase in the intensity of NF-penetratin indicate that initial uptake is mediated by endocytosis, a characteristic left unaltered by any of the treatments. Therefore, the lack of statin-induced increase in the total uptake of penetratin implies that endocytosis must be hindered in atorvastatin-treated samples. Indeed, it has been shown that statins inhibit endocytosis by interfering with the prenylation-dependent function of certain G proteins (Sidaway et al., 2004; Yilmaz et al., 2006). Alternatively, the compensatory increase in membrane viscosity after atorvastatin treatment may also impede endocytosis.
Despite the aforementioned uncertainties, two of the treatments, phloretin and atorvastatin, significantly enhanced the concentration of penetratin in the cytosol, the compartment most relevant from a therapeutical point of view. Analysis of the correlation between penetratin uptake and different biophysical properties of the membrane revealed that the dipole potential is the strongest predictor of penetratin uptake. Although the dipole potential, membrane compactness and viscosity characterize different membrane properties, all of them are related to membrane order. Therefore, it is not surprising that besides the dipole potential, other membrane characteristics also correlate with penetratin uptake. Membrane viscosity is correlated with endo-lysosomal release of penetratin. However, the predictive value of this correlation is undermined by the following two points: (i) While phloretin exerts no effect on membrane fluidity, it significantly enhances penetratin uptake in both SKBR-3 and MDA-MB-231 cells. (ii) Although penetratin uptake was modulated by the treatments in SKBR-3, none of them altered membrane viscosity in this cell line. Membrane compactness or hydration, characterized by the generalized polarization of Laurdan, was correlated weakly and in a statistically non-significant way with penetratin uptake. The predictive value of this correlation is further deteriorated by the fact that the generalized polarization of Laurdan was non-significantly modified by any of the treatments in MDA-MB-231 cells, while large and significant effects in penetratin uptake were observed.
In the present manuscript, not only did we identify the positive, intramembrane dipole potential inhibiting the uptake and endo-lysosomal release of penetratin, but we could also enhance the cytoplasmic concentration of the cell-penetrating peptide in a medically relevant way by statin treatment. In order for a treatment modality to be of potential medical importance, it must be well-tolerated and therapeutically effective. In contrast to many in vitro studies, in which statins are often overdosed, the nanomolar concentration range of atorvastatin used in the presented experiments is identical to the serum concentration at therapeutic doses (Bjorkhem-Bergman, Lindh & Bergman, 2011). By applying atorvastatin not requiring metabolic activation, we also circumvented another pitfall of in vitro cellular studies, the application of statins in a prodrug form, which are unlikely to be converted to the active metabolite in cell cultures. Although certain side-effects are associated with statin treatment, extensive experience indicates that they are safe, well-tolerated even in long-term applications in combinations with other drugs (Fievet & Staels, 2009; Luo, Wang, Zhu, Du, Wang & Ding, 2016). These circumstances further support the potential medical relevance of our findings.
Although reduction of the dipole potential in both MDA-MD-231 and SKBR-3 cells resulted in enhanced penetratin accumulation in the cytosol after phloretin treatment, the latter cell line exhibited lower sensitivity to atorvastatin. Sensitivity to statins correlates inversely with HMG-CoA reductase activity (Göbel, Breining, Rauner, Hofbauer & Rachner, 2019; Kimbung, Lettiero, Feldt, Bosch & Borgquist, 2016). While the expression of this enzyme is higher by ∼20-30% in SKBR-3 cells according to a publication (Kimbung, Lettiero, Feldt, Bosch & Borgquist, 2016) and the Expression Atlas of the European Bioinformatics Institute (https://www.ebi.ac.uk/gxa/home), the confidence intervals of the expression of HMG-CoA reductase in the two cell lines completely overlap according to the Genevestigator platform comparing transcriptomic data from several public repositories, with the level of expression corresponding to the MDA-MB-231 cell line spanning three orders of magnitude. In addition, atorvastatin, albeit at a micromolar concentration, resulted in 4-fold induction of HMG-CoA reductase expression in SKBR-3 cells, while no change in enzyme expression was observed in MDA-MD-231 cells (Kimbung, Lettiero, Feldt, Bosch & Borgquist, 2016). Cellular uptake of statins is believed to be mediated by transporters to a large extent. Organic anion transporter polypeptides (OATP) have been implicated in transmembrane import of statins (Dobson & Kell, 2008; Kalliokoski & Niemi, 2009; Wu, Whitfield & Stewart, 2000). However, none of the OATP transporters is expressed significantly differently according to the previous databases. Therefore, the most likely cause for the different atorvastatin sensitivity of the two cell lines in terms of penetratin uptake and reduction in cellular cholesterol content seems to be the difference in the baseline and statin-induced expression of HMG-CoA reductase expression, but solid conclusions cannot be drawn due to inconsistencies in the literature.
In conclusion, we have shown that a decreased, positive membrane dipole potential significantly increases both the total cellular uptake and endocytic escape of penetratin depending on what kind of treatment is used for modifying the dipole potential. As a result, both medically relevant (atorvastatin) and irrelevant (phloretin) treatments decreasing the dipole potential enhance the concentration of penetratin in the cytoplasm, the compartment most relevant for its therapeutic action. This discovery could permit the delivery of drugs and drug candidates exhibiting low or no transmembrane permeability into cells in animal experiments, human trials or in the clinical setting after further studies clarify the cell type dependence and the in vivo potential of this treatment.