br Str br VES br unimers
unimers in the low 0–20 μM range used (Fig. S1C,D and Table S2). For both lipid-modified cationic peptides, there was a significant increase in IC50 (p < .001) on MDR cancer cells compared to the parental cell line, suggesting that the activity of the H8R8-based amphiphiles was partially dependent on Pgp efflux activity. Pgp is able to efflux both hydrophobic and hydrophilic drugs, thereby reducing the effective intracellular drug concentration. For example, Pgp has been shown to efflux docetaxel and doxorubicin, requiring 100-fold more drug to kill MDR cancer cells than non-resistant cancer cells, whereas lipid-modified cationic pep-tides bypass the efflux capabilities [30,31]. Consistent with the litera-ture, VES alone exhibited some anti-cancer activity, with IC50 values of 23.0 ± 2.0 μM and 36.0 ± 4.6 μM on EMT6/P and EMT6/AR-1, re-spectively [32,33]. Covalent modification of VES to H8R8 led to a 5-fold decrease in IC50 on both EMT6/P and EMT6/AR-1 relative to VES alone.
We compared the anti-cancer activity of protease resistant D-amino acid-based lipid-modified cationic peptides to that of their faster de-grading L-amino 50-76-0 counterparts . Surprisingly, only a modest decrease in the IC50 was observed using the D-amino acid containing H8R8-based amphiphiles (Table S1). The enhanced activity may be at-tributed to the lower binding affinity to membrane-associated heparin sulfates and/or slower rates of internalization, as previously reported for similar cell penetrating peptides . Thus, we used L-amino acid-based amphiphiles going forward as the L-amino acid peptides are naturally present in the body and the reduced degradability of D-amino acid containing peptides may be toxic to non-cancer cells .
To confirm the anticancer selectivity of H8R8-based amphiphiles, the IC50 of each amphiphile against MDR cancer cells was compared to that of non-cancer cells. NIH/3T3 cells were selected as the non-cancer cell line as fibroblasts are major stromal cells and they would typically be in close proximity to cancer cells . Cancer selectivity is typically deduced from the difference in IC50 between cancer and endothelial or fibroblast cells . The IC50 of VES-H8R8 and Str-H8R8 on healthy cells was up to 8.4-fold and 4.3-fold higher, respectively, than that of breast cancer cells (p > .001, Fig. 2A), demonstrating greater selective toxi-city to cancer vs. healthy cells. Interestingly, the selectivity that we observed is consistent in order of magnitude with that observed of the clinical chemotherapeutic, docetaxel, where the IC50 on NIH/3 T3 cells is 6.4 fold higher than that on breast cancer EMT6/P cells (50 nM vs 7.8 nM) [38,39]. In both breast cancer cell lines investigated, VES-H8R8 was twice as selective compared to Str-H8R8, which may be due to the cancer selective activity reported for VES alone [37,40,41]. The anti-cancer selective activity of both VES-H8R8 and Str-H8R8 may arise from differences in cell membrane composition (e.g., o-glycosylated mucin concentration), mitochondrial polarization and greater cell prolifera-tion of cancerous vs healthy cells [42–44]. The difference in membrane potential is attributed to healthy cells exposing more zwitterionic phospholipids vs cancer cells exposing more anionic phospholipids and negatively charged glycoproteins [44,45]. The mitochondrial mem-brane potential is typically more negative in cancerous than healthy cells. For example, Neu4145 cancer cells exhibit a mitochondria membrane potential of −210 mV whereas healthy cells typically ex-hibit a mitochondrial membrane potential in the range of −108 to
Fig. 2. H8R8-based amphiphiles exhibit potent and selective anti-cancer activity. (A) Comparison of the IC50 of Str-H8R8 and VES-H8R8 on the parental breast cancer cell line, EMT6/P, the multi-drug resistant breast cancer cell line, EMT6/AR-1, and the healthy fibroblast cell line, NIH/3T3. (B) Relationship be-tween the anti-cancer activity on EMT6/AR-1 cells and the lipid length of the various H8R8-based am-phiphiles. Data are presented as a mean ± SD (n = 3) and statistical analyses was performed using one-way ANOVA and Tukey’s multiple comparison test (***p < .001).
−159 mV, which makes cancer cells more sensitive to mitochondria selective treatment [9,15]. As drug sensitive and MDR cancer cells exhibit more polarized mitochondria relative to non-cancerous cells, we anticipate that VES-H8R8 and Str-H8R8 will be potent against other drug sensitive and MDR cancer cells.
To gain greater insight into the role of the hydrophobic tail to anti-cancer activity, we measured the IC50 of a series of H8R8-based am-phiphiles as a function of the lipid length of the amphiphilic peptides on EMT6/AR-1 (Fig. 2B) and EMT6/P cells (Fig. S3B). Both unmodified H8R8 and butyl-modified H8R8 showed limited anti-cancer activity, with IC50 above 200 μM. However, octyl-modified H8R8 exhibited an IC50 of 25.6 ± 1.9 μM, and longer hydrophobic segments (C12 and longer) exhibited IC50 in the low μM range (< 7 μM). These results suggest that H8R8-based amphiphiles modified with a lipid greater than or equal to 12 carbons is the threshold for anti-cancer activity. The enhanced anti-cancer activity of longer hydrophobic tails is likely due to enhanced membrane association and cell penetration, as was ob-served for C10-C16 modified heptaarginine (R7) . Interestingly, mitochondrial-penetrating peptides with a logP in the range of −1.0 to −1.4 exhibited superior mitochondrial targeting relative to less hy-drophobic peptides, highlighting a potential relationship between mi-tochondrial colocalization and anti-cancer activity . We chose to further investigate Str-H8R8 and VES-H8R8 for cytotoxicity and tar-geting as the stearyl and VES moieties share similar theoretical logP values. While VES is known to be bioactive and Str-H8R8 has been used in gene delivery and nanoparticle formulations, there may be important and untapped synergistic effects encapsulating VES-H8R8 or Str-H8R8 [47,48]. Moreover, stearyl-modified cationic peptides have been used for nanoparticle-based drug delivery systems in vivo, where we an-ticipate that safe doses may be used with VES-H8R8 or Str-H8R8, prompting further in vivo investigations [49,50].