Guanidinium-rich cell-penetrating peptidomimetic–doxorubicin conjugates show linker-dependent, reduced acute toxicity to cardiomyocytes and normal breast epithelium while retaining activity against a HER2-positive breast cancer model (SKBR3 spheroids)

Figures

• 

  SCHEME 1. Panels A and B. (A) Synthesis of azide-functionalized doxorubicin (DOX-ACA-N3). (B) Copper-catalyzed azide–alkyne cycloaddition (CuAAC) of DOX-ACA-N3 to the alkyne/hexynoyl-terminated eight-guanidinium peptidomimetic to give DOX-ACA-HEX-G8. Panel C. (C) SPAAC click conjugation of DOX-ACA-N3 to the DBCO-terminated eight-guanidinium peptidomimetic to give DOX-ACA-DBCO-G8. The two conjugates differ in the linker and in the click chemistry (alkyne/CuAAC versus DBCO/SPAAC).

• 

  FIGURE 1. Dose–response of free DOX, DOX-ACA-HEX-G8, and DOX-ACA-DBCO-G8 in cancer spheroids (MDA-MB-231, top; SKBR3, bottom) at 24, 48, and 72 h. Points are mean ± SD of the replicate wells from three independent biological experiments (four to eight technical replicates per concentration in each); curves are four-parameter logistic fits. Dotted line, 50% viability. All panels were recomputed from the raw plate readings, each plate normalized to its own control wells (Section 2.4). MDA-MB-231 was tested to 20 μM at 48 and 72 h and to 50 μM at 24 h; SKBR3 was tested to 50 μM at all three time points (DOX to 20 μM at 48 and 72 h).

• 

  FIGURE 2. Dose–response in non-malignant spheroids (HB2, top; AC16 cardiomyocytes, bottom). Note the marked low-dose hormesis (apparent viability >100%) for both conjugates, most extreme for DOX-ACA-DBCO-G8 in AC16 at 48 h. Points are mean ± SD of the replicate wells from three independent biological experiments (four to eight technical replicates per concentration in each); curves are four-parameter logistic fits.

• 

  FIGURE 3. Cellular uptake by doxorubicin autofluorescence in SKBR3 at 24 h, compared at an equimolar, sub-lethal 2 µM. Panels: untreated control, free DOX, DOX-ACA-HEX-G8, and DOX-ACA-DBCO-G8. The red signal is the intrinsic fluorescence of doxorubicin and indicates intracellular drug accumulation; it is NOT a measure of viability or cell number. The untreated control is essentially signal-free, whereas all three DOX-bearing agents produce intracellular fluorescence, indicating that both conjugates are internalized. A single sub-lethal concentration is deliberately shown because fluorescence intensity scales with drug amount rather than cell number; a dose escalation would conflate drug loading with viability and must not be read as spheroid growth.

• 

  FIGURE 4. LDH release for the control panel on non-malignant cells (AC16 cardiomyocytes; single concentration 10 μM, 72 h; mean ± SD, n = 3 independent experiments), expressed as percent of the maximum (full-lysis) LDH signal (sample/maximum × 100, without subtraction of the spontaneous signal; the untreated control therefore reports the low background release, ∼3%). Free DOX, the DOX-ACA-N3 intermediate, and both non-covalent DOX + carrier mixtures give near-maximal LDH release (87%–99%), whereas the covalent conjugates give only 26% (DOX-ACA-HEX-G8) and 13% (DOX-ACA-DBCO-G8) and the bare carriers are inert (≤4%). On non-malignant cells, covalent conjugation, not mere co-administration, spares the cells from lytic damage. K3 (death/max) defines 100% release; K2 (untreated) and vehicle mark the assay floor (∼3%). Brackets show Games-Howell post hoc comparisons on the per-experiment means (***p < 0.001; **p < 0.01; *p < 0.05; n. s., p > 0.05); exact p values for every contrast are given in Supplementary Table S11. With n = 3 these p values are unstable and indicate only which contrasts the data support; the effect sizes carry the argument.

• 

  FIGURE 5. ATP versus DNA (biomass) for the control panel on non-malignant cells (AC16; 10 μM, 72 h; mean ± SD, n = 3), each normalized to untreated control. The cytotoxic arms (free DOX, DOX-ACA-N3, mixtures) fall to ∼1% (ATP) and ∼23–28% (DNA) and the carriers match the control. For the covalent conjugates, ATP drops below control (44% for DOX-ACA-HEX-G8, 82% for DOX-ACA-DBCO-G8) while DNA/biomass stays near control (∼90–94%); the ATP–biomass gap indicates reduced metabolic activity without proportional biomass loss at this single dose. K2 (untreated) is the 100% reference; K3 (death/max) marks the floor of both read-outs (ATP ∼0%, DNA ∼6%). Brackets show Games-Howell post hoc comparisons on the per-experiment means (***p < 0.001; **p < 0.01; *p < 0.05; n. s., p > 0.05); exact p values for every contrast are given in Supplementary Table S11. With n = 3 these p values are unstable and indicate only which contrasts the data support; the effect sizes carry the argument.

• 

  FIGURE 6. LDH release (% of maximum) on HER2-positive SKBR3 cells (10 μM, 72 h; mean ± SD, n = 3 independent experiments), expressed as sample/maximum × 100 without subtraction of the spontaneous signal. Free DOX, DOX-ACA-N3, and the non-covalent mixtures cause near-maximal release (∼100%), whereas the covalent conjugates release little LDH (∼17%) despite reducing biomass to ∼22% of control, indicating a non-lytic mode of action (apoptosis versus cytostasis pending caspase/annexin confirmation). K3 (death/max) defines 100% release; K2 (untreated) and vehicle mark the assay floor (∼3%). Brackets show Games-Howell post hoc comparisons on the per-experiment means (***p < 0.001; **p < 0.01; *p < 0.05; n. s., p > 0.05); exact p values for every contrast are given in Supplementary Table S11. With n = 3 these p values are unstable and indicate only which contrasts the data support; the effect sizes carry the argument.

• 

  FIGURE 7. ATP versus DNA (biomass) on HER2-positive SKBR3 cells (10 μM, 72 h; mean ± SD, n = 3 independent experiments), each normalized to untreated control. The cytotoxic arms (free DOX, DOX-ACA-N3, mixtures) fall to ∼1% (ATP) and ∼22% (DNA) and the carriers match the control. Both covalent conjugates reduce both read-outs below control (ATP 21% for DOX-ACA-HEX-G8 and 43% for DOX-ACA-DBCO-G8; DNA/biomass ∼22%), with no hormesis, confirming a genuine loss of biomass by an orthogonal measure rather than an ATP-only effect. K2 (untreated) is the 100% reference; K3 (death/max) marks the floor of both read-outs (ATP ∼1%, DNA ∼12%). Brackets show Games-Howell post hoc comparisons on the per-experiment means (***p < 0.001; **p < 0.01; *p < 0.05; n. s., p > 0.05); exact p values for every contrast are given in Supplementary Table S11.

• 

  FIGURE 8. Cell-free ATP spike-recovery control. Recovery of a known ATP spike measured in the absence of cells, in the medium and reagent used for the AC16 spheroid experiments, in a white 96-well flat-bottom plate, and read immediately, reproducing the interval over which ATP and test article are in contact in the cell-based assay. Each point is one determination; color denotes the ATP spike level (50, 100, 230, 500 or 1,000 nM) and the horizontal bar is the mean of the arm across all compound concentrations and spike levels. Recovery is the background-corrected signal of the arm divided by the background-corrected signal of the buffer arm at the same spike level; a compound-only blank was subtracted for every arm and concentration. Test articles were assayed at 0.01–0.50 µM and ethanol was held at 1% (v/v) in every well. The shaded band and the dotted lines mark the 85%–115% acceptance criterion. No determination fell outside the band (range 92.2%–108.9%, median 99.6%, n = 155 recovery values; the five buffer wells define 100% and are plotted for reference, 160 wells in total). Recovery was independent of the spike level and did not differ between arms containing the DAPEG carrier and…