Synthesis and Investigation of Peptide-Drug Conjugates Comprising Camptothecin and a Human Protein-Derived Cell-Penetrating Peptide

Abstract

Drug targeting strategies, such as peptide-drug conjugates (PDCs), have arisen to combat the issue of off-target toxicity that is commonly associated with chemotherapeutic small molecule drugs. Here we investigated the ability of PDCs comprising a human protein-derived cell-penetrating peptide-platelet factor 4-derived internalization peptide (PDIP)-as a targeting strategy to improve the selectivity of camptothecin (CPT), a topoisomerase I inhibitor that suffers from off-target toxicity. The intranuclear target of CPT allowed exploration of PDC design features required for optimal potency. A suite of PDCs with various structural characteristics, including alternative conjugation strategies (such as azide-alkyne cycloaddition and disulfide conjugation) and linker types (non-cleavable or cleavable), were synthesized and investigated for their anticancer activity. Membrane permeability and cytotoxicity studies revealed that intact PDIP-CPT PDCs can cross membranes, and that PDCs with disulfide- and protease-cleavable linkers liberated free CPT and killed melanoma cells with nanomolar potency. However, selectivity of the PDIP carrier peptide for melanoma compared to noncancerous epidermal cells was not maintained for the PDCs. This study emphasizes the distinct role of the peptide, linker, and drug for optimal PDC activity and highlights the need to carefully match components when assembling PDCs as targeted therapies.

Keywords: camptothecin; cell‐penetrating peptide; cleavable linker; melanoma; peptide–drug conjugate.

FIGURE 1

The proposed characteristics varied to generate a suite of PDIP‐CPT PDCs, including the peptide conjugation handle and linker type (colored pink). For cleavable linkers, the cleavage site is indicated with a dotted black line. For the peptide sequences, disulfide bonds are indicated with a grey line and the residues in the spacer region are colored teal; c[ ] = backbone cyclized; X = azidoalanine; *amidated C‐terminus.

SCHEME 1

Modification of CPT to produce constructs with (A) a non‐cleavable linker, (B) a disulfide‐cleavable linker, and (C) a cathepsin B protease‐sensitive linker.

FIGURE 2

(A) General scheme and conditions for PDC synthesis. (B) Structures and yields of the four PDCs. The yields were determined using the molecular weight of the TFA salt, assuming all basic residues, the terminal amine (if present) and the quinoline moiety within CPT are protonated.

FIGURE 3

(A) and (B) Cytotoxicity of PDCs, CPT, and peptides against cultured HT144 (melanoma) and HaCaT (noncancerous) cells, following incubation for 72 h. Cell death was measured using resazurin, with 0.1% (v/v) Triton X‐100 as a control for 100% cell death. Data points are expressed as mean ± SD for at least two biological replicates. Dose response curves were fitted using [inhibitor] versus response with variable slope, Graphpad Prism v10. (C) Hemolysis of human RBCs (0.25% v/v) following incubation with PDCs, CPT, and peptides for 72 h. RBC lysis was determined by measuring released hemoglobin, with 0.1% (v/v) Triton X‐100 as a control for 100% cell lysis. Melittin was included as a membranolytic control. Data points are expressed as mean ± SD for two technical replicates. Dose response curves were fitted using [inhibitor] versus response with variable slope, Graphpad Prism v10. (D) Half maximal cytotoxicity concentration (CC₅₀) and minimal hemolytic concentration (HC₁₀) values of PDCs, CPT, and peptides. CC₅₀ values are expressed as mean (SEM). The legend applies to all three graphs. c[A]PDIP is a desulfurized variant of cPDIP 2 (see Table S1) to prevent the presence of free thiol during assays. n.d., not determined.

FIGURE 4

(A) Membrane permeability and cellular uptake of PDCs, scaffold peptides, and CPT. Compounds were added to PAMPA wells (4 μM) and incubated for 4 h. Supernatant from apical and basolateral sides of the membrane were collected and analyzed using LC–MS. The amount in the combined soluble fractions was compared to the starting amount to determine % recovery, and the apparent permeability (Papp) was calculated as before (Sevin et al. 2013). Data is expressed as mean ± SD from a single experiment with three technical replicates. (B) Internalization/cell membrane association of PDCs 12–14 and PDIP with HT144 cells. Cells were treated with PDIP or PDCs (4 μM) for 1 h, washed to remove extracellular compound, then extracted with 75% (v/v) MeCN in mQH₂O (containing 1.75% (v/v) TFA), to recover soluble peptide or PDC from precipitated cell proteins. The relative amount of internalized peptide or PDC was determined by integrating area under the curve (AUC) for the [M + 6H]⁶⁺ m/z peak (for the intact mass) using time of flight–mass spectrometry (see Figure S2). The amount of internalized compound was compared to the AUC present when 4 μM of each analogue was added to untreated cells after addition of the extraction solution, to determine the percentage of internalized compound. Data is expressed as mean ± SD from a single experiment with three technical replicates for each treatment and control and was analyzed with ANOVA (Tukey multiple comparisons; **p < 0.01, *p < 0.05). (C) Detection of intracellular CPT following 6 h treatment of HT144 cells with CPT and PDCs 12 and 14. Cells were treated with CPT, 12 or 14 (4 μM) for 6 h, washed to remove extracellular compound, and the relative amount of CPT was compared by determining AUC from targeted multiple reaction monitoring using tandem mass spectrometry (see Figure S3). Data is expressed as mean ± SD from a single experiment with three technical replicates for each treatment and was analyzed with ANOVA (Fisher multiple paired comparisons; **p < 0.01, *p < 0.05).