Pharmacokinetics, biodistribution and toxicology of novel cell-penetrating peptides

Abstract

Cell-penetrating peptides (CPPs) have been used in basic and preclinical research in the past 30 years to facilitate drug delivery into target cells. However, translation toward the clinic has not been successful so far. Here, we studied the pharmacokinetic (PK) and biodistribution profiles of Shuttle cell-penetrating peptides (S-CPP) in rodents, combined or not with an immunoglobulin G (IgG) cargo. We compared two enantiomers of S-CPP that contain both a protein transduction domain and an endosomal escape domain, with previously shown capacity for cytoplasmic delivery. The plasma concentration versus time curve of both radiolabelled S-CPPs required a two-compartment PK analytical model, which showed a fast distribution phase (t_1/2α ranging from 1.25 to 3 min) followed by a slower elimination phase (t_1/2β ranging from 5 to 15 h) after intravenous injection. Cargo IgG combined to S-CPPs displayed longer elimination half-life, of up to 25 h. The fast decrease in plasma concentration of S-CPPs was associated with an accumulation in target organs assessed at 1 and 5 h post-injection, particularly in the liver. In addition, in situ cerebral perfusion (ISCP) of L-S-CPP yielded a brain uptake coefficient of 7.2 ± 1.1 µl g^-1 s^-1, consistent with penetration across the blood-brain barrier (BBB), without damaging its integrity in vivo. No sign of peripheral toxicity was detected either by examining hematologic and biochemical blood parameters, or by measuring cytokine levels in plasma. In conclusion, S-CPPs are promising non-toxic transport vectors for improved tissue distribution of drug cargos in vivo.

Figure 1

Pharmacokinetic profiles of enantiomers L- and D-S-CPP, according to a bicompartment model after a single intravenous injection in rats. Ten-month-old wild-type female Wistar rats were injected in the caudal vein at T₀ and blood was collected at different time points until 24 h. Linear graphical representation of plasma concentrations of ³H-L-S-CPP (dose = 6.3 × 10⁷ dpm  kg⁻¹, or 32.0 µg  kg⁻¹) (A) and ³H-D-S-CPP (dose = 16.2 × 10⁷ dpm  kg⁻¹, or 3.5 µg  kg⁻¹) (B). Plasma concentrations are presented as the % of the respective calculated initial concentration (C₀). Both sets of curves followed a bicompartment PK model with two phases: a rapid distribution phase (illustrated in green with intercept A as distribution coefficient) and a much longer elimination phase (illustrated in orange with intercept B as elimination coefficient) (C). Data are presented as the mean ± SEM. AUC Area under the curve, C₀ initial estimated concentration, Cl clearance, D₀ initial dose, dpm disintegration per minute, S-CPP Shuttle cell-penetrating peptides, T_1/2α/β half-life of distribution (α) and elimination (β), Vd estimated area of the compound’s distribution.

Figure 2

Pharmacokinetic profile of ³H-IgG combined to L-S-CPP, according to a linear model after a single intravenous injection in rats. Ten-month-old wild-type female Wistar rats were injected in the caudal vein at T₀ and blood was collected at different time points until 48 h. Linear graphical representation of plasma concentrations of ³H-IgG (at 1.3 × 10⁷ dpm  kg⁻¹ = 19.9 µg.kg^{- 1}, or with L-S-CPP (1.4 mg  kg⁻¹). Plasma concentrations are represented as the % of the respective calculated initial concentration (C₀). Both sets of curves followed a linear PK model. PK parameters are shown in the inserted Tables. Data are presented as mean ± SEM. Statistical analysis: Student’s t-test between ³H-IgG alone compared to the combination with L-S-CPP at equivalent doses for calculated PK parameters. AUC Area under the curve, C₀ initial estimated concentration, Cl clearance, D_i initial theorical dose, dpm disintegration per minute, Dose_extrapolated graphically estimated dose, S-CPP Shuttle cell-penetrating peptides, IgG immunoglobulin G, T_1/2 half-life, V_D estimated area of the compound’s distribution.

Figure 3

Apparent volume of distribution (μl  g⁻¹) of tritiated enantiomers L- and D-S-CPP, 1 and 5 h after intravenous injection. Ten-week-old male CD-1 mice were injected in the caudal vein and sacrificed by intracardiac perfusion 1 or 5 h post injection of ³H-L-S-CPP = 10.4 µg  kg⁻¹ (A) or ³H-D-S-CPP = 15.9 µg  kg⁻¹ (B) Tissues were homogenized and comparisons between the two forms of S-CPP are illustrated (C,D). The apparent volume of distribution (μl  g⁻¹) in each organ was calculated by dividing radioactive counts (dpm  g⁻¹) of each tissue by plasma counts (dpm  µl⁻¹) at the same time point. Data are represented on a logarithmic scale with a mean of N = 2–4 ± SEM. Statistical analyses were performed on values ​​after logarithmic transformation with an unpaired Student t-test (^£p < 0.05; ^££p < 0.01; ^£££p < 0.001). dpm disintegration per minute, S-CPP Shuttle cell-penetrating peptides, V_D estimated area of the compound’s distribution.

Figure 4

Apparent volume of distribution (μl g⁻¹) of tritiated IgG_antiNUP 1 h after co-injection with either forms of S-CPP or a scrambled peptide. Ten-week-old male CD-1 mice were injected in the caudal vein and sacrificed by intracardiac perfusion at 1 h post injection, thereby removing blood from the brain. IgG_antiNUP targets nuclear pore proteins. The ³H-IgG_antiNUP dose was 82.5 µg  kg⁻¹ ± L/D-S-CPP or combined with a control peptide ("Scramble" or SCR) at 3.6 mg  kg⁻¹. The apparent volume of distribution (μl g⁻¹) in each organ was calculated by dividing radioactive counts (dpm g⁻¹) of each tissue by plasma counts (dpm µl⁻¹). Data are represented on a logarithmic scale with the mean of N = 6–12 ± SEM. Statistical analyses were performed on values ​​after logarithmic transformation by a One-Way ANOVA parametric test followed by a Tukey post-hoc test (**p < 0.01; ***p < 0.001; ****p < 0.0001), a Welch ANOVA parametric test followed by a Dunnett’s post-hoc test ^(¤p < 0.05; ^¤¤¤p < 0.001). ns not significant, dpm disintegration per minute, S-CPP Shuttle cell-penetrating peptides, IgG immunoglobulin G, V_D estimated area of the compound’s distribution.

Figure 5

L-S-CPP crossed the blood–brain barrier, as assessed with in situ cerebral perfusion (ISCP). (A) The brain uptake coefficient (Clup; µl s⁻¹ g⁻¹) of ³H-L-S-CPP was calculated as V_brain/T, where T is the time perfusion (60 s). Injected concentrations were: 0.08, 0.16 and 0.32 µg/ml, corresponding to doses per mice of 0.2, 0.4 and 0.8 µg. The rate of entry of ³H-L-S-CP in the brain did not decrease with higher concentrations, consistent with the absence of a saturable transport mechanism. (B) The pmol/g of ³H-L-S-CPP in the brain were estimated from the dpm/g (L-S-CPP) and the specific activity and showed a linear increase (r² = 0.88; p < 0.0001), in accordance with free diffusion across the BBB. (C) Comparison of the proportion of ³H-L-S-CPP found in the vascular or extravascular fractions of the brain, showing a fast distribution into the brain parenchyma. Data are represented with the mean of N = 4–7 ± SEM. Statistical analyses were performed by a One-Way ANOVA parametric test followed by a Tukey post-hoc test (*p < 0.01) or a Welch ANOVA parametric test followed by a Dunnett’s post-hoc test (^¤p < 0.05; ^¤¤¤p < 0.001). Clup brain uptake coefficient, S-CPP Shuttle cell-penetrating peptides.