Impact of Peptide Sequence on Functional siRNA Delivery and Gene Knockdown with Cyclic Amphipathic Peptide Delivery Agents

Abstract

Short-interfering RNA (siRNA) oligonucleotide therapeutics that modify gene expression by accessing RNA-interference (RNAi) pathways have great promise for the treatment of a range of disorders; however, their application in clinical settings has been limited by significant challenges in cellular delivery. Herein, we report a structure-function study using a series of modified cyclic amphipathic cell-penetrating peptides (CAPs) to determine the impact of peptide sequence on (1) siRNA-binding efficiency, (2) cellular delivery and knockdown efficiency, and (3) the endocytic uptake mechanism. Nine cyclic peptides of the general sequence Ac-C[XZ]₄CG-NH₂ in which X residues are hydrophobic/aromatic (Phe, Tyr, Trp, or Leu) and Z residues are charged/hydrophilic (Arg, Lys, Ser, or Glu) are assessed along with one acyclic peptide, Ac-(WR)₄G-NH₂. Cyclization is enforced by intramolecular disulfide bond formation between the flanking Cys residues. Binding analyses indicate that strong cationic character and the presence of aromatic residues that are competent to participate in CH-π interactions lead to CAP sequences that most effectively interact with siRNA. CAP-siRNA binding increases in the following order as a function of CAP hydrophobic/aromatic content: His < Phe < Tyr < Trp. Both cationic charge and disulfide-constrained cyclization of CAPs improve uptake of siRNA in vitro. Net neutral CAPs and an acyclic peptide demonstrate less-efficient siRNA translocation compared to the cyclic, cationic CAPs tested. All CAPs tested facilitated efficient siRNA target gene knockdown of at least 50% (as effective as a lipofectamine control), with the best CAPs enabling >80% knockdown. Significantly, gene knockdown efficiency does not strongly correlate with CAP-siRNA internalization efficiency but moderately correlates with CAP-siRNA-binding affinity. Finally, utilization of small-molecule inhibitors and targeted knockdown of essential endocytic pathway proteins indicate that most CAP-siRNA nanoparticles facilitate siRNA delivery through clathrin- and caveolin-mediated endocytosis. These results provide insight into the design principles for CAPs to facilitate siRNA delivery and the mechanisms by which these peptides translocate siRNA into cells. These studies also demonstrate the nature of the relationships between peptide-siRNA binding, cellular delivery of siRNA cargo, and functional gene knockdown. Strong correlations between these properties are not always observed, which illustrates the complexity in the design of optimal next-generation materials for oligonucleotide delivery.

Keywords: cell-penetrating peptide; cyclic peptide; drug delivery; siRNA delivery.

Figure 1

(A) Graphical representation of condensation of disulfide-constrained CAPs with siRNA into CAP siRNA nanoparticles that facilitate translocation of siRNA across cell membranes into the cytosol. (B) Structural representation of the cyclic-Ac-C[WR]₄G-NH₂ CAP in which the indole side chain groups of Trp residues are shown in green, the Arg guanidium-presenting side chains are shown in blue, and the Cys sulfur residues forming the constraining disulfide bond are shown in yellow.

Figure 2

Cellular uptake of CAP–siRNA complexes into A549 lung adenocarcinoma cells. (A) Delivery of fluorescently labeled siRNA (siGlo Red-siRNA) with our CAPs in A549 lung cancer cells. Nuclei are stained with DAPI. Cell membranes are stained with Wheat Germ Agglutinin-Alexa 488. CAP concentration was 13.3 μM; siGlo Red-siRNA concentration was 132 nM. (B) Quantification of the number of CAP–siRNA complexes in A549 lung cancer cells. Amount of siRNA per nucleus represented as the average of at least two images (n = 2, **p ≤ 0.05, ****p ≤ 0.0001).

Figure 3

Knockdown efficiency of CAP–siRNA complexes against TTF-1 mRNA expression in A549 lung adenocarcinoma cells. Relative TTF-1 mRNA levels as percent knockdown normalized against GAPDH loading control determined by RT-qPCR. Scrambled is the control siRNA delivered with lipofectamine 3000. TTF-1 is delivery of TTF-1 siRNA with lipofectamine 3000 (n = 5, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001).

Figure 4

Quantification of siRNA/nucleus in the presence of DNS (400 μM) or MDC (200 μM) in A549 lung adenocarcinoma cells. (A) [FKFE]₂ (n = 2), (B) [FK]₄ (n = 2), (C) [FR]₄ (n = 3, **p ≤ 0.01, ***p ≤ 0.001), (D) [LR]₄ (n = 4, *p ≤ 0.05), and (E) [YR]₄ (n = 3, *p ≤ 0.05). (F) [WR]₄ (n = 4, **p ≤ 0.01, ***p ≤ 0.001), (G) [WK]₄ (n = 5, *p ≤ 0.05), (H) [WH]₄ (n = 5, ****p ≤ 0.0001), and (I) (WR)₄G (n = 5, ****p ≤ 0.0001).

Figure 5

Quantification of siRNA/nucleus in A549 lung adenocarcinoma cells previously treated with scrambled, Cav1, CLTC, or Cav1 + CLTC siRNA. (A) [FKFE]₂ (n = 4), (B) [FK]₄ (n = 3), (C) [FR]₄ (n = 4, **p ≤ 0.01, ***p ≤ 0.001), (D) [LR]₄ (n = 3, ***p ≤ 0.001, ****p ≤ 0.0001), (E) [YR]₄ (n = 3, *p ≤ 0.05), (F) [WR]₄ (n = 5, **p ≤ 0.01, ***p ≤ 0.001), (G) [WK]₄ (n = 7, *p ≤ 0.05, **p ≤ 0.01), (H) [WH]₄ (n = 3, **p ≤ 0.01), (I) [WS]₄ (n = 2, *p ≤ 0.05), and (J) (WR)₄G (n = 4, ***p ≤ 0.001, ****p ≤ 0.0001).