Intracellular redistribution of cell-penetrating peptide p28: A mechanism for enhanced anti-cancer activity

Abstract

Cell-penetrating peptides (CPPs) have been studied as they provide an efficient strategy for the intracellular delivery of bioactive molecules in various biomedical applications such as cancer diagnosis and therapy. We have developed an anionic CPP, p28, that can preferentially enter cancer cells and induce cell cycle arrest and apoptotic cell death preclinically and that showed preliminary efficacy in humans. Yet, the underlying intracellular fate after cell entry remains largely uncharacterized. To better understand the intracellular trafficking of p28, we investigated more closely the role of endosomal acidification and retrograde transport in cancer cells. Here, we show that agents such as chloroquine (CQ), NH₄Cl, and nordihydroguaiaretic acid (NDGA) that alter discrete steps of endocytosis or Golgi-mediated transport remarkably increased p28 access to the cytosol and nucleus. Moreover, CQ significantly improved the antiproliferative effects induced by p28. Our findings suggest that inadequate intracellular localization is the major limiting factor for the efficacy of CPPs such as p28. Taken together, these results indicate that the effective and adequate alternation of intracellular localization of CPPs can be a promising drug-delivery strategy to specifically deliver the therapeutic molecule to its intracellular therapeutic active site or organelle that leads to potentiate the efficacy for cancer diagnosis and therapy.

1. Introduction

In the recent advances in drug development, bioactive macromolecules such as proteins, peptides, particles, DNA, and RNA have shown their immense potential as therapeutic agents [1,2]. However, their application is limited due to insufficient cellular, tissue uptake and target specificity [3]. Cell-penetrating peptides (CPPs) have been shown to have a promising potential to transport such macromolecules cargos directly into cells [4,5]. It has been reported that several CPPs can even preferentially deliver macromolecules to pathological tissues (e.g., tumors) whilst sparing healthy tissue [6–11]. The characteristics of preferential penetration can allow a reduction in the off-target effects. However, the main barrier to efficient intracellular delivery is not only the translocation of macromolecules across the cellular membrane, also their adequate intracellular localization. As they are often trapped inside endocytic organelles, it causes limiting access to their site(s) of action in pathological tissues (e.g., cytoplasm, mitochondria, or nucleus) [12]. Thus, intracellular targeting-delivery affords selective localization of CPPs to their target site, improving their efficacy and safety.

The tumor-targeted delivery of imaging agents or drugs is one of the major challenges in cancer therapy [13]. We previously reported that a small bacterial protein azurin from Pseudomonas aeruginosa and its derived peptide p28 preferentially enter various cancer cells independent of p53 expression status [7]. Upon entry, they induce p53-mediated anti-proliferative effects [7,8,14,15]. As a single therapeutic agent, p28 (NSC745104) was tested in two phase I clinical trials and granted the FDA Orphan Drug and Rare Pediatric Disease designations due to its demonstrated preliminary efficacy without apparent adverse effects, toxicity or immunogenicity in patients with advanced solid tumors and in pediatric patients with recurrent and refractory central nervous system (CNS) tumors (NIH/NCI and Pediatric Brain Tumor Consortium) [16–18]. Therefore, p28 is a potentially ideal CPP that can serve as a tumor-targeted vehicle [7,17,18] since it preferentially penetrates cancer cells [7], is highly water soluble [19], and clinically exhibits no significant adverse effects, toxicity or immunogenicity in humans [17,18]. In this study, we aimed to investigate more closely the intracellular trafficking of p28 in cancer cells. Here, we report that the inadequate intracellular localization of p28 is the major limiting factor for its efficacy. These findings hold promise for an improved efficacy of macromolecules like p28 and can guide the development of a promising and effective drug-delivery strategy.

2. Methods and materials

2.1. Peptide synthesis and conjugation

CPPs, p28 (an azurin fragment: Leu⁵⁰-Asp⁷⁷ LSTAA DMQGV VTDGM ASGLD KDYLK PDD) and octaarginine (Arg8), were synthesized by CS Bio at >95 % purity and mass balance. Peptides labeling with Alexa Fluor 568 and ICG was performed as described before [7,20]. Briefly, peptides were dissolved in PBS mixed with NHS-functionalized Alexa Fluor 568 dye (#A20003, Invitrogen, CA, USA) at a 1:2 protein/dye ratio, sodium bicarbonate was added, and the mixture were incubated overnight at 4 °C with continuous stirring. Labeled peptides were dialyzed against cold-PBS using Slide-A-Lyzer Dialysis Cassettes 2000 MWCO (#66205, ThermoFisher Scientific, CA, USA). Similarly, NHS-functionalized gold nanoparticles (GNPs) and gold nanorods (GNRs) (#CS11–15, Nanopartz Inc., CO, USA) were conjugated with p28 according to the manufacturer’s instructions. GNPs (15 nm diameter) or non-spherical GNRs with lengths of 73 nm × diameters of 25 nm were conjugated with 500 M excess of p28 in 0.1 M borate buffer (pH 8.0) at room temp for 4 h with continuous stirring. The labeled peptides were washed with 0.1 % Tween 20 in 1 % PBS at 9000 r.c.f. for 10 min and resuspended in PBS. For indocyanine green (ICG) labeled p28, maleimide-ICG was added to p28 (a cysteine residue was attached at the N-terminal of p28) in PBS incubated overnight at 4 °C with continuous stirring at dark. The labeled peptide was dialyzed against cold-PBS using Slide-A-Lyzer Dialysis Cassettes 2000 MWCO (#66205, ThermoFisher Scientific). Final products were analyzed by Mass Spectrometry Core (University of Illinois at Chicago, IL, USA).

2.2. Cell cultures

The human colon cancer cells, p53^wt HCT116, normal colon cell lines (CCD33Co), and fibroblasts were obtained from American Type Culture Collection (ATCC, Manassas, VA, USA). p53^null HCT116 cell line was supplied by Dr. B. Vogelstein [21]. Human melanoma cells (UISO-Mel29) were established and characterized in our laboratory [22]. All cells except UISO-Mel29 (DMEM) were cultured in MEME (Invitrogen) supplemented with 10 % heat-inactivated fetal bovine serum (R&D Systems, MN, USA), 100 units/ml penicillin, and 100 μg/ml streptomycin at 37 °C in 5 % CO₂.

2.3. Confocal imaging

HCT116, UISO-Mel29, CCD33Co, and fibroblasts were seeded overnight on glass coverslips at 37 °C under 5 % CO₂, rinsed with fresh medium, and incubated at 37 °C for 2 h in prewarmed medium containing Alexa Fluor 568–labeled p28 (AF-p28, 20 μM), Arg8 (AF-Arg8, 5 μM), or medium alone. For intracellular trafficking of p28, cells were pretreated with nordihydroguaiaretic acid (NDGA, 25 μM) chloroquine (CQ, 10 μM [23]) and ammonium chloride (NH₄Cl, 2 mM [24]) for 24 h, followed by staining with AF-28 for 2 h. For intracellular distribution of p28, cells were stained with AF-28 for 2 h. Cells were rinsed with PBS, fixed in 2.5 % formalin for 5 min and then stained with anti-Golgi-AF488 conjugate (sc-58770, Santa Cruz, Texas, USA) and anti-caveolae 1 antibody (#ab2910, Abcam) both diluted at 1:500 and incubated overnight. This was followed by anti-rabbit secondary antibody (#ab150077, Abcam). For LysoTracker^™ Green (#L7526, Invitrogen) and ER-Tracker^™ Green (#E34251, Invitrogen) staining, the manufacturer’s instructions were followed. Coverslips were mounted in medium containing 1.5 μg/ml 4′,6-diamidino-2-phenylindole to counterstain nuclei (#H-1200–10, DAPI, Vector Laboratories, CA). The coverslip slides were visualized using a confocal laser-scanning microscope (Model LC710; Carl Zeiss Inc., Baden-Württemberg, Germany).

2.4. Flow cytometry

Cells at 1.0 × 10⁶ in PBS were incubated for 2 h at 37 °C with AF-p28 (20 μM), AF-Arg8 (5 μM) in the presence or absence of 5 mM methyl-β-cyclodextrin (#332615, MβCD, Sigma-Aldrich) [8], 4.5 μM filipin (#SAE0087, Sigma-Aldrich), and 10 μM nocodazol (#R17934, Med-ChemExpress, New Jersey, USA) [7]. After incubation, cells were washed with PBS, fixed in 2.5 % formalin, resuspended in PBS, and subsequently analyzed with a MoFlo Cell Sorter (Dako, CA, USA). The fold increase of the mean fluorescence intensity (MFI) over background levels represents the mean fluorescence of three separate experiments.

2.5. Cell proliferation assay

The MTT assay was performed as we previously described [7]. Briefly, cells were seeded at a density of 8,000 cells per well in 100 μl of growth medium in 96-well plates and grown overnight. Then, the cells were exposed to p28 in the presence or absence of CQ at 10 μM for 24 h.

2.6. Synergistic assay

To assess the synergistic effects, cells were treated with p28, and CQ individually or in combination with p28 and CQ. Briefly, cells were seeded at a density of 5,000 cells per well in 100 μl of growth medium in 96-well plates and grown overnight. Cell viability was collected and percent inhibition was calculated as described before [7]. Data were analyzed using SynergyFinder Ver.3.0 [25].

2.7. Transmission electron microscopic images

After HCT116 or UISO-Mel29 cells were exposed to p28-GNPs or GNPs alone as a control (300 μg/ml) for 2–24 h, approximately 1 mm³ cell culture samples were fixed in 4 % phosphate-buffered glutaraldehyde. For in vivo tumor samples, p28-GNRs or GNRs alone at 5 mg/kg was injected i.v. to athymic mice with xenografted HCT116 tumor. After 2 and 24 h, tumors and surrounding tissues (skin) were resected and fixed. Cells from in vitro and tissues/tumors from in vivo preparations were washed with 2 % sucrose in 0.1 M Sorensen’s phosphate buffer at room temperature. Samples were then treated with 1 % osmium tetroxide in 0.1 M Sorensen’s phosphate buffer for 1 h at room temperature. After washing with dH₂O and dehydration with acetone, resin infiltration (EMBed 812 resin [EMS]) was performed as follows: 30 min, 2:1 mix of 100 % acetone:resin; 30 min, 1:1 mix of 100 % acetone:resin; 1 h, 1:2 mix of 100 % acetone:resin; and 1.5 h, 100 % resin. Samples were then placed in embedding molds and polymerized overnight at 60 °C. Sections (60 nm thick) were made using an ultratome and picked up on 200 mesh copper grids. TEM images of uranyl acetate- and lead citrate-stained samples were taken with a JEOL 1220 TEM (JEOL Ltd., Tokyo, Japan).

2.8. p53 transcription factor assays

HCT116 and UISO-Mel29 cells were treated with p28 at 20 μM, either in the presence or absence of 20.7 μM chloroquine (FisherScientific), for 16 h at 37 °C in a humidified atmosphere containing 5 % CO₂. After washing with PBS and collecting the cells, cytosolic and nuclear fractions were prepared using a nuclear extract preparation kit, following the manufacturer’s protocol (#40010, Active Motif, CA, USA). Protein was quantified using the Bradford Protein Assay (#5000001, Bio-Rad, CA, USA) and samples were diluted to equal loading volumes. The activation of p53 was quantified using TransAM p53 Transcription Factor Assay Kits (#41196, Active Motif), following the manufacturer’s instructions. Briefly, nuclear factions were added to a 96-well plate containing immobilized oligonucleotides with the p53 consensus binding site. A primary antibody, which recognizes an epitope of p53 exposed upon DNA binding, was applied, followed by a secondary antibody conjugated to horseradish peroxidase (HRP). The chemiluminescent signal was measured using a spectrophotometer at 450 nm, with a reference wavelength of 655 nm for quantification.

2.9. Animal models

All animal work was performed under a protocol approved by the University of Illinois at Chicago. Human colon cancer cell lines, p53^wt HCT116 (1.0 × 10⁶ cells/mouse), were injected subcutaneously on the right flank of 4- to 6-week-old male athymic mice. When tumors reached 10 mm in diameter, animals received 0.5 mg/kg of ICG-p28 via intravenous (i.v.) [26]. After 24 h, near-infrared (NIR) images of tumor were obtained by the FDA 510(k)-cleared NIR imaging system PDE neo (Photodynamic Eye; Mitaka USA, UT) as described previously [20]. For tumor growth studies, HCT116 (p53^wt) and the isogenic cell line (p53^null) were injected subcutaneously (1.0 × 10⁶ cells/mouse). Once the mice developed a tumor (5 mm in diameter), animals were randomized into control and treatment groups: p28 at 10 and 20 mg/kg (3.4, and 6.8 μmol/kg, respectively) i.p. daily for 4 weeks, 5-Fluorouracil (5FU) at 20 mg/kg i.p. for the first 3 weeks. For a combination approach, once the mice developed a tumor (5 mm in diameter), animals were randomized into control and treatment groups: PBS intraperitoneal (i.p.) injection (control), chloroquine (CQ; i.p., 25 mg/kg) for twice times per week, p28 (i.p., 10 mg/kg) for thrice times per week, and combination: chloroquine (i.p., 25 mg/kg) and p28 (i.p., 10 mg/kg) was co-administrated. At necropsy, tumors were dissected, weighed, and frozen.

2.10. Histological analysis

Resected tissues were placed in containers prefilled with buffered 3.7 % formalin and fixed for 24 h. Formalin was replaced with 70 % ethanol following fixation. Samples were paraffin-embedded, and blocks were cut into 4 μm-thick sections and mounted onto slides. A p53-downstream biomarker, p21 (WAF-1 Ab-5, Labvision) stained slides were analyzed blindly to the control and experimental (p28 at 10 mg/kg) groups of p53^wt and p53^null tumors. The pathologist was responsible for determining a tumor’s presence and general characteristics of resected tissues.

2.11. Anti-tumor response

Patient selection, study design, and assessments for p28 in the clinical trial have been published previously [17]. The patient with colon cancer (Stage IV) received three doses of p28 i.v. Tumor lesions were accurately measured in at least one dimension (longest diameter) with a minimum size of 10mm by CT scan. Tumor response and progression were evaluated by Response Evaluation Criteria in Solid Tumors (RECIST 1.1).

2.12. Statistical analysis

GraphPad Prism 9.0 was used for statistical analysis of data. Two-tailed Student t-tests were used for single comparisons, and group differences were evaluated using ANOVA. Data were presented as mean ± SEM, and all experiments were performed in triplicate, at least.

3. Results

3.1. Inhibition of tumor growth by p28

In our previous Phase I clinical trial of p28 monotherapy, fifteen patients with p53-positive solid tumors, including four colon and seven melanoma received at least one dose level of p28 [16,17]. Based on RECIST 1.1 analysis, two melanoma patients and one colon cancer patient demonstrated a partial response (PR), while one melanoma patient showed a complete response (CR) to p28 treatment among all participants in the trial. In preclinical settings, p28 produces a time and dose-related inhibition of various cancer types, including melanoma and breast cancer [19,27]. Indeed, p28 significantly reduced the weight of p53^wt HCT116 colon xenograft tumors in athymic mice in a dose-dependent manner (Fig. 1A). Our previous studies indicated that p28 enters cancer cells, p28 binds to the p53 DNA-binding domain, leading to an increase in intracellular levels of wild-type and some of mutant p53 [17,28], thereby inducing cell cycle arrest at G₂/M and then apoptotic cell death [29]. To determine the role of p53, another xenograft study was conducted with the HCT116 isogenic cell line with p53^null. Unlike p53^wt HCT116, p28 did not inhibit p53^null HCT116 colon xenograft tumors at any concentration (Fig. 1B). At the end of the treatment, tumors were evaluated for a p53-downstream maker, p21. Consistent with tumor growth, p21-positive cells were significantly increased in the p28 group compared to the control group of p53^wt tumors, but not for p53^null tumors (Fig. 1C). In the clinical study, the patient with colon cancer (p53 positive) showed a preliminary response to p28 treatment based on Response Evaluation Criteria in Solid Tumors (RECIST 1.1) (Fig. 1D). Thus, our findings showed that p53 significantly contributes to the p28’s anti-proliferative effects. Therefore, it is important to study the detailed mechanisms of cell penetration and intracellular trafficking of p28 to improve its efficacy.

Fig. 1.

Inhibition of colon cancer tumor growth by p28. p28 significantly reduced the weight of colon cancer HCT116 wild-type (A: p53^wt HCT116) colon xenograft tumors in a dose-dependent manner, unlike p53 null tumor (B: p53^null HCT116). (C) Tumor sections from PBS-treated (control) and p28-treated groups were stained for the p53-downstream marker p21 and subsequently quantified. (D) CT scan images showing tumor size (green line) before and after p28 treatment. 5FU: 5-Fluorouracil. n.s., not significant; ** p < 0.01, *** P < 0.001.

3.2. Cellular entry of p28 peptide into human colon and melanoma cancer cells is p53 independent

We previously demonstrated that p28 has two modes of cellular entry [7,8]. Unlike many cationic CPPs, p28 is not bound by cell membrane glycosaminoglycans (e.g., heparin) and preferentially penetrates cancer cells via the caveolae-mediated endocytotic pathway as well as the clathrin/caveolae-independent direct translocation [7,8]. Using confocal microscopy, we demonstrated that p28 preferentially penetrates human colon cancer cells (HCT116) and melanoma cells (UISO-Mel29), compared to normal colon cells (CCD33Co) and fibroblasts (Fig. 2A). As expected, Arg8, highly cationic CPP, penetrated both cancer and normal cells (Fig. 2A). Additionally, near-infrared fluorescent ICG dye-conjugated p28 was found to preferentially localize in xenografted tumors of HCT116 cells (Fig. 2B). As we reported, the primary anti-tumor effect of p28 is p53-mediated cell-cycle inhibition [7,28–30]. To evaluate whether the p53 expression status is important for cellular entry of p28, quantitative analyses were performed using isogenic cell lines of HCT116 colon cancer cells with either p53-positive or p53-null (negative) status. Since no significant differences were observed between the two cell lines, these findings indicate that the cellular entry of p28 occurs independently of p53 expression (Fig. 2C). More importantly, we confirmed that cellular entry of p28 was, at least partly, due to caveolae-mediated endocytosis. Cellular entry of p28 was blocked by methly-β-cyclodextrin (MβCD; inhibitor of caveolae-mediated endocytosis [31]), filipin (inhibitor of the caveolae pathway by directly binding to cholesterol within the caveolae membrane structure [32]), and nocodazol (inhibitor of the caveolae pathway by depolymerizing microtubules [33]) in HCT116 and UISO-Mel29 cells. These inhibitors did not completely prevent p28 entry, reducing internalization by approximately 60 % in cancer cells and 30–40 % in normal cells (Fig. 2D and Fig. S1). It suggests that the caveolae-mediated endocytosis pathway is the major route of p28 internalization and trafficking in cancer cells compared to normal cells.

Fig. 2.

Cell penetration of p28 peptide into cancer cells. (A) Human colon cancer cells (HCT116), melanoma cells (UISO-Mel29), normal colon cells (CCD33Co), and fibroblasts were incubated with Alexa Fluor 568–labeled p28 (AF-p28, 20 μM) or Alexa Fluor 568–labeled Arg8 (AF-Arg8, 5 μM) for 2 h. Cells were imaged using a confocal laser-scanning microscope. The red colour represents Alexa Fluor 568-labeled p28, while the green colour indicates the intrinsic autofluorescence of CCD33Co cells. Scale bar: 20 μm. (B) Athymic mice bearing s.c. HCT116 tumors (~10 mm in diameter) were intravenously injected with either PBS or 0.5 mg/kg of ICG-p28. Imaging was performed 24 h later using the FDA 510(k)-cleared NIR imaging system PDE neo. Red arrows indicate the tumor, while green arrow indicates the kidney. (C) HCT116 cells with wild-type p53 (p53^wt) and p53 null (p53^null) status were incubated with AF-p28 (20 μM) and Alexa Fluor 568–labeled Arg8 (AF-Arg, 5 μM) for 2 h, followed by analysis using a MoFlo Cell Sorter. (D) HCT116, UISO-Mel29, CCD33Co, and fibroblasts were pretreated with methyl-β-cyclodextrin (5 mM, MβCD) for 16 h, followed by incubation with AF-p28 (20 μM) for 2 h. Cells were analyzed using a MoFlo Cell Sorter. n.s., not significant.

3.3. Intracellular trafficking of p28

To explore the intracellular trafficking of p28 in cancer cells, we employed transmission electron microscopy (TEM) at high magnification using gold nanoparticles/nanorods. Gold nanoparticles (GNPs) were utilized due to their excellent potential as intracellular imaging agents, enabling detailed visualization of p28 movement within the cells [34]. We found GNP-p28 in vesicles (e.g., endosome) inside the cytoplasm space of HCT116 and UISO-Mel29 cancer cells for 2 h incubation in vitro (Fig. 3A–B). In control experiments using free GNP, only a minimal amount of GNP itself was detected in these cell lines under the same experimental conditions (Fig. S2A–B). In most cases, multiple GNP-p28 particles were observed within a single endosome. Notably, after 24 h of incubation, GNP-p28 particles were detected inside the nucleus (Fig. 3D–E), unlike free GNP (Fig. S2C–D). In general, the direct observation of CPPs trafficking into the nucleus in cancer cells remains poorly characterized. Furthermore, the evidence was confirmed in HCT116 xenografted tumors. It has been reported that the shape and size of GNPs significantly influence their ability to penetrate cells [35,36]. To eliminate potential artificial effects associated with particle shape and size, we used rod-shaped gold nanorods (GNRs) conjugated with p28, enabling more precise and unambiguous TEM observations in in vivo studies. Consistent with the results shown in Fig. 3A–D, TEM showed direct evidence of GNR-p28 particles trafficking into the nucleus of cancer cells in vivo (Fig. 3C, F). In contrast, very few GNR particles were detected in tumors (Fig. S3A). Also, neither free GNR nor GNR-p28 particles were found in tumor-surrounding tissues (skin) (Fig. S3B, C). This finding supports the role of p28 as a tumor-preferential vehicle for delivering cargo molecules specifically to cancer cells.

Fig. 3.

Intracellular trafficking of p28. (A, B) GNP-p28 was observed within vesicles (e.g., endosomes) in the cytoplasmic space of UISO-Mel29 and HCT116 cancer cells following 2 h of incubation. (C) GNR-p28 particles were found within vesicles (e.g., endosomes) in the cytoplasmic space of xenograft HCT116 tumors. (D, E) After 24 h of incubation, GNP-p28 was distributed within the cytoplasm and nuclei of UISO-Mel29 and HCT116 cancer cells. (F) GNR-p28 particles were detected trafficking into the nuclei of xenograft HCT116 tumors 24 h post-injection. Blue arrows indicate the locations of GNP-p28 and GNR-p28 particles. The yellow ahead-arrows indicate the nuclear membrane. M: mitochondria, N: nucleus.

3.4. Interrupt discrete steps of endocytosis

Macromolecules are often trapped within vesicles and endocytic organelles, which limit their access to the target sites such as the cytoplasm, mitochondria, or nucleus. We confirmed the intracellular trafficking and localization of p28. It involved caveolae-mediated transport (Fig. 2D, S1, 4A–F) through lysosome pathways, accumulation at the Golgi apparatus and endoplasmic reticulum (ER) (Fig. 4G–X). We then employed specific inhibitors targeting distinct steps of endocytosis and Golgi-mediated transport to further investigate these pathways. These inhibitors were used to investigate the roles of endosomal acidification and retrograde transport in the intracellular trafficking/localization of p28. First, the intracellular localization of p28 was examined after treatment with lysosomotropic amines, including chloroquine (CQ) and ammonium chloride (NH₄Cl) [37]. As controls, the intracellular localization of p28 in cancer cells without NH₄Cl or CQ treatment showed that a substantial amount of p28 was localized at the Golgi, with lesser amounts observed in nuclear and cytoplasmic spaces (Fig. 5A–D). In contrast, the cytoplasmic spaces and nuclear transport of p28 were markedly increased, suggesting that endosomal escape induced by CQ or NH₄Cl enhanced p28 redistribution in HCT116 and UISO-Mel29 cells (Fig. 5E–L). Next, we also investigated the cellular trafficking of p28 involving the Golgi complex. The endocytosis process and intracellular retrograde trafficking are closely interconnected, particularly for protein cargo molecules [38]. The trans-Golgi network (TGN) receives extracellular materials and recycled molecules from endocytic compartments [39]. Notably, nordihydroguaiaretic acid (NDGA) has been reported to stimulate the rapid retrograde movement of both Golgi stack and TGN membranes back to the endoplasmic reticulum [40]. To explore this effect, HCT116 and UISO-Mel29 cells were exposed to p28 followed by NDGA treatment. Our results showed that NDGA induced a redistribution of p28, remarkably enhancing its localization within the nuclear and cytoplasmic space of HCT116 and UISO-Mel29 cells (Fig. 5M–P).

Fig. 4.

Intracellular trafficking and localization of p28 in cancer cells. HCT116 and UISO-Mel29 cells were pretreated with AF-p28 (20 μM) for 2 h. Cells were then incubated with an anti-caveolin-1 antibody followed by anti-rabbit-AF488 (A–F), anti-Golgi-AF488 (G–L), ER-Tracker (M–R), and LysoTracker (S–X). Images were acquired using a confocal laser-scanning microscope. Scale bar: 20 μm.

Fig. 5.

Intracellular redistribution of p28 in cancer cells. HCT116 and UISO-Mel29 cells were pretreated with chloroquine (CQ, 10 μM), ammonium chloride (NH₄Cl, 1 mM) or nordihydroguaiaretic acid (NDGA, 25 μM) for 16 h followed by incubation with AF-p28 (20 μM) for 2 h. Cells were then imaged using a confocal laser-scanning microscope. (A-D) In the control group (no treatment), AF-p28 was predominantly localized to the Golgi or perinuclear region. Following CQ (E-H), NH₄Cl (I-L), and NDGA (M-N) treatment, AF-p28 was redistributed, expanding into the nuclear and cytoplasmic spaces. White arrows indicate the location of AF-p28 in each group. Scale bar: 20 μm.

3.5. Effect of p28 in combination with inhibitor that induce redistribution of p28

Among three intracellular trafficking inhibitors tested above, CQ is the only FDA-approved drug (e.g., for rheumatoid arthritis and other autoimmune diseases [41]). As reported previously [42,43], we confirmed the dose-dependent cytotoxic activity of CQ in cancer cells (Fig. S4). To investigate whether intracellular redistribution of p28 induced by an intracellular trafficking inhibitor could affect its anti-proliferative activity, HCT116 and UISO-Mel29 cells were treated with CQ in combination with various concentrations of p28. Concentrations of CQ were selected at IC_20–50. The results showed that CQ significantly enhanced the effect of p28 in both HCT116 and UISO-Mel29 cell lines (Fig. 6A, B). To elucidate the synergistic effect of p28 and CQ, HCT116 and UISO-Mel29 cells were treated with various concentrations of p28 and CQ for 24 h. As expected, the combination of p28 and CQ showed the clear synergistic effects (Fig. 6C, D). To investigate whether this improvement was due to the increased activity of p53 through intracellular redistribution of p28, the p53 activity in HCT116 and UISO-Mel29 cells was determined. The combination of CQ and p28 significantly increased the DNA-binding activity of p53 compared to treatment with either CQ or p28 alone in both cell lines (Fig. 6E), suggesting that the intracellular redistribution of p28 enhances its anti-proliferative activity by effectively activating tumor suppressor protein, p53.

Fig. 6.

Intracellular redistribution of p28 enhances its anti-cancer activity. (A-B) HCT116 and UISO-Mel29 cells were treated with chloroquine (CQ, 10 μM) in combination with varying concentrations of p28 for 24 h. Cell viability was assessed using the CCK-8 assay. (C-D) The synergistic effect of CQ and p28 was assessed by combining various concentrations of CQ and p28. The results were analyzed using SynergyFinder Ver.3.0. (E) Cells were treated with p28 (20 μM) in the presence or absence of CQ (20.7 μM) for 16 h. The activation of p53 was quantified using the TransAM p53 Transcription Factor Assay Kit. The p53 activity in each cell line exposed to p28 was used as a control (expressed as 1.0). n.s., not significant; *** p < 0.001.

To determine whether the combined effects of p28 and CQ observed in vitro also translate in vivo, a mouse xenograft study was performed (Fig. 7A). In mice bearing HCT116 xenograft tumors, CQ or p28 alone significantly (ANOVA, p < 0.001 vs. control) inhibited the tumor growth over a four-week treatment period, as expected. More importantly, tumor growth in the animals co-administrated with p28 and CQ was significantly (ANOVA, p < 0.001 vs. single agents) inhibited than that in animals treated with p28 or CQ alone (Fig. 7B). Compared with the PBS control, combined treatment with p28 and CQ significantly decreased the tumor volume by nearly half at week 4. At these treatment conditions, all treatment groups were not accompanied by a decrease in body weight (Fig. 7C). These results suggest that the intracellular trafficking inhibitor, CQ can be useful to improve the efficacy of macromolecules like p28 and can guide the development of a promising and effective drug-delivery strategy.

Fig. 7.

Inhibition of HCT116 xenograft tumor growth. (A) Schematic representation of the in vivo study design. Athymic mice with s.c. HCT116 tumors (~5 mm in diameter) were randomized into four groups: control (PBS) and treatment groups (CQ, p28, or a combination of CQ and p28). (B) Tumor volumes were quantified in mice treated with PBS, CQ (25 mg/kg), p28 (10 mg/kg), or a combination of CQ (25 mg/kg) and p28 (10 mg/kg). (C) None of the treatments significantly affected the body weight of the animals. Results are presented as mean ± SEM, with N = 6 in each group. *** p < 0.001 (One-way ANOVA).

4. Discussion

It has been suggested that hydrophilic molecules can generally be internalized by the classical endocytosis pathway. However, several CPPs can transport hydrophilic molecules to eukaryotic cells through energy-independent pathways [44,45]. CPPs (Protein Transduction Domains; PTDs) are short peptides (<30 residues) that have been predominantly evaluated and developed for disease diagnosis and therapy [46,47]. They are not only able to cross the lipophilic barrier of the cellular membranes, also deliver small particles as well as large molecules (e.g., contrast imaging agents, nanoparticles, proteins and DNA/RNA) inside the cells [47]. Given that the majority of published CPPs is cationic (net positive charge) due to their positive side-chains, electrostatic interaction between cationic CPPs and negatively charged surface of cells is believed to be the initial step for their cell penetration [47,48]. In contrast, p28 is a unique CPP as it is an anionic peptide with a tumor-targeting motif (Fig. 2) [20,26,49]. Moreover, p28 has a cytostatic activity via p53-mediated cell cycle arrest [28]. Thus, to investigate the detailed mechanisms of penetration and intracellular localization of p28 is important. Based on these data, it also helps modify the p28-treatment protocol or design new types of CPPs. Gold nanoparticles have been utilized in various biomedical applications, including as intracellular imaging agents, therapeutics, therapeutic delivery vectors, and in clinical trials involving PEGylated gold nanoparticles [50,51]. It has been suggested that cellular uptake as well as nuclear entry of gold nanoparticles was size and shape-dependent [52,53]. Small nanoparticles (e.g., <10 nm) can efficiently penetrate into tumor spheroids [52] and enter the nucleus [53]. Since our objective in this study was to determine the effect of p28 as a targeted-delivery vehicle, we used relatively large size with two different shapes of gold nanoparticles, spherical shape with diameter of 15 nm and rods shape with lengths of 73 nm × diameter of 25 nm. With this approach, we eliminate the possibility that p28-GNR nuclear entry is due to the GNR motif. Despite very few GNPs/GNRs showed nuclear entry, p28 can deliver even quite heavy molecules [GNP (20.5 MDa) and GNR (417 MDa)] to nuclear in cancer cells (Fig. 3). Further experiments are required to fully elucidate nuclear entry mechanism of p28.

Based on our previous studies, p28 penetrates cancer cells through the caveolin-dependent endocytosis as well as direct translocation [6,7,54]. In most cases, the significant part of p28 relies on the endocytic pathway entry, leading to its entrapment within endosomes and lysosomes (Figs. 2, 4, 5A–D). A substantial amount of p28 accumulated at the cell perinuclear structures was observed (Figs. 4, 5E–L). In addition, the retrograde transport inhibitor NDGA that blocks proteins/peptides transport from the ER to the Golgi apparatus complex [55] redistributed the p28 intracellular localization (Fig. 5M–P). The functioning of the Golgi apparatus and ER are both so closely linked to the lysosomes [38,56]. These results support the hypothesis that p28 can escape the endocytic compartments and that p28 entrapment inside the endosome and lysosome can be a critical rate-limiting factor. It suggests that enhancement of endosomal escape and effective p28 transport to the nuclear in cancer cells can improve the p28 efficacy without changing its dosing schedule.

The treatment of the FDA-approved drug, CQ leads to an increase in the pH inside the lysosomes and the inhibition of several enzymes such as proteases and nucleases [57]. This causes the rupture of the vesicle because of the increase in the osmotic pressure with increasing concentrations of the drug [58]. With this action, the nuclear localization of p28 was substantially increased by enhancing endosomal escape (Fig. 5E–L), suggesting that the localization of CPPs at target organelles can be optimized after penetration. Our previous studies have shown that p28 is a cell-penetrating peptide that significantly inhibits cell proliferation by stabilizing and increasing the levels of the tumor suppressor protein p53, leading to cell cycle arrest at the G₂/M phase, effectively preventing cancer cell growth [28,29]. Similarly, CQ has demonstrated anti-proliferative effects in various cancer types, including colon cancer, breast cancer, glioblastoma, head and neck cancer, lung cancer, and liver cancer [42,43]. Our results showed that the combination of p28 and CQ exhibits synergistic effects, suggesting that CQ not only promotes the intracellular redistribution of p28 but also enhances its synergistic effects (Fig. 6). In vivo results confirmed that the efficacy of p28 combined with CQ was significantly greater than that of either CQ or p28 alone (Fig. 7). This combination approach can be further optimized to improve efficiency. For example, the bioavailability/biodistributions of CQ and p28 are quite different in murine models. The elimination half-life t1/2, clearance, and volume of distribution of CQ are 46.6 h, 9.9 l/h/kg, and 667 l/kg, respectively [59]. Whereas the elimination half-life t1/2, clearance, and volume of distribution of p28 are 0.06 h, 2.2 l/h/kg, and 0.29 l/kg, respectively [19]. In addition, renal clearance is the primary route of p28 excretion, but hepatic clearance is the primary route for CQ [59]. More importantly, CQ does not have tumor-specificity [60]. As we demonstrated, the combination of p28 and CQ showed a substantial synergistic effect on cancer cells in vitro. These suggest that specific delivery of CQ or other similar compounds is an appropriate approach. Given that this is the first study demonstrating that entrapment of p28 within endosomes and lysosomes significantly hinders its intracellular delivery and co-treatment with lysosomotropic amines like CQ can overcome this obstacle by enhancing endosomal escape, thereby facilitating the effective intracellular transport of p28 and increasing its efficacy in cancer cells, it would be interesting to test the effect of p28 in combination with targeted CQ delivery in the future. Nevertheless, these results support evidence from the previous results shown in Fig. 6, indicating that the improved p28 efficacy was due to the effective transport of p28 to the nuclear in cancer cells.

In summary, multiple macromolecules, ICG dye (0.7 kDa), Alexa Fluor dye (1.0 kDa), GNP (20.5 MDa), and GNR (417 MDa), were used as cargo materials in this study. These results showed that p28 could deliver various types of macromolecules with a wide range of molecular weight at different experimental conditions, suggesting that p28 can be used as a useful vehicle. The entrapment of p28 within endosomes and lysosomes presents a significant bottleneck in its intracellular delivery. Co-treatment with lysosomotropic amines, such as CQ and ammonium chloride, can overcome this limitation by enhancing endosomal escape. This improvement facilitates the effective transport of p28 to the nucleus, thereby increasing its efficacy in cancer cells.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jconrel.2025.113660.

Data availability

No data was used for the research described in the article.