Enhanced tumor uptake and activity of nanoplex-loaded doxorubicin

Abstract

Doxorubicin (Dox) has widespread use as a cancer chemotherapeutic agent, but Dox is limited by several side effects including irreversible cardiomyopathy. Although liposomal Dox formulations, such as Doxil, mitigate side effects, they do not prolong survival in many patients. As a result, efforts have continued to discover improved formulations of Dox. We previously found that a peptide-based nanoplex delivered plasmid DNA efficiently to tumors in murine models. Unlike the majority of nanoparticles that depend solely on enhanced permeability and retention (EPR) for their transport into the tumor, our peptide-based nanoplex has a potential advantage in that its uptake primarily depends on neuropilin-1 receptor targeting. Because Dox binds to DNA, we tested whether this delivery platform could effectively deliver Dox to tumors and reduce their size. The nanoplexes increased the levels of Dox in tumors by about 5.5-fold compared to aqueous (free) Dox controls. Consistent with enhanced levels in the tumor, the nanoplex-Dox treatment had significantly greater anti-tumor activity. Whereas low dose free Dox did not reduce the size of tumors compared to untreated controls, the low dose nanoplex-Dox reduced the size of tumors by nearly 55% (p<0.001). The high dose nanoplex-Dox also inhibited the size of tumor significantly more than the comparable high-dose free Dox (p<0.001). Furthermore, apoptosis and proliferation markers (Ki67) of tumors observed in the different treatment groups correlated with their ability to inhibit tumor size. This study shows the efficacy of an NRP-1 targeted nanoplexes to deliver Dox to tumors in vivo and lays the groundwork for more complex and effective formulations.

1. Introduction

Dox alone or in combination with other chemotherapy is a common first-line therapy for numerous cancers including breast, ovarian, bladder, and lung. Although the mechanism of action for Dox is still being studied, several mechanisms have been proposed including intercalation into DNA disrupting gene expression, generation of reactive oxygen species, and inhibition of topoisomerase II [1]. Regardless of whether one or several of these proposed mechanisms are in play, Dox has cell-cycle specific activity and binds to DNA with high affinity.

The most severe long-term adverse effect of Dox therapy is irreversible cardiomyopathy, which is based on the total cumulative dose [2]. Increased levels of reactive oxygen species resulting in apoptosis in the heart appear to have a significant role in Dox cardiomyopathy. Efforts to mitigate Dox cardiomyopathy and other adverse effects have been undertaken via liposomal drug delivery [1, 3]. Several systemically administered liposomal Dox formulations have been investigated and approved including pegylated (Doxil^®) and unpegylated forms. Importantly, there is a better safety profile with liposomal Dox formulations [1, 9].

Nevertheless, there has been a debate as to whether pegylated or non-pegylated liposomal doxorubicin increased survival in cancer patients [1, 4] with one notable exception. In cancer patients with cardiac disease, it has been demonstrated that liposomal Dox does improve survival compared to conventional Dox [4]. These liposomal nanoparticles (NPs) enter the tumors preferentially through the EPR effect. The EPR effect is thought to result from a combination of leakiness of tumor blood vessels resulting in the flux of NPs from the blood into the tumor tissue and reduced numbers of lymphatic vessels in tumors associated with decreased drainage of NPs.

In a previous study, we determined that the histidine-lysine (HK) plasmid nanoplex transfected tumors efficiently [5]. However, unlike most nanoparticles, it is likely that HK nanoplexes do not depend on the EPR mechanism but instead target NRP-1. Because linear H2K peptides share a common amino acid sequence motif of –KXXK- (where X is any amino acid) with peptides that target NRP-1 and activate this pathway [5, 6], the H2K nanoplex is likely transported through this pathway in the tumor. In addition, antibodies, which target NRP-1, almost completely block the uptake of the H2K nanoplexes into the tumor, further validating that these nanoplexes enter tumors through this pathway [5]. By targeting NRP-1 with the H2K peptide, the nanoplexes intrinsically incorporate a tumor targeting ligand.

Based on its NRP-1 tumor targeting and Dox binding tightly to plasmid DNA, the HK plasmid nanoplex may be particularly well-suited to deliver Dox to tumors. In the current study, we determined that Dox-containing nanoplexes delivered Dox more effectively to tumors and had greater antitumor activity than the free Dox control.

2. Materials and methods

2.1. Animals

Female athymic mice (4–8 weeks) were purchased from Envigo (Indianapolis, IN, USA). Animal experiments were performed in accordance with regulations by the Institutional Animal Care and Use Committee.

2.2. Cell Lines

The malignant cancer cell, MDA-MB-435, was cultured and maintained in DMEM with 10% fetal calf serum and 20 mM glutamine.

2.3. Plasmids

The pCpG-TdTomato(TdT) plasmid was prepared for systemic injection as described previously [7]. The plasmids were purified with an EndoFree Maxi kit (Qiagen, Valencia, CA, USA).

2.4. Peptides

The linear HK peptides were synthesized by Genscript (Piscataway, NJ, USA). As previously detailed [5], sequences of HK peptides are the following: KHKHHKHHKHHKHHKHHKHK-amide capped (H2K); KHKHHKHHKHHKHHKHHKHK-CO2H (H2K-CO2H); KHHKHHKHHHKHHHKHHHKHHHKHHHKHHKHHK (H3K-33). The cRGD-PEG-H3K4b conjugate was synthesized as described [5] with an average of 4 ligands per H3K4b molecule.

2.5. Dox binding to DNA and NP

Fluorescence of the Dox solution (0. 1 μg, 1.84 μM) was determined alone or after addition of plasmid DNA (3.2 μg, 1.6 μg, or 0.8 μg) for 30 minutes. Similarly, the fluorescence of Dox in the nanoparticle (NP) was determined after addition of the HK peptide (H2K/H2KCO2H/H3K: wt/wt/wt: 4.5/0.5/1) to the Dox: DNA conjugate for 40 min.

To form the nanoparticle, the ratio of HK to plasmid DNA was about 1:2 (wt:wt). The fluorescence of Dox was measured at excitation of 489 and an emission of 582 nm (Synergy H1 microplate reader, Biotek, Winooski, VT, USA).

2.6. Release of Dox from the nanoplex and plasmids

The time-release profile of Dox from the nanoplexes was determined by the recovered Dox in the filtrate [8] using Microcon YM-50 filters (Millipore, Billerica, MA, USA). After measuring the fluorescence of Dox in the filtrates, the percent of Dox release was determined. In addition, release of Dox from supercoiled and degraded plasmids (DNAseI, 4 units, 4 h) was determined. After incubation, control buffer or acidic isopropanol (which releases Dox from DNA) was added overnight at 4°C. The fluorescence of these samples was then measured.

2.7. HK nanoplexes formation in vivo

To the plasmid DNA (36 μg in 140 μl of water), HK (H2K/H2KCO2H/H3K: 11.25/1.25/2.5 μg in 110 μl water) was added quickly and mixed by pipetting. Forty-five minutes after mixing, cRGD-PEG-H3K4b (1.63 μg in 30 μl) was added to nanoplex for 20 min prior to the intravenous injection of the nanoplex unless otherwise stated. Dox-containing HK nanoplexes were prepared similarly except that Dox (2.2 or 4.4 μg) was added to the plasmid for 30 minutes at RT prior to the addition of the HK peptides.

2.8. Particle size and zeta potential

The size (Z-A) and zeta potential of the nanoplexes were determined with the Zetasizer (Malvern, Westborough, MA, USA).

2.9. Dox detection in tumors in vivo

When the tumors were approximately 150 mm³, the mice were injected intravenously with either free Dox (4.4 μg) or Dox-containing nanoplexes (2.2 or 4.4 μg of Dox). Two hours after the injection of the different Dox therapies, the mice were euthanized, and the amount of Dox in tumors was determined (6 tumors per group). To quantitate Dox in tumors, we used the acidic isopropanol method [9]. The fluorescence of Dox from the tumor extracts was measured at an excitation of 489 and an emission of 582 nm. A standard curve was created by adding known amounts of Dox to tumor lysates (0.5, 0.25, 0.125 μg).

2.10. In-vivo tumor experiments

MDA-MB-435 cells (4 × 10⁶ cells per injection) were subcutaneously injected bilaterally into each mouse. After 4 days, when the subcutaneous tumors were about 45 mm³, the mice were separated into groups of 6 mice to determine the efficacy of treatments. The groups were as follows: untreated, NP alone, low dose Dox (2.2 μg/ injection), high dose Dox (4.4 μg/injection, low dose Dox NP (2.2 μg/injection), high dose Dox NP (4.4 μg/injection). The mice were given five intravenous injections of the various treatments. Nanoplexes were administered every 2 to 3 days (Monday, Wednesday, Friday). Tumor size was assessed with calipers before each injection and 2 days after the last injection; the size was calculated by formula ½ × d₁ × d₂², where d₂ is the smaller of the two measurements. Mice were weighed before each injection with tumor weight subtracted to obtain the final weight.

2.11. Histology of tumor and tissues

Two days after the last injection, mice were euthanized, and the deparaffinized tumors/tissues were stained with hematoxylin and eosin.

2.12. In-vivo apoptosis of tumors

The TUNEL assay was performed using the In Situ Cell Death Detection Kit (Roche Applied Science, Mannheim, Germany), following the manufacturer’s instructions. After the tumors were incubated with TdT enzyme and label solution, the nuclei were stained with DAPI (Abcam, Cambridge, MA).

2.13. Immunofluorescence of Ki67.

After the tumors were fixed and a heat-mediated antigen retrieval step was performed, the tumors were permeabilized (0.1% Triton X-100) for 10 min and then blocked with 5% BSA in 0.1% PBS-Tween for 2 h. The samples were then incubated with primary rabbit antibodies Ki67 (Abcam, ab 16667) or a monoclonal control (Abcam, ab 172730) at 1/200 dilution overnight at 4°C, followed by incubation at RT for 1 h with a goat anti-rabbit IgG (Alexa Fluor ®488) pre-adsorbed (ab150077) secondary antibody (shown in green). Nuclear DNA was labeled with DAPI.

2.14. Statistical Analysis

The results reported as mean ± standard deviation represent at least 6 separate data measurements, unless otherwise indicated. Results were analyzed using a one-way ANOVA analysis with Holm-Sidak multiple pairwise comparisons posthoc test. P <0.05 was considered statistically significant.

3. RESULTS

3.1. Dox is incorporated within the HK nanoplex.

Upon Dox binding to DNA, the fluorescence of Dox is quenched enabling determination of entrapment efficiency [10]. In Fig. 1, we validated that the tdT plasmid could bind to Dox as seen by quenching of its fluorescence. Upon binding to DNA, more than 95% of the fluorescence of Dox is quenched at several Dox:DNA ratios (1/32, 1/16, 1/8, and 1/4). Since Dox is positively charged, we investigated whether the addition of positively charged HK peptide would competitively compete with Dox for the DNA. At low HK to DNA ratios (~1:2) to form nanoplexes, there was no evidence that Dox was released from the DNA since quenching of the Dox was maintained.

Fig. 1. Dox is intercalated in the NP.

Compared to Dox alone, its fluorescence is significantly quenched with the addition of DNA. More than 95% of the fluorescence of Dox was decreased at several ratios of Dox: DNA (1:32, 1:16, 1:8, 1:4). With the addition of HK peptides to the Dox: DNA adduct, there was no effect on quenching. In the absence of DNA, the addition of the HK peptide to Dox did not reduce fluorescence compared to Dox alone.

3.2. Size and zeta potential of the nanoplexes.

The particle size and zeta potential of different HK plasmid-Dox nanoplexes were modestly different. The size of the HK nanoplex without Dox was slightly smaller than those with Dox. Although zeta potential of all the formulations had a negative surface charge, the zeta potential of the nanoplexes increased as more Dox was added (Table 1).

Table 1.

Biophysical properties of HK polyplexes in vivo

Dox/DNA1	Z-A2 size (nm)	ζ Potential (mv)
-3	145.8±13.2	−25.3±6.4
1/16	157.1±5.0	−23.5±3.2
1/8	172.8±6.7	−19.7±4.7

¹Ratio of Dox: DNA (w/w)

²Z-A, size of the polyplex prepared in water

³-, No Dox added to the DNA plasmid sample

3.3. Release of Dox.

The HK plasmid-Dox NP was stable up to 8 h to changes in the pH (pH 7.4 and 5.0). Less than 7% of Dox was released under these conditions (Fig. S1a). However, degradation of the plasmid does markedly enhance the release of Dox (Fig. S1b).

3.4. Dox-containing nanoplexes inhibit the size of tumors significantly more than free Dox.

Because the linear HK nanoplexes previously showed poor uptake and activity in vitro yet marked antitumor activity in vivo [5], we proceeded directly to in vivo studies. The low dose NP and high dose NP had significantly more antitumor activity than comparable dosages of free Dox (Fig. 2). Whereas the low dose Dox had no effect on the growth of tumors, the low dose NP inhibited the growth by about 60% at the sixth time point measurement (p<0.001). Similarly, the high dose NP inhibited the size of tumors by about 58% compared to the high dose of free Dox (p<0.001). Importantly, the Dox-containing NP showed a dose-dependency in their inhibition of the size of tumors. The high dose NP inhibited the tumor size by about 35% more than the low dose NP (p<0.01).

Fig. 2. Dox-containing nanoplexes significantly inhibit tumors.

Four days after mice received injections of tumor cells, mice were divided into various treatment groups. Each group had six mice with 12 tumors, and tumors were measured prior to each injection. There were 5 injections with a final measurement 2 days after the last injection. At the sixth measurement, the low dose NP (2.2 μg Dox) and high dose NP (4.4 μg Dox) were significantly more effective than comparable dosages of free Dox (p<0.001). The high dose NP inhibited the tumor size by about 35% compared to the low dose NP (p<0.05).

3.5. Dox levels in tumors were higher with the nanoparticle.

Consistent with the therapeutic response, the low dose and high-dose Dox NPs showed higher tumor levels of Dox than comparable dosages of free Dox (Fig. 3). The Dox levels in tumors of the low dose and high dose nanoplex were approximately 2.4- and 5.5-fold higher than those of the high dose free-Dox. The Dox levels of the high-dose HK nanoplex were statistically different than those of the free Dox (p<0.001).

Fig. 3. Tumor levels of Dox are increased with NP.

When the tumors were approximately 150 mm³, the mice were injected intravenously with free Dox (4.4 μg) or Dox-containing nanoplexes (2.2 or 4.4 μg of Dox). Two hours after injection of different Dox therapies, the mice were euthanized, and the amount of Dox in tumors (μg) was determined (6 tumors per group). The low and high dose NP had increased levels of Dox in the tumor compared to the free Dox. Furthermore, the high dose Dox nanoplex was significantly higher than the comparable free Dox (p<0.001).

3.6. Reduced growth and enhanced apoptosis in tumors treated with Dox-containing nanoplexes.

The histology, apoptosis, and mitotic index of tumors closely correlated to therapeutic responses with the different therapies. H&E staining showed large areas of necrosis in Dox-nanoplexes compared to the comparable free Dox (Fig. 4, S2). The expression of Ki67, a marker of proliferation, and apoptosis were also analyzed in tumor sections (Fig. 4, S3-5). We found that about 30% of the tumor cells in the low dose Dox nanoplex stained positive for Ki67 compared to 80% of cells in the low dose free Dox group. Moreover, only 14% of the tumor cells in the high dose Dox nanoplex group stained positive for Ki67 whereas about 57% of cells stained positive in the high dose free Dox group (Fig. 4, S4). In addition, enhanced apoptosis, as demonstrated by the positive TUNEL staining, was observed in the tumors of mice treated with the Dox containing nanoplexes (Fig. 4, S5). Thus, both decreased proliferation and increased apoptosis accounted for the reduction in tumor size with the Dox-containing nanoplexes compared to the free Dox group.

Fig. 4. In vivo histology.

After the last treatment injection, mice were euthanized, and tumor sections were stained with hematoxylin and eosin. There was significantly greater tumor necrosis in mice treated with Dox-containing nanoplexes. Scale bars, 20 μM (Upper panels). Immunofluorescent detection of Ki67. A fluorescent secondary antibody was used to visualize the primary antibody binding. There were increases in Ki67 expression in the free Dox treatment groups compared to Dox-containing nanoplex Scale bars, 10 μM (Middle panels). Apoptosis of tumor-induced by various treatment groups with the TUNEL assay. Increased apoptosis was observed in Dox-containing nanoplexes. Scale bars, 20 μM (Lower panels).

3.7. Toxicity.

At the dosages used in this study, no weight loss or other adverse effects were observed with Dox-nanoplexes (Fig. S6). Furthermore, no histological evidence of toxicity was found in normal tissues (heart, liver, lungs, spleen, and kidneys) from animals treated with either low or high dose Dox-nanoplexes (data not shown).

4. Discussion.

DNA was an early carrier of an anthracycline that showed antitumor activity [11]. When combined with DNA, Daunorubicin, which is closely related to Dox, prolonged the life-span of mice bearing L1210 compared to free daunorubicin [12]. This early study stimulated additional studies including the present one to combine an anthracycline (i.e., Dox) with nucleic acids to target tumors [13, 14].

Most reports indicate that Dox binds with higher affinity to 5’-GC-3’bp (or 5’-CG-3’) of DNA than to other bp [15-17]. Nevertheless, base pairs of adjacent 5’-CA-3’ have been reported to bind anthracyclines tightly and release them more slowly than adjacent GC base-pairs [18, 19]. Whereas numerous 5’-CG-3’ sequences (~110) are located within the TdT insert, both 5’-CA-3’ and 5’-GC-3’ sequences are scattered throughout the plasmid including the insert. To our knowledge, current methodologies (i.e., isothermal calorimetry) have not been done to compare the binding and release of Dox to GC-rich with CA-rich double-stranded oligonucleotides or plasmids. Such comparisons would be of interest to understand and construct an improved Dox-DNA nanoparticle. In addition to these nucleotide patterns with increased affinity toward Dox, plasmids have been noted to bind more tightly to Dox than oligodeoxynucleotides [17].

From previous studies, accumulation in the tumor of HK nanoplexes reaches its peak between 2 and 6 h [20]. Modification of the HK peptide to improve the stability of the nanoplex may enhance its half-life in the blood and tumor accumulation. In this study, nearly 10% of the Dox is within the tumor two hours after administering the high-dose Dox nanoplex, which is significantly greater than levels from the free Dox treatment.

Levels of Dox in the tumors correlated with the antitumor efficacy of the various treatment groups. Moreover, as shown in Fig. S1, the release of Dox is accelerated as the plasmid DNA is degraded. Since nanoplexes protect plasmids from enzymatic degradation, these HK-plasmid-NPs likely release doxorubicin with disassembly of the nanoplexes in the tumor. That Dox is quickly released from the nanoplex is further validated by the statistically greater antitumor activity observed after the first injection of the Dox-containing nanoplexes compared to free Dox.

In contrast, a problem concerning liposome preparations, including Doxil^®, is the inadequate release of Dox in the tumor. Increased tumor levels of Dox from the PEGylated liposomes in tumors do not always exhibit increased antitumor efficacy. For example, although PEGylated Dox liposome preparation had a 3 to 4-fold greater accumulation than free Dox in two different fibrosarcoma mouse models [21], free Dox had greater antitumor activity in one of these models and only had marginally less activity in the other. The authors suggested that the antitumor efficacy of Doxil was partly limited by the reduced release of Dox from the liposomes. To increase the bioavailability of Dox from Doxil, the copolymer P85 was administered at various times after mice with human ovarian cancer were given Doxil [22]. By disrupting the membranes of the liposomes, P85 aided in the release of Dox from Doxil. Although P85 administered 1 h after Doxil significantly improved the antitumor efficacy, P85 given 48 hours after Doxil was optimal. Unlike Doxil, the antitumor activities of the free Dox and the Dox nanoplexes correlated closely with the levels of Dox measured in the tumors. Despite these promising results, more PK studies are required for the HK Dox nanoplexes.

There are also potential disadvantages for NPs that depend solely on EPR for their entry into the tumor. Because blood vessel permeability may vary significantly within tumor tissue as well as between the primary and metastatic tumors, the antitumor efficacy of NPs, which depend solely on EPR, may be inconsistent [23]. An alternative pathway of entry which enables the NP to cross the vessel and traverse the tumor may circumvent this obstacle. A previous study provided strong evidence that the uptake of H2K polyplexes into the tumor was not dependent on ERP but instead on the NRP1 pathway [5]. Whereas NRP1 is frequently upregulated in tumor cells, NRP-1 expression is increased in nearly all tumor endothelial cells [24]. Thus, even when the NRP-1 transport system of tumor cells is not increased, upregulation of NRP1 in tumor endothelial cells ensures that the nanoplex traverses the vascular barrier of the tumor.

The current study lays the groundwork for improving the tumor-targeted delivery of Dox with a nanoplex. As shown in Fig. 2 and 4, dose-dependency was observed in the therapeutic response. Based on this and other preliminary studies, one could readily increase tumor levels of Dox by increasing the amount of DNA delivered with each injection or by repeating the injections. The current approach used 8- to 16-fold lower amounts of a linear peptide compared to those of a branched peptide in prior studies [25]. Notably, even at higher dosages of branched peptide and their nanoplexes, we observed no toxicity in mice. Consequently, one would anticipate that increased amounts of nanoplexes comprised of the linear HK can be safely administered with greater antitumor efficacy.

Moreover, the use of plasmid-based therapy raises the possibility of gene therapy coupled with Dox chemotherapy. Our preliminary data have demonstrated that Dox which is incorporated into shRaf-1 expressing plasmids (Dox: DNA-1/16 ratio) have a synergistic antitumor effect. Nevertheless, because Dox intercalates with DNA, it may decrease gene expression at increased Dox: DNA ratios, and thus, efforts are underway to determine the optimal DNA ratio. In addition to this approach, we are examining the antitumor efficacy of administering shRaf-1 tumor-inhibitory nanoplexes and Dox-containing nanoplexes separately. Further studies are required to determine the optimal delivery and strategy for shRaf-1 expressing and Dox plasmid nanoplexes.

In summary, this study demonstrated an effective yet easily assembled nanoparticle formulation of Dox with improved activity and without observed toxicity. Consistent with greater accumulation in tumors of Dox, the nanoplex had a greater antitumor efficacy compared to free Dox. Moreover, the enhanced tumor delivery of Dox from the nanoplexes corresponded to greater apoptosis and reduced mitosis in treated tumor xenografts.

• Doxorubicin (Dox)-plasmid conjugates and HK peptides formed nanoplexes

• Nanoplexes increased tumor levels of Dox compared to the free Dox

• The nanoplex-Dox had markedly greater anti-tumor activity than the free Dox

• Dose-response tumor inhibition was observed with the nanoplex-Dox treatments

• The nanoplex-Dox treatments significantly enhanced tumor apoptosis

Abbreviations:

Dox

Doxorubicin

EPR

enhanced permeability and retention

NP

nanoparticle

NRP-1

neuropilin-1

HK

histidine-lysine peptide

P85

Pluronic acid 85

shRaf-1

plasmid targeting Raf-1