Combination wt-p53 and MicroRNA-125b Transfection in a Genetically Engineered Lung Cancer Model Using Dual CD44/EGFR-targeting Nanoparticles

Abstract

Mutations in KRAS and p53 signaling pathways contribute to loss of responsiveness to current therapies and a decreased survival in lung cancer. In this study, we have investigated the delivery and transfection of wild-type (wt-) p53 and microRNA-125b (miR-125b) expressing plasmid DNA, in SK-LU-1 human lung adenocarcinoma cells as well as in Kras^G12D/p53^fl/fl (KP) genetically engineered mouse model of lung cancer. Systemic plasmid DNA delivery with dual CD44/EGFR-targeted hyaluronic acid (HA)-based nanoparticles (NPs) resulted in a 2- to 20-fold increase in wt-p53 and miR-125b gene expression in SK-LU-1 cells. This resulted in enhanced apoptotic activity as seen with increased APAF-1 and caspase-3 gene expression. Similarly, in vivo evaluations in KP mouse model indicated successful CD44/EGFR-targeted delivery. Tumor growth inhibition and apoptotic induction were also observed with (wt-p53+miR125b) combination therapy in KP tumor model. Lastly, J774.A1 murine macrophages co-cultured with transfected SK-LU-1 cells showed a 14- to 35-fold increase in the iNOS-Arg-1 ratio, supportive of previous results demonstrating a role of miR-125b in macrophage repolarization. Overall, these results show tremendous promise of wt-p53 and miR-125b gene therapy using dual CD44/EGFR-targeting HA NP vector for effective treatment of lung cancer.

Introduction

Lung cancer is the leading cause of cancer deaths worldwide, and non–small-cell lung cancers (NSCLC) accounts for majority (>85%) of all the lung cancer cases.^1,2,3 Cisplatin is often used in combination with other cytotoxic agents, such as etoposide or gemcitabine, as the first-line treatment of advanced NSCLC in the clinic. However, due to the rapid acquisition of multidrug resistance in NSCLC, alternative therapeutic strategies are in critical demand and being fervently explored in preclinical and clinical settings.^4,5,6

Among these investigative therapies, the use of nucleic acid constructs has enormous potential in improving the clinical outcomes in cancer. Effective systemic delivery of nucleic acid constructs for transfection in the tumor mass is a major challenge. The use of viral vectors for gene therapy in solid tumors is limited by severe toxicity concerns.⁷ Nonviral vectors, such as nanoparticle (NP)-based drug delivery, for nucleic acid–based therapies (NATs) can improve drug loading, enhance delivery, reduce toxicity, and decrease immunogenicity. Moreover, NPs surface decorated with tumor-targeting agents, such as antibodies and peptides, can minimize off-target effects.^8,9 We recently reported the effective use of CD44-targeting hyaluronic acid (HA)–based self-assembling nanosystems for delivery of NATs in combination with cisplatin for the treatment of drug-resistant lung cancer.⁴

KRAS (~20–30%) and p53 (also known as TP53 or Trp53) mutations (~50%)^2,3 are most frequently associated with NSCLC. Developing animal models with targeted mutations in these two genes can result in lung adenocarcinoma, which closely mimics the histopathological progression of the human NSCLC. The Kras^G12D p53^fl/fl model also known as the KP model of lung adenocarcinoma is one such model which aims to recapitulate human lung cancer progression.¹⁰ Intranasal or intratracheal administration of recombinant viral particles expressing Cre-recombinase activates the oncogenic KRAS allele and inactivates two conditional alleles of p53 (ref. 11) Although tedious, this model offers several key advantages over xenograft models, including tumor development in a native, immune-competent environment where the cancer cells obtain signals and challenges from surrounding tissues as would occur in patients. Thus, accurate evaluation of therapeutic efficacy can be obtained. Recent important study reported nonviral delivery of nucleic acid construct miR-34a in the KP model, which specifically downregulated its target genes and slowed tumor growth.¹⁰ However, restoration of miR-34a levels only partially substitutes downstream signaling cascades of p53. Hence, we decided to employ wt-p53 plasmid–encoding NPs (p53 NPs) and investigated the chemotherapeutic effects in the KP model.

Tumor tissues are structurally complex, consisting of multiple cell types derived from different cell lineages. Major component of this tumor microenvironment (TME) are tumor-associated macrophages (TAMs)^12,13 with different phenotypes such as M0, M1, and M2. Environmental cues can decide and exhibit a pleotropic mixture of both tumor-promoting and tumor-inhibitory macrophages. Clinical data indicate that TAMs are predominantly of M2 phenotype and contribute significantly in almost every stage of cancer development from initiation, immunosuppression, progression, and lymphangiogenesis to metastasis.^12,14 NATs have been evaluated to repolarize TAMs to the M1 (proinflammatory state) for anticancer therapy.^12,13,14,15 MiR-125b is one such nucleic acid–based therapy whose effectiveness in NSCLC is still controversial.¹⁶ Hence, in the current investigation, we have assessed the delivery of miR-125b–encoding plasmid with the aim of modulating TAM in the TME.

Along with nucleic acid–based therapy, we also assessed the targeting capability of our NPs for epidermal growth factor receptor (EGFR) due to its elevated expression in KP mouse model.^17,18,19,20 Thus, we have investigated a multipronged approach by using multiple nucleic acid constructs along with chemotherapy and a targeted delivery system for enhanced therapeutic efficacy. Although few reports in the past including our own work has captured these aspects individually in relation to NATs in cancer,^4,17 the current investigation answered these molecular conundrums in a mouse model which recapitulates the human disease, holding greater translational potential of these therapeutic strategies.

Results

NP formulation and characterization

We formulated self-assembling HA-PEI/PEG NPs encapsulating plasmid DNA (Figure 1a). The NPs showed a spherical morphology in TEM (Figure 1b) with particles in the size range of 200–400 nm. The polydispersity index of the NPs was ~0.2 to 0.3, and the surface charge was in the range of −35 to −38 mV (Figure 1c). The encapsulation efficiency of NPs was evaluated to be about 92–95% as determined by polyanion competition–based gel retardation as well as by PicoGreen dsDNA fluorometric assay (Figure 1c,d).

Figure 1.

Formulation and characterization of control and targeted hyaluronic (HA) acid–based nanoparticles (NPs). (a) Schematic representation of HA-poly(ethylene imine)/HA-poly(ethylene glycol)/HA-EGFR targeting peptide (HA-PEI/PEG/EGF) self-assembled NPs encapsulated with plasmid DNA. (b) Transmission electron microscopy of the self-assembled NPs showed a spherical morphology. (c) The NP diameter ranged from 235 to 410 nm, polydispersity index ranged from 0.2 to 0.3, and surface charge was in the range of −33 to −38 mV. (d) Electrophoretic retardation analysis of plasmid binding by HA-based NPs with the release of intact plasmid upon decomplexation with poly(acrylic acid). (e) Scheme for combination therapy with the addition of two different plasmids with a time lag.

Quantitative assessment of plasmid-encapsulated HA NP uptake

SK-LU-1 cells are human cell lines that express mutation in Kras^G12D and in p53, closely resembling oncogenic KRAS- and p53-deficient lung adenocarcinoma, both in genetically engineered mouse model and in humans. Hence, they provide an effective in vitro model to determine therapeutic effectiveness of our NP formulation. We quantified the gene expression levels for the plasmids after SK-LU-1 transfection with plasmid encoding p53 alone or in combination with miR-125b plasmid encapsulated in either CD44-targeted (CD44TNP) or dual (EGFR + CD44)-targeted NPs (DualTNP) (Figure 2a). The p53 expression was increased with CD44-targeted delivery system irrespective of whether it was administered alone (CD44-p53TNP) or in combination with miR-125b (CD44-comboTNP) (P < 0.05 versus untreated controls). However, the dualp53TNPs showed 10-fold increase in p53 expression as compared to CD44-p53TNPs. The dual-comboTNPs resulted in a 20-fold increase in p53 expression as compared to CD44-comboTNPs. Unlike p53 gene expression, the miR-125b increased twofold over baseline and remained the same irrespective of transfection with single or dual targeted NPs (CD44125bTNPs and dual125bTNPs). However, a 20-fold increase in miR-125b expression was observed when combined with p53 plasmid (dual-comboTNPs versus dual-125bTNPs; P < 0.05). Lastly, the dual-comboTNPs also exhibited higher miR125b expression as compared to CD44-comboTNPs (P < 0.05).

Figure 2.

Quantitative transfection efficiency using qRT-PCR of the expression profiles of (a) wild-type p53 and (b) microRNA-125b (miR-125b) in SK-LU-1 cells following transfection with plasmid DNA encapsulated in hyaluronic acid (HA)-based nanoparticles(NPs). CD44 T p53 NP, CD44-targeting NPs encapsulating p53 plasmid; CD44 T 125b NP, CD44-targeted NPs encapsulating miR-125b encoding plasmid; dual (CD44/EGFR) T p53 NP, CD44 and EGFR-targeted NP encapsulating p53 encoding plasmid; dual T 125b NP, CD44 and EGFR-targeted NPs encapsulating miR-125b encoding plasmid, CD44 T combo NP, CD44-targeted NPs encapsulating p53 or miR-125b encoding plasmid given as combination therapy with an 18-hour time lag; dual T combo NP, CD44 and EGFR-targeted NPs encapsulating p53 or miR-125b encoding plasmid provided as combination therapy with an 18-hour time lag. TP53 mRNA levels are normalized to β-actin, whereas the miR125b levels are quantified by normalization against U6 snRNA levels. Both p53 and miR125b are quantified against naked plasmid-treated cell samples using relative quantification by ΔΔCt method. Asterisk indicates comparison against naked plasmid-treated cell controls. Data represented as fold changes mean ± SEM, n = 6. (-) indicates comparisons between different groups with an observed statistical significance of P < 0.01. **P < 0.001 comparison against naked plasmid-treated cell controls.

Since few studies have indicated a dynamic interaction between miR-125b and p53 expression,^21,22 we investigated whether combination therapy with both plasmids would interfere/alter gene expression of either plasmids. We observed that when both the plasmids were delivered concurrently, the expression of miR-125b was reduced (Supplementary Figure S1). Hence, we employed a strategy with time lag introduced between the deliveries of the two plasmids. No significant changes were observed with p53 expression, but miR-125b was significantly reinstated following a time lag of 18 hours (Supplementary Figure S1 and Figure 1e). Thus, time-lag approach was employed when combination therapy was administered both in vitro and in vivo.

Combination therapy with dualTNPs induces apoptosis in SK-LU-1 cells

Following the time-lag strategy, we measured the apoptotic markers in the SK-LU-1 cells posttransfection with nanovectors encapsulated with plasmids coding for either p53 alone or in combination with miR-125b (Figure 3a). Dual-p53TNP resulted in elevated caspase-3 levels; however, when treated in combination with miR-125b, both the CD44-comboTNPs and dual-comboTNPs resulted in elevated caspase-3 levels (P < 0.05; Figure 3a). A corresponding increase was also observed in the APAF-1 and PUMA levels (P < 0.05; Figure 3a). Interestingly, dual-125bTNPs also showed a ~3-fold increase in APAF-1 mRNA expression (P < 0.05). For the antiapoptotic marker analysis, treatment with dual-comboTNPs therapy resulted in decreased levels of both the antiapoptotic genes, namely, Survivin and Bcl-2 (P < 0.05). No significant changes were observed with other treatment groups (Figure 3b). Hence, these results support the previous results and confirm that a time lag between deliveries of the two plasmids is permissive of the intended therapeutic approach for apoptosis induction in SK-LU-1 cells. It is important to mention that these apoptotic markers were also evaluated post transfection with appropriate controls (i.e., naked plasmid–encapsulated HA particles and lipofectamine-encapsulated plasmid DNA; Supplementary Figure S2).

Figure 3.

Analysis of (a) proapoptotic and (b) antiapoptotic genes SK-LU-1 posttransfection with plasmid-encoding p53, 125b encapsulated in HA nanoparticles (NPs) following single and combination therapy. Normalized to β-actin and untreated control samples using relative quantification by ΔΔCt method. Asterisk indicates comparison against control = 1. One-way analysis of variance followed by post hoc t-test with multiple comparisons, *P < 0.05. Data represented as mean ± SEM, n = 6. CD44 T p53 NP, CD44-targeting NPs encapsulating p53 plasmid; CD44 T 125b NP, CD44-targeted NPs encapsulating miR-125b encoding plasmid; dual T p53 NP, CD44 and EGFR-targeted NP encapsulating p53 encoding plasmid; dual T 125b NP, CD44 and EGFR-targeted NPs encapsulating miR-125b encoding plasmid; CD44 T combo NP, CD44-targeted NPs encapsulating p53 or miR-125b encoding plasmid given as combination therapy with an 18-hour time lag; dual T combo NP, CD44 and EGFR-targeted NPs encapsulating p53 or miR-125b encoding plasmid provided as combination therapy with an 18-hour time lag. (-) indicates comparisons between different groups with an observed statistical significance of P < 0.01.

Repolarization of J774.A1 macrophages following co-culture with transfected SK-LU-1

The polarization of TAMs can be stimulated in vitro with stimulants such as IFN-γ, LPS, and IL-4. We characterized macrophage polarization using the iNOS/Arg1 ratio. The inducible nitric oxide synthase (iNOS) is tightly correlated with M1 phenotype during inflammation, whereas arginase-1 (Arg1) is generally involved with anti-inflammatory role in the M2 phenotype. Arg1 uses the same substrate as iNOS, and hence, only one is preferably synthesized in the macrophages, thus allowing iNOS/Arg1 to be a good marker of the repolarization of macrophages. We measured the iNOS/Arg1 ratio in activated J774 macrophages post IL-4–mediated M2 stimulation and co-cultured these cells with SK-LU-1 cells post transfection with NPs (Figure 4). Co-culturing the M2-polarized J774 cells with SK-LU-1 cells transfected with NPs coding for miR125b or p53 plasmid showed a ~3–5-fold increase in iNOS/Arg-1 ratio in J774 macrophages (P < 0.05, relative to nontransfected but M2-activated cells). On the other hand, co-culturing with SK-LU-1 cells transfected with CD44-comboTNPs showed ~14-fold increase in iNOS/Arg-1 ratio in J774 macrophages (P < 0.05). But dual-comboTNPs showed ~35-fold increase in iNOS/Arg-1 ratio (P < 0.05). Other cytokines and interleukins, e.g., TNF-α, IL-1β, IL-6, showed a similar effect (Supplementary Figure S3).

Figure 4.

iNOS/Arg-1 mRNA expression ratio post repolarization of IL-4–stimulated J774 cells post co-culture with p53 and 125b transfected SK-LU-1 cells. Normalized to β-actin and untreated control samples using relative quantification by ΔΔCt method. Asterisk indicates comparison against control = 1. One-way analysis of variance followed by post hoc t-test with multiple comparisons, *P < 0.05. Data represented as mean ± SEM, n = 6. CD44 T p53 NP, CD44-targeting nanoparticles (NPs) encapsulating p53 plasmid; CD44 T 125b NP, CD44-targeted NPs encapsulating miR-125b encoding plasmid; dual T p53 NP, CD44 and EGFR-targeted NP encapsulating p53 encoding plasmid; dual T 125b NP, CD44 and EGFR targeted NPs encapsulating miR-125b encoding plasmid; CD44 T combo NP, CD44 targeted NPs encapsulating p53 or miR-125b encoding plasmid given as combination therapy with an 18-hour time lag; dual T combo NP, CD44 and EGFR-targeted NPs encapsulating p53 or miR-125b encoding plasmid provided as combination therapy with an 18-hour time lag. (-) indicates comparisons between different groups with an observed statistical significance of P < 0.01.

In vivo evaluation of plasmid expression following delivery in HA NPs in genetically modified lung cancer mice models

After confirming that our NPs could effectively deliver plasmids in vitro, induce apoptosis, and stimulate macrophage repolarization, we evaluated the effective delivery of our NP formulation in a genetically engineered KP mouse model. A scheme we used for the induction of tumor in these mice, as well as the treatment strategy employed, is depicted in Figure 5a. Consistent with our in vitro approach, we also used a time lag for the in vivo studies. Cisplatin therapy was also introduced in combination with our NP formulation, not only to compare our treatment strategy to current therapeutic approach but also to enhance its efficacy.²³ At the end of therapy, the tumor samples were isolated and evaluated for efficacy. CD44-p53TNPs encapsulating p53 plasmid treatment resulted in threefold increase in p53 expression relative to phosphate-buffered saline (PBS)–treated animals (Figure 5b). The combination therapy with miR-125b plasmid did not alter p53 expression when delivered in CD44T-comboNPs. However, both the dual-p53TNPs alone and dual-comboTNPs showed ~9–10-fold increased p53 expression relative to the expression levels of PBS-treated animals (P < 0.05; Figure 5b). The primers were designed specifically for the flag sequence on plasmid, allowing us to differentiate between the native and delivered p53 gene. Corresponding changes in protein levels of p53 expression were also confirmed using western blot analysis with an antiflag antibody (Figure 5c), and quantitative analysis of the western blot analysis was also performed (Figure 5c). Lastly, the CD44-125bTNPs showed a ~7-fold increase, and the dual-125bTNPs showed a ~10–12-fold increase in miR-125b expression. Similarly, the treatment using the combination therapy, i.e., p53 + miR-125b also showed a ~7-fold and ~12-fold increase with CD44-comboTNPs and dual-comboTNPs, respectively (P < 0.05; Figure 5d). It is important to mention that no inflammatory effect of the null plasmid was noted in any of our previous in vivo study. Also, our extensive previous work on HA-based NPs did not indicate any nonspecific effects, including any inflammatory response measured by absence of macrophage repolarization, as well as unaltered interleukin and cytokine levels including TNF-α levels.^5,6,24,25,26 As such, these controls were not included in the in vivo experiments.

Figure 5.

In vivo results: (a) dosing scheme in KP mouse model 10 weeks post Ade-Cre administration. The animals were treated with plasmid DNA encapsulated in HA NP on days 1, 3, and 5 for weeks 2, 4, and 6. When combination therapy was to be provided, p53-NPs were administered 18 hours after the last miR-125b plasmid DNA NP dose. In vivo gene expression studies: (b) flag p53 and (d) 125b expression following treatment of tumor-induced mice with plasmid encapsulated HA NPs as single or combination therapy. Normalized to β-actin and untreated control samples using relative quantification by ΔΔCt method. Asterisk indicates comparison against control = 1. One-way analysis of variance followed by post hoc t-test with multiple comparisons, *P < 0.05. Data represented as Mean ± SEM, n = 5. (c) Protein expression for flag-p53 using western blot analysis in tumor samples. Quantification performed as described in the Materials and Methods section. CD44 T p53 NP, CD44-targeting NPs encapsulating p53 plasmid; CD44 T 125b NP, CD44-targeted NPs encapsulating miR-125b encoding plasmid; dual T p53 NP, CD44 and EGFR-targeted NP encapsulating p53 encoding plasmid; dual T 125b NP, CD44 and EGFR-targeted NPs encapsulating miR-125b encoding plasmid; CD44 T combo NP, CD44-targeted NPs encapsulating p53 or miR-125b encoding plasmid given as combination therapy with an 18-hour time lag; dual T combo NP, CD44 and EGFR-targeted NPs encapsulating p53 or miR-125b encoding plasmid provided as combination therapy with an 18-hour time lag. (-) indicates comparisons between different groups with an observed statistical significance of P < 0.01.

Combination therapy with dualTNPs along with cisplatin reduces tumor growth and tumor cell-proliferation in KP mouse model of lung cancer

The in vivo efficacy of our NP formulation was evaluated using immunohistological staining (Figure 6). Lung tissue from PBS-treated mice showed highly aggressive tumor development with several tumor nodules showing grade 4 tumors along with some invasive tumor nodules. The treatment with cisplatin also resulted in tumor nodules similar to grade 3 and 4 tumors. Lung tissues from mice treated with single plasmids resulted in tumor nodules with grade 2 and 3 nodules with an exception of CD44TNPs encapsulating p53, which showed relatively lower tumor burden. However, treatment with combination therapy resulted in relatively fewer tumor nodules, both in CD44-comboTNPs and dual-comboTNPs. Hence, our combination and targeted approach resulted in reduced tumor aggressiveness/progression. No associated liver toxicity was observed (Supplementary Figure S5).

Figure 6.

Tumor growth and proliferation analysis. (a) H&E and (b) Ki67 images of lung sections indicating tumor growth and proliferation in lung tissues PBS and plasmid encapsulated HA nanoparticles (NPs) as single or combination therapy. CD44 T p53 NP, CD44-targeting NPs encapsulating p53 plasmid, CD44 T 125b NP, CD44-targeted NPs encapsulating miR-125b encoding plasmid; dual T p53 NP, CD44 and EGFR-targeted NP encapsulating p53 encoding plasmid; dual T 125b NP, CD44 and EGFR-targeted NPs encapsulating miR-125b encoding plasmid; CD44 T combo NP, CD44-targeted NPs encapsulating p53 or miR-125b encoding plasmid given as combination therapy with an 18-hour time lag; dual T combo NP, CD44 and EGFR-targeted NPs encapsulating p53 or miR-125b encoding plasmid provided as combination therapy with an 18-hour time lag. H&E, Hematoxylin and eosin.

Next, we performed immunostaining with Ki-67 dye (Figure 6b), to investigate if there were changes in the proliferation capacity of tumor cells posttreatment with our therapy. As evident, the PBS-treated tissues indicated elevated Ki-67 staining and fluorescence, indicative of increased tumor cell proliferation. However, all the other treatment groups showed decreased proliferative activity relative to PBS-treated groups. Most importantly, a marked decrease or almost complete absence of proliferation was observed with the combination therapy in CD44-targeted and dual targeted approach, i.e., CD44-comboTNPs and dual-comboTNPs (Supplementary Figure S6).

Combination therapy with dualTNPs along with cisplatin induces apoptosis in KP mouse model of lung cancer

We analyzed if the changes in decreased tumor growth were associated with apoptotic induction (Figure 7a,b). Among proapoptotic genes, elevated levels of APAF-1, BAX, and caspase-3 were observed with CD44-125bTNPs and cisplatin (P > 0.05). On the other hand, treatment with both CD44-p53TNPs and dual-p53TNPs increased proapoptotic genes including APAF-1 and caspase-3 by about ~3–5 folds (P < 0.05). While in combination with miR125b, this increase was substantiated, and a ~4–7-fold increase was observed in all proapoptotic genes including APAF1, Caspase-3, Bax, and PUMA (P < 0.05), and the dual-comboTNPs had higher efficacy compared to the single CD44-comboTNPs. Similarly, the combination therapy with the dual targeted approach (dual-comboTNPs) resulted in greater than 50% decreased expression among antiapoptotic genes Survivin and Bcl2 (P < 0.05 versus PBS).

Figure 7.

Analysis of (a) proapoptotic and (b) antiapoptotic genes in lung tissues following treatment with PBS and plasmid encapsulated HA nanoparticles (NPs) as single or combination therapy. Normalized to β-actin and PBS control samples using relative quantification by ΔΔCt) method. Asterisk indicates comparison against control = 1. One-way analysis of variance followed by post hoc t-test with multiple comparisons. *P < 0.05. Data represented as mean ± SEM, n = 5. CD44 T p53 NP, CD44-targeting NPs encapsulating p53 plasmid; CD44 T 125b NP, CD44-targeted NPs encapsulating miR-125b encoding plasmid; dual T p53 NP, CD44 and EGFR-targeted NP encapsulating p53 encoding plasmid; dual T 125b NP, CD44 and EGFR-targeted NPs encapsulating miR-125b encoding plasmid; CD44 T combo NP, CD44-targeted NPs encapsulating p53 or miR-125b encoding plasmid given as combination therapy with an 18-hour time lag. dual T combo NP, CD44 and EGFR-targeted NPs encapsulating p53 or miR-125b encoding plasmid provided as combination therapy with an 18-hour time lag. (-) indicates comparisons between different groups with an observed statistical significance of P < 0.01.

Combination therapy with dualTNPs along with cisplatin induces a proinflammatory tumor microenvironment in lung tissue in KP mouse model

As described previously, our in vitro studies depicted a proinflammatory induction of macrophages after treatment with the combination therapy, thus we aimed to assess if similar trends were observed in vivo. Both the CD44- and dual-p53 TNPs showed ~6–8-fold increase in iNOS/Arg-1 ratio (Figure 8); however, both CD44- and dual-125b TNPs showed ~11–18-fold increase in iNOS/Arg-1 ratio (P < 0.05 versus PBS; Figure 8). Lastly, combination therapy showed ~23–46-fold increase in iNOS/Arg-1 ratio (CD44-comboTNPs and dual-comboTNPs versus PBS, P < 0.05). Similar changes were also observed in the other proinflammatory markers, i.e., TNFα and IL-1β (Supplementary Figure S4). The above results indicate an induction of proinflammatory tumor microenvironment, which could possibly synergize with p53-mediated apoptosis to prevent tumor growth.

Figure 8.

iNOs/Arg-1 gene expression following treatment with cisplatin and other plasmid HA nanoparticles (NPs). Normalized to β-actin and untreated control samples using relative quantification by ΔΔCt method. Asterisk indicates comparison against control = 1. One-way analysis of variance followed by post hoc t-test with multiple comparisons. *P < 0.05. Data represented as mean ± SEM, n = 5. CD44 T p53 NP, CD44-targeting NPs encapsulating p53 plasmid; CD44 T 125b NP, CD44-targeted NPs encapsulating miR-125b encoding plasmid; dual T p53 NP, CD44 and EGFR-targeted NP encapsulating p53 encoding plasmid; dual T 125b NP, CD44 and EGFR-targeted NPs encapsulating miR-125b encoding plasmid; CD44 T combo NP, CD44-targeted NPs encapsulating p53 or miR-125b encoding plasmid given as combination therapy with an 18-hour time lag; dual T combo NP, CD44 and EGFR-targeted NPs encapsulating p53 or miR-125b encoding plasmid provided as combination therapy with an 18-hour time lag. (-) indicates comparisons between different groups with an observed statistical significance of P < 0.01.

Discussion

The use of NATs and its delivery using viral vectors has several limitations, e.g., insertional mutagenesis, toxicity concerns, limited cargo capacity, and manufacturing challenges. However, alternative approaches to gene delivery using nonviral biomaterials such as inorganic NPs, cationic lipids, liposomes, and polymers have been limited either due to low efficiency resulting in limited efficacy or due to limited preclinical evaluation, currently only evaluated in xenograft animal models. In the current study, we have demonstrated that HA polymer–based NP encapsulating plasmid DNA can be delivered systemically in an autochthonous mouse model of lung cancer, and more importantly, this dual targeted approach can elicit a potent antitumor response. Specifically, we illustrate the delivery and effectiveness of p53- and miR-125b-expressing plasmids in both in vitro and in vivo models of KP lung tumors. However, in addition to our usual CD44-targeting capability demonstrated before,^5,6 we also surface functionalized our NPs with EGF peptide to enable higher selectivity to cancer cells. Significant proportion of human tumors including lung tumors overexpress EGFR and carry mutations in KRAS and p53 genes, hence developing therapies similar to our current strategy targeted to tumors with such mutations would have a high clinical translation value, not only for NSCLC but also to other tumors.

The loss of p53 function results in chemoresistance, and restoration of endogenous expression of wt-p53 influences reliable cancer response to chemo- and radiation therapy in solid tumors including lung cancer.^27,28,29 Our recent study reported the successful induction of apoptosis in vivo after wt-p53 transfection in tumor cells.²³ In the current study as well, the induction of proapoptotic genes following NP treatment both in vitro and in vivo strongly indicates that our delivery system could successfully elevate the wt-p53 gene expression, resulting in induction of endogenous apoptotic pathway in cancer cells. MiR-125b provided a synergistic effect and resulted in greater apoptotic induction as compared to single therapy, as evident from the immunohistology and immunostaining studies indicating decreased tumor aggressiveness/proliferation. Hence, combination therapy of p53 with miR-125b in dualTNPs did not interfere with each other and resulted in elevated apoptosis, reduced tumor growth, and cell proliferation.

Previously, some studies have reported interaction between miR-125b and p53;^16,21,22 hence, we explored the possibility of introducing a time lag between the administration of the two plasmids and were successful in inducing a synergistic effect. Taking advantage of this approach, the dual targeted NPs with combination therapy showed an elevated antitumor effect in both SK-LU-1 cells and the lung tumor tissues isolated from KP mouse models. We also ensured that use of the two plasmids with the time-lag would not introduce resistance to cisplatin therapy. Hence, these results provide a preliminary investigation into a novel yet simplistic strategy to counteract the opposing effects of NATs, align them for synergistic therapeutic efficacy, and prevent resistance to current treatment regime, e.g., cisplatin. This is highly critical and still unexplored especially for the KRAS- and p53-specific lung tumors.

A dual role has been proposed for miR-125b in tumor tissues.¹⁶ However, in the current study, we report the antitumor effects of miR-125b in lung tissues in combination with p53. We suggest that the apoptotic effects of miR-125b might be attributable to its role in macrophage polarization and activation toward a proinflammatory status. Previous studies have reported such antitumor role of miR-125b along with other such proinflammatory microRNAs, e.g., miR-146a,³⁰ by modifying the macrophage activation and polarity resulting in altered inflammatory state of tumor tissue. Here, the combination therapy with dual targeted NPs resulted in a more proinflammatory tumor microenvironment status with elevated iNOS/Arg1 ratio, TNF-α, and IL-1β levels, and reduced IL-10 levels. Additionally, in our in vitro studies, post IL-4 stimulated M2 macrophages when co-cultured with dualTNPs-transfected SK-LU-1, the J774 cells showed an increase in the iNOS/Arg-1 ratio and a shift toward M1 phenotype. The specific mechanisms employed by miR-125b for induction of proinflammatory changes in TAMs are still not fully understood. Previous studies have indicated that miR-125b can repress the interferon regulatory factor during activation of M1 macrophages³¹; however, concrete evidence is still not established. Based on the current results, we suggest that combination therapy of p53 with miR-125b can also influence the tumor microenvironment as evident by an increase in proinflammatory marker status. Hence, an amplified reduction in tumor progression and proliferation is a result of elevated antitumor effect supplemented by the apoptotic effects of p53 on cancer cells. As evident from Ki-67 immunostaining studies, a greater antitumor effect and decreased tumor progression was observed with combination therapy as compared to single therapy of either plasmid. Modulating TME for antitumor effect has been explored before; however, we provide preliminary evidence for a synergistic approach by combining the tumor cell apoptosis along with modulation of the tumor microenvironment as a potential strategy for cancer therapy.

Interestingly, apart from the macrophage repolarization, miR-125b has also previously been indicated to induce apoptosis in cancer cells, including downregulation of the antiapoptotic genes, e.g., Bcl-2. One of the possible contributing mechanisms to this downregulation is based on the presence of a binding sequence for miR-125b in the promoter region of Bcl-2 (ref. 32) and is reported to inhibit the antiapoptotic activity of Bcl-2, similar to the results of our current study. Moreover, similar binding site is also identified in the 3′-UTR region of Survivin gene,³³ although no other studies except ours have been performed to evaluate the effects of miR-125b on survivin. It is well known that p53 regulates the levels of all the antiapoptotic genes as evident from our results, and hence the downregulation of antiapoptotic pathway might be an additional synergistic action of these two plasmids. Lastly, initiation of p38-MAPKinase signaling is also reported to be involved for mediating the effects of miR-125b as well as p53 (ref. 21) and could be another possible contributive mechanism. However, the current study was focused on the successful use of NPs for in vitro and in vivo delivery of these plasmids with high specificity, and future studies could be performed to investigate the mechanisms involved in mediating the synergistic effect of these two plasmids.

Overall, the current study builds on our extensive previous work using nonviral HA-based NP drug delivery systems for in vitro and in vivo cancer therapy.^5,6,17 However, our current approach greatly extends our previous work and overall the application of HA-based drug delivery systems for use in cancer therapeutics, as well as for TME reprogramming, especially in KP lung cancer model. It is also important to mention that the main goal of the current study was to investigate the effective delivery and therapeutic characterization of the combination therapy of miR-125b and p53 plasmids with dual targeted approach with highly aggressive tumor. Future studies can involve an investigation of detailed mechanism behind these synergistic effects; for example, the communication between tumor cell and macrophages in vivo and potential transfer of miR-125b or p53 from tumor cell to macrophages in microvesicles/exosomes. Hence, this is the first study aimed to assess the synergistic potential of inducing cancer cell apoptosis as well as reprogramming the tumor microenvironment, resulting in elevated antitumor effect in the KP mouse model of lung cancer.

Conclusions

Due to the limitations of current small-molecule cytotoxic therapies, NATs have been investigated as a potential therapeutic strategy for drug-resistant lung cancer. In this study, we have investigated wt-p53–mediated cancer cell apoptosis and delivery of plasmid DNA encoding miR-125b to evaluate cancer cell–mediated macrophage repolarization using NP-based drug delivery strategies. Both in vitro and in vivo studies indicated successful uptake, gene expression, and apoptosis of cancer cells postdelivery with plasmid DNA encapsulated HA particles. In vivo delivery in a KP mouse model, which aims to recapitulate stepwise development of lung tumors showed decreased tumor progression and increased apoptotic gene expression following plasmid therapy. Further studies are warranted to assess the direct influence of TAMs on cancer cells post plasmid DNA encapsulated NP delivery to enable development of therapeutic strategies affecting cancer cells and cells of the tumor microenvironment.

Materials and Methods

Materials. Sodium hyaluronate (HA) with an average molecular weight of 20 kDa was obtained from Lifecore Biomedical (Chaska, MN). Poly(ethylene imine) (PEI molecular weight 20,000 Da) was obtained from Polysciences (Warrington, PA). Mono-functional poly(ethylene glycol)-amine (PEG_2K-NH₂, molecular weight 2,000 Da) was purchased from Creative PEG Works (Winston Salem, NC). Cisplatin was purchased from Fisher Scientific.

Cell culture experiments. SK-LU-1 human lung adenocarcinoma cells were obtained from ATCC (Manassas, VA). The cells were cultured in Dulbecco's Modified Eagle's medium supplemented with 10% fetal bovine serum and grown at 37 °C, 5% CO₂.

Plasmid DNA-encapsulated HA NP formulation and characterization. We have previously published information regarding preparing combinatorial designed HA formulations.^4,5,6 The HA-PEI and HA-PEG conjugates were prepared using this combinatorial approach. For the synthesis of HA-PEG-EGF, 50 mg of maleimide-PEG-amine was added to EDC/NHS-activated HA. Following synthesis of HA-PEG-maleimide, an EGFR-specific peptide, YHWYGYTPQNVI, designated as GE11 was used for conjugation with maleimide. The GE11 peptide was originally synthesized and screened as an EGFR-specific peptide by Li et al.³⁴ For this study, the GE11 peptide with a spacer sequence of GGGGC was synthesized at Tufts University Core Facility, Boston, MA. We have previously successfully prepared EGFR-targeted polymeric NPs using this peptide sequence.¹⁷ The carboxyl group of terminal cysteine of the peptide was reacted with the maleimide of maleimide-PEG-HA in HEPES buffer (pH 7.4) at 1:1 molar ratio while mixing under N2 at 4 °C for 24 hours. The peptide conjugate was then purified by dialysis (3.5 kDa), lyophilized, and characterized by NMR spectroscopy.

The HA-PEI and HA-PEG solutions (3 mg/ml) were prepared by dissolving the polymer in PBS. NP size and charge were determined on a Malvern Nano ZS (Malvern Instruments, UK). Transmission electron microscopy (JEM-1000; JEOL, Tokyo, Japan) was performed to assess the formation of plasmid-loaded NPs. Uranyl acetate ribonucleic acid stain was used to demarcate plasmid from the polymer. The ability of these complexes to release the plasmid was determined by treating them with poly(acrylic acid), followed by gel electrophoresis. Encapsulation efficiency of plasmid in the NP was assessed using the picogreen assay. The definitions of NPs are as follows: CD44TNP, CD44-targeted NPs; DualTNP, CD44 + EGFR-targeted NPs; CD44 125b TNP, CD44-targeted microRNA-125b; dual comboTNPs, delivery of CD44 + EGFR-targeted NPs encapsulating p53 and miR125b plasmids. Other plasmid-encapsulated NPs are also named similarly.

Quantitative assessment of plasmid encapsulated HA NP uptake.SK-LU-1 cells (0.2 million cells) were plated overnight in six-well plates and treated with 20 μg plasmid encapsulated HA NPs. Gene expression was assessed following 24 and 48 hours.

Apoptotic analysis of SK-LU-1 cells posttransfection with plasmid-encapsulated NPs. SK-LU-1 cells transfected with p53 and miR-125b plasmid were collected 48 hours posttransfection and pro- and antiapoptotic marker expression was assessed using qPCR. Messenger RNA was isolated, and complementary DNA (cDNA) was synthesized as previously described.³⁵

Repolarization of J774.A1 macrophages following co-culture with transfected SK-LU-1. SK-LU-1 cells were transfected with plasmid-encapsulated HA NPs (miR-125b and p53) as single or combination therapy. Forty-eight hours posttransfection, the transfected SK-LU-1 cells were co-cultured with IL-4–stimulated J774 cells. The co-cultured J774 cells were collected after 24 hours and M1 and M2 gene expression was assessed by qRT-PCR. RNA was isolated, and cDNA was synthesized as described previously.³⁵

In vivo evaluation of plasmid-encapsulated HA NPs in genetically modified lung cancer models. All animal handling and procedures were performed according to an approved protocol by Northeastern University's Institutional Animal Care and Use Committee. Cohorts of KP mice were infected with 2.5 × 10⁷ pfu of AdenoCre (University of Iowa) by intranasal inhalation 10 weeks after tumor initiation as described previously (Supplementary Figure S7).⁷ Mice (n = 8) were injected with PBS, cisplatin (5 mg/kg), and plasmid-encapsulated NPs (single therapy and in combination) (Figure 5). Following completion of the therapeutic regimen, the mice were euthanized by carbon dioxide asphyxiation. The lungs were inflated with PBS (for PCR studies), fixed in 4% formalin (histology studies), and stored in appropriate conditions until further analysis.

Tissue immunohistochemistry. Lung lobes were embedded in paraffin, sectioned at 4 μm, and stained with H&E for tumor pathology. Lung tumor sections were dewaxed, rehydrated, and stained using standard immunohistochemistry protocols.²² Anti-Ki67 antibody conjugated to Alexa Fluor 488 (1:100; Abcam, Boston, MA) was used for staining.

Qualitative transfection efficiency using western blot. Protein expression for flag-p53 was evaluated using western blot as previously described.²³ Briefly, proteins were extracted from tumors using Total Protein Extraction Kit (Millipore, Billerica, MA) and Powergen 125 tissue homogenizer (Fisher Scientific, Waltham, MA). Tissue lysate samples were analyzed for total protein concentration using the BCA assay (Pierce, Rockford, IL). Fifty micrograms of total protein extract was run on a precast 4–20% sodium dodecyl sulfate-polyacrylamide gel electrophoresis system at 200 V for 30 minutes. Subsequently, protein bands on the gel were transferred onto a polyvinylidene difluoride membrane by an iBlot Dry Blotting System (Invitrogen, Carlsbad, CA). The membrane was blocked with 5% milk in Tween-containing Tris-buffered saline (TBST) for 1 hour at room temperature. Membrane was cut and incubated with 1:1,000 dilution of primary rabbit β-actin antibody or 1:1,000 dilution of primary mouse monoclonal antiFLAGM2 antibody (Sigma-Aldrich, St Louis, MO) separately overnight at 4 °C. Membranes were then washed three times with TBST and incubated with 1:2,000 dilutions of secondary antirabbit or antimouse horseradish peroxidase–conjugated IgG (Cell Signaling Technology, Danvers, MA) in TBST for 1 hour at room temperature. After rinsing excess antibody with TBST and water, 4 ml ECL substrate (Pierce) was added and mixed with membranes for 5 minutes, which is cleaved by peroxidase to give a chemiluminescent product. The membranes were visualized using a Kodak Digital X-ray Specimen System. β-Actin was used as a protein-loading control. Quantification was performed using Image J software, and ratios were calculated respective to the β-actin concentrations.

Isolation of tumor-associated macrophages. The TAMs were isolated based on the protocol previously described by our group.³⁶ Briefly, the tumor tissue was cut into 3-mm fragments, followed by collagenase digestion (0.3 mg/ml, Worthington Biochemical, NJ) for 60 minutes at 37 °C in an incubator. The samples were then brought into a biosafety cabinet and placed into a 100-mm Petri dish containing 10 ml of the same media. The resultant suspension was then filtered via a 70-μm stainless steel wire mesh which resulted in almost a single-cell homogenous suspension. Three milliliter of ACK lysis buffer (Invitrogen) was added for 6 minutes to remove red blood cells, followed by addition of 10 ml of RPMI 1640 media. Following this, the suspension was further centrifuged, washed with PBS three times, and transferred to a 15-ml falcon tube with serum-free RPMI 1640 for 60 minutes. At this point, the cells collected from each group of animals were combined and placed in a T-175 tissue culture flask and incubated at 37 °C for 1 hour. The nonadherent cells were removed, and cells were washed twice with ice-cold 1× PBS. Using Accutase solution (Invitrogen, CA), the adherent cells were removed, and cells were pelleted by centrifugation. Following wash with 1× PBS, the pellet was used for magnetic bead–based separation using commercially available kit from Miltenyi Biotec system (Auburn, CA) as per manufacturer's protocol. Briefly, the cell pellet was re-suspended in 90 μl of proprietary Miltenyi buffer and stained with the 10 μl of CD11b-FITC-conjugated antibody (AbD Serotec, NC). Following dark condition incubation for 30 minutes at 4 °C in a refrigerator, the cells were washed with 1× PBS, and any unbound antibody was removed. Resultant cell pellet was re-suspended in 90 μl of buffer per one million cells followed by addition of 10 μl of anti-FITC–coated magnetic microbeads (Miltenyi Biotec). Cells were incubated for 20 minutes at 4 °C in a refrigerator under dark condition, washed, and supernatant was removed. The resultant cells were re-suspended in 500 μl of MACS buffer and separated using magnetic cell sorting (MACS) apparatus (Miltenyi Biotec) as per the manufacturer's instructions. After separation, the cells were used for mRNA isolation as described previously³⁷ using High Pure RNA Isolation Kit from Roche Applied (Indianapolis, IN). cDNA was synthesized using First Strand cDNA synthesis kit from Roche Applied as described previously. Quantitative data were represented in Supplementary Figure S8.³⁷

Statistical data analysis. All statistical analysis was performed using Prism 6.0 software (Graph Pad Software, San Diego, CA). Results were expressed as mean ± SEM. Data was analyzed using a one-way analysis of variance and/or Student's t-test, followed by Bonferroni's post hoc analysis for multiple comparisons. Differences were considered statistically significant at P < 0.05.

SUPPLEMENTARY MATERIAL Figure S1. Combination therapy—time Lag expression studies. Figure S2. qRT-PCR analysis of pro-apoptotic and anti-apoptotic genes SK-LU-1 post transfection with plasmid encoding p53 and/or 125b using lipofectamine transfection reagent. Figure S3. Re-polarization of IL-4 stimulated J774 cells post co-culture with p53 and 125b transfected SK-LU-1 cells. Figure S4. Analysis of inflammatory status in lung tissue post treatment with HA-encapsulated plasmids and cisplatin. Figure S5. Liver toxicity profile. Figure S6. Quantitative analysis of the Ki-67 staining. Figure S7. Histopathological analysis for induction of tumor after administration of Ad-Cre. Figure S8. Quantification of plasmid transfection in vivo in macrophages.