Improved therapeutic efficacy against murine carcinoma by combining honokiol with gene therapy of PNAS‐4, a novel pro‐apoptotic gene

Abstract

PNAS‐4, a novel pro‐apoptotic gene activated during the early response to DNA damage, can inhibit proliferation via apoptosis when overexpressed in some tumor cells. Recent studies have indicated that honokiol can induce apoptosis, inhibit angiogenesis, and suppress tumor growth. In the present study, we investigated whether mouse PNAS‐4 (mPNAS‐4) could augment the apoptosis of tumor cells induced by honokiol in vitro, and whether the antiangiogenic activity of honokiol and induction of apoptosis by mPNAS‐4 could work cooperatively to improve the antitumor efficacy in vivo. In vitro, mPNAS‐4 inhibited proliferation of murine colorectal carcinoma CT26 and Lewis lung carcinoma LL2 cells through induction of apoptosis, and significantly augmented the apoptosis of CT26 and LL2 cells induced by honokiol. Compared with treatment with mPNAS‐4 or honokiol alone, in vivo systemic administration of an expression plasmid encoding mPNAS‐4 and low‐dose honokiol significantly suppressed tumor growth through the enhanced induction of apoptosis and the augmented inhibition of angiogenesis. Our data suggest that the combined treatment with mPNAS‐4 plus honokiol augments antitumor effects in vitro and in vivo, and that the improved antitumor activity in vivo may be associated with enhanced induction of apoptosis and augmented inhibition of angiogenesis. The present study may provide a novel and effective method for the treatment of cancer. (Cancer Sci 2009; 100: 1757–1766)

The traditional trinity of cancer therapies is comprised of chemotherapy as the first‐line treatment for a wide variety of tumors, surgical intervention, and radiation. Although chemotherapy is effective in combating the growth and spread of some tumors, dose‐dependent toxicities such as bone marrow suppression, gastrointestinal reactions, liver dysfunction, and renal toxicity, and development of drug resistance greatly limit its therapeutic efficacy.^{( 1 )} To date, none of the treatment regimens using conventional chemotherapy drugs such as cisplatin, adriamycin, and 5‐fluorouracil is curative. Thus, continued preclinical studies on innovative therapeutic strategies are warranted.

Natural products have been the source of many medically beneficial drugs, and their importance in the prevention and treatment of cancer is becoming increasingly apparent.^{( 2 )} Honokiol, a naturally occurring compound isolated from Magnolia officinalis, is of medicinal use. Honokiol shows some biological activities, including antioxidative activity,^{( 3 )} anxiolytic potency,^{( 4 )} antiplatelet aggregation,^{( 5 )} hepatoprotective effect,^{( 6 )} neuroprotective effect,^{( 7 )} and anti‐inflammatory and antimicrobial effects.^{( 8} , ⁹ , ^{10 )} Previous reports have shown the anticancer activities of honokiol. Honokiol induces apoptosis in various cancer cell lines^{( 11} , ¹² , ¹³ , ^{14 )} and represses tumor growth in nude mice.^{( 15} , ^{16 )} Moreover, honokiol inhibits angiogenesis by interfering with the phosphorylation of VEGFR‐2.^{( 15 )} Recently, we found that liposomal honokiol significantly inhibits tumor‐associated lymphangiogenesis and metastasis.^{( 17 )} More importantly, honokiol can overcome drug resistance in various cancer cell lines.^{( 18} , ¹⁹ , ²⁰ , ^{21 )} These findings indicate that honokiol may be a novel, promising anticancer agent.

PNAS‐4 was identified as a novel apoptosis‐related protein in the human acute promyelocytic leukemia cell line NB4. The PNAS‐4 protein contains a conserved N‐terminal portion belonging to the DUF862 family with unknown function and two or three hydrophobic motifs. The amino acid sequence identities of PNAS‐4 among various organisms are very high. Recent studies showed that PNAS‐4 is upregulated in peripheral blood mononuclear cells that are exposed to carcinogenic agents such as benzene,^{( 22 )} in human papillomavirus 16 E6‐expressing U2OS cells (U2OSE64b) following mitomycin C treatment,^{( 23 )} in human papillomavirus‐infected invasive cervical cancer,^{( 24 )} and in androgen‐independent prostate cancer,^{( 25 )} indicating that it might be related to the genesis of some cancers. hPNAS‐4 was identified as a novel pro‐apoptotic gene activated during the early response to DNA damage, and when overexpressed in osteosarcoma U2OS cells, it could induce significant apoptosis.^{( 26 )} In addition, a recent report showed that the expression of porcine pPNAS‐4 tended to decrease gradually during the period for development of muscle fibers, that is, the muscle fiber number appeared to increase as a result of the decrease in pPNAS‐4 mRNA expression level during prenatal muscle development, implying that PNAS‐4 may be involved in apoptosis.^{( 27 )}

Some original investigations on the functions of PNAS‐4 have been carried out in our laboratory. By using zebrafish as the animal model, we observed that overexpression of PNAS‐4 leads to elongation of the antero‐posterior body axis, and knocking down PNAS‐4 causes gastrulation defects, indicating that PNAS‐4 may regulate convergence and extension during vertebrate gastrulation.^{( 28 )} By searching mouse Reference mRNA sequences (refseq‐rna) database from NCBI with the full‐length amino acid sequence of hPNAS‐4, we found a mouse PNAS‐4 homolog and named it mPNAS‐4 (GenBank accession no. NM_024282). The mPNAS‐4 cDNA encodes a protein possessing a putative DUF862 domain and three hydrophobic motifs, with a calculated molecular weight of 21 434 Da. In addition, we found that overexpression of PNAS‐4 efficiently inhibited growth of LL2 cells, human ovarian cancer SKOV3 cells, and lung adenocarcinoma A549 cells through induction of apoptosis.^{( 29} , ³⁰ , ^{31 )}

More recently, we have confirmed that PNAS‐4 enhances the sensitivity of lung cancer to gemcitabine.^{( 32 )} Similarly, we decided to investigate whether PNAS‐4 could also potentiate the antineoplastic effects of honokiol. This decision was based on the following considerations:

• 1Honokiol has strong anticancer activities including pro‐apoptotic activity, antiangiogenic activity, anti‐invasive activity, antimetastatic activity, and antiproliferative activity.^{( 11} , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ^{17 )}
• 2More importantly, honokiol shows superiority over conventional chemotherapy drugs, that is, it can overcome drug resistance in some cancer cell lines.^{( 18} , ¹⁹ , ²⁰ , ^{21 )}
• 3We have carried out systematic studies on honokiol, and applied for a Chinese patent (patent no. 200610021277.8, The preparation of liposomal honokiol and its application in the treatment of cancer).

On the basis of these considerations and the observations that PNAS‐4 induces apoptosis when overexpressed in some tumor cells, we subsequently investigated the antitumor efficacy of mPNAS‐4 plus low‐dose honokiol in vitro and in vivo, using CT26 and LL2 murine models. The present study demonstrated that in vitro mPNAS‐4 induces apoptosis and also enhances the apoptosis of cancer cells induced by honokiol, and that the antiangiogenic activity of honokiol and the induction of apoptosis by mPNAS‐4 could work cooperatively to enhance efficacy in vivo. Taken together, our results may provide a novel and effective approach to treat cancer.

Materials and Methods

Preparation of liposome.  The liposome for honokiol was prepared in our laboratory as described previously,^{( 21 )} and the liposome for plasmid treatment in animal experiments was a cationic liposome (DOTAP/cholesterol). The cationic liposome was prepared using the procedure described previously.^{( 33 )} Briefly, the cationic lipid DOTAP (Avanti Polar Lipids, Alabaster, AL, USA) and the neutral lipid cholesterol (Sigma, St Louis, MO, USA) were mixed at equimolar concentrations and dissolved in chloroform in a round‐bottomed flask. The solution was rotated on a rotary evaporator at 30°C for 30 min to obtain a thin lipid film. The dried lipid film was then hydrated in 5% glucose solution to give a final concentration of 7 mM DOTAP and 7 mM cholesterol, referred to as 7 mM DOTAP:Chol. The hydrated DOTAP:Chol film was rotated in a water bath at 50°C for 45 min and then 35°C for 10 min. Then, the mixture was left overnight and sonicated at low frequency for 5 min at 50°C. After sonication, it was transferred to a tube and heated at 50°C for 10 min. The mixture was sequentially extruded through polycarbonate membrane (Millipore, Billerica, MA, USA) to decrease size (five times at 0.2 µm and three times at 0.1 µm) using syringes. The final cationic liposome (DOTAP/cholesterol) was a small multilamellar liposome in a size range of 100 ± 20 nm. It was stored under argon gas at 4°C.

Preparation of liposomal honokiol.  Separation and purification of honokiol were carried out according to methods described previously.^{( 34 )} In order to make it soluble, the purified honokiol was encapsulated with liposome according to the method reported previously.^{( 21 )}

Plasmid construction and purification.  The pcDNA3.1‐mPS plasmid was constructed as previously described.^{( 31 )} pcDNA3.1 vector was used as a control. Both pcDNA3.1‐mPS and pcDNA3.1 were purified by two rounds of passage over EndoFree columns (Qiagen, Chatsworth, CA, USA), as reported previously.^{( 35} , ^{36 )} The purified plasmids were mixed with liposome to form a DNA–liposome complex, and then used for subsequent animal experiments.

Cell culture and transfection.  LL2 and CT26 cells were obtained from American Type Culture Collection (Manassas, VA, USA) and cultured in DMEM (Gibco, Carlsbad, CA, USA) and RPMI‐1640 (Gibco), respectively, containing 10% FBS in a 37°C incubator with a humidified 5% CO₂ atmosphere. For transfection, each well of six‐well or 96‐well plates was seeded with 2 × 10⁵ or 2 × 10³ cells, respectively. Transfection was carried out using Lipofectamine 2000 (Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions.

Treatments of cells in the in vitro experiments.  LL2 and CT26 cells were classified into the following five groups, and treated as follows:

• • Control: the cells were left untreated, and when cultured for 72 h, cells were harvested.
• • pcDNA3.1 (empty vector) alone: the cells were first incubated for 24 h, then transfected with pcDNA3.1 plasmid. Forty‐eight hours after transfection, cells were harvested.
• • Honokiol alone: when the cells were cultured for 48 h, honokiol was added at a concentration of 10 µg/mL. Twenty‐four hours later, cells were harvested.
• • mPNAS‐4 alone: the cells were first incubated for 24 h, then transfected with pcDNA3.1‐mPS plasmid. Forty‐eight hours after transfection, cells were harvested.
• • mPNAS‐4 plus honokiol (combination): the cells were first incubated for 24 h, then transfected with pcDNA3.1‐mPS plasmid. Twenty‐four hours after transfection, honokiol was added at a concentration of 10 µg/mL. Forty‐eight hours after transfection, cells were harvested.

The harvested cells above were used for MTT assay, flow cytometric analysis, agarose gel DNA electrophoresis, and morphological analysis. In the morphological analysis, different cell groups were treated as described above. However, in order to identify transfected cells from non‐transfected cells upon green fluorescence, the cells were cotransfected with pcDNA3.1 or pcDNA3.1‐mPS plus pEGFP‐N1 plasmids (Clontech Laboratories, Mountain View, CA, USA). The concentration of pEGFP‐N1 vector was one‐fifth that of the pcDNA3.1 or pcDNA3.1‐mPS vector.

Detection of mPNAS‐4 expression in vitro and in vivo.  For detection of mPNAS‐4 expression in vitro, LL2 cells were treated according to the schedules described above. One part of the harvested cells was used to isolate RNA and then subjected to RT‐PCR for amplification of the coding region of mPNAS‐4, and the rest was used for the detection of protein expression by western blot analysis according to the method described previously,^{( 37 )} using the anti‐mPNAS‐4 polyclonal antibodies produced by us.^{( 29 )} For detection of mPNAS‐4 expression in vivo, when mice were killed at the end of the experiment, the tumor tissues were collected for detecting mPNAS‐4 expression by RT‐PCR. The primers used for amplification of mPNAS‐4 (608 bp) were as follows: 5′‐GCGGATCCGCCACC ATGGCCAACCAGCCC ATCATC‐3′ and 5′‐CCGCTCGAGCTATAGTTTTGTGTGGCGC CCAGG‐3′, whereas the primers used for amplification of GAPDH (187 bp) were designed as reported previously.^{( 38 )}

MTT assay.  Survival of cells after treatments was quantified using the MTT assay.^{( 39 )} Cells were seeded in triplicate for each group and the processes were repeated three times. Media‐only treated cells served as the indicator of 100% cell viability.

Flow cytometric analysis.  Flow cytometric analysis was carried out to identify sub‐G₁ cells and apoptotic cells, and to measure the percentage of sub‐G₁ cells after propidium iodide staining in hypotonic buffer as previously described.^{( 40 )}

Agarose gel DNA electrophoresis.  The pattern of DNA cleavage was analyzed by agarose gel electrophoresis as described previously.^{( 40 )}

Morphological analysis.  LL2 cells were treated according to the schedules above. Apoptotic morphology was analyzed according to previously reported methods with a slight modification.^{( 41} , ^{42 )} Briefly, the cells were cotransfected as described above and cultured for the indicated period of time, then used for apoptotic morphology analysis. The apoptotic morphology of tumor cells was determined at 48 h after transfection and the percentage of apoptotic cells was determined by the number of green cells with apoptotic morphology divided by the total number of green cells. The images of enhanced green fluorescent protein (EGFP) expression were captured by a fluorescence microscope.

Animal tumor models and treatment.  Female mice at 6–8 weeks of age were transplanted with 5 × 10⁵ tumor cells. The LL2 and CT26 models were established in C57BL/6 and BALB/c mice, respectively. Ten days after tumor cell inoculation, when tumors were 5–8 mm in diameter, the mice were assigned randomly to one of the following five groups (10 mice/group), and treated with: (i) 100 µL PBS; (ii) 100 µg pcDNA3.1 plasmid and 300 µg liposome complexes in 100 µL PBS; (iii) 100 µL of 0.5 mg liposomal honokiol (25 mg/kg bodyweight); (iv) 100 µg pcDNA3.1‐mPS plasmid and 300 µg liposome complexes in 100 µL PBS; and (v) 100 µg pcDNA3.1‐mPS plasmid and 300 µg liposome complexes in 100 µL PBS and 100 µL of 0.3 mg liposomal honokiol (15 mg/kg bodyweight). The mice were treated with DNA–liposome complex by intravenous administration via the tail vein twice a week, and received honokiol by intraperitoneal route every day for 4 weeks. Tumor size was measured with a dial caliper at 3‐day intervals, and tumor volume was calculated by the following formula: tumor volume (mm³) = 0.52 × length (mm) × width (mm) × width (mm).^{( 43 )} All studies involving mice were approved by the Institutional Animal Care and Treatment Committee of Sichuan University.

Alginate‐encapsulated tumor cell assay.  Alginate‐encapsulated tumor cell assays were carried out as described previously with a slight modification.^{( 44 )} Briefly, CT26 cells were resuspended in sodium alginate solution (w/v, 1.8%) and added dropwise to 250 mM calcium chloride solution. One of the formed alginate beads contained approximately 1 × 10⁵ tumor cells. BALB/c mice were then anesthetized, and four beads were implanted subcutaneously into an incision made on the dorsal side. Mice were subdivided into five groups (five mice per group) and treated as described above. The treatments were started 24 h after the beads were implanted. Two weeks later, the mice were injected i.v. with 100 µL of 100 mg/kg FITC–dextran (Sigma) solution through the tail vein. Beads were removed and photographed 20 min after FITC–dextran injection. The FITC–dextran uptake was measured against a standard curve of FITC–dextran.

Histological analysis.  Dissected tumors were divided in half, one‐half for paraffin sections fixed in 10% NBF and embedded in paraffin, and the other half frozen at –80°C. For MVD analysis, frozen sections (7 µm) were fixed in acetone, incubated, and stained with an antibody reactive to CD31 as reported previously.^{( 45 )} MVD was determined by counting the number of microvessels per high‐power field as described previously.^{( 46} , ^{47 )} For quantitative assessment of apoptosis, TUNEL was carried out with an In situ Cell Death Detection Kit (Hoffmann‐La Roche, Basel, Switzerland) following the manufacturer's protocol.^{( 48 )} For observations of potential side effects, the tissues (including heart, liver, spleen, lung, kidney, and brain) were fixed in 10% NBF and embedded in paraffin. Sections (3–5 µm) were stained with H&E.^{( 49 )} Sections in H&E staining and immunohistochemical staining were observed by two pathologists in a blinded manner.

Statistical analysis.  The statistical analysis was carried out using SPSS software (version 12.0 for Windows, SPSS UK Ltd., Woking, Surrey, UK). All of the values were expressed as means ± SD. Data were analyzed by one‐way ANOVA, and then differences among the means were analyzed using Tukey–Kramer multiple comparison test. Survival curves were constructed according to the Kaplan–Meier method,^{( 50 )} and statistical significance was determined by the log‐rank test.^{( 51 )} Differences were considered significant at P < 0.05.

Results

Overexpression of mPNAS‐4 in vitro and in vivo. In vitro overexpression of recombinant mPNAS‐4 in transfected LL2 cells was confirmed by RT‐PCR (Fig. 1a), and the production of mPNAS‐4 protein was further confirmed by western blotting analysis (Fig. 1b). To exactly investigate whether intravenous injections of pcDNA3.1‐mPS expression plasmid led to an apparently improved expression level of exogenous mPNAS‐4 within the tumor tissues, we decided to detect it by RT‐PCR analysis. As expected, in vivo overexpression of recombinant mPNAS‐4 was further confirmed by RT‐PCR (Fig. 1c).

Figure 1.

In vitro and in vivo expression of recombinant mouse PNAS‐4 (mPNAS‐4). (a) RT‐PCR analysis of mPNAS‐4 expression. Results from LL2 cells untreated, from cells transfected with pcDNA3.1 plasmid, from cells treated with honokiol, from cells transfected with pcDNA3.1‐mPS, and from cells treated with pcDNA3.1‐mPS plus honokiol are shown in lanes 1–5, respectively. (b) Western blotting analysis of mPNAS‐4 expression in vitro after transfection of LL2 cells. Total protein samples were obtained from LL2 cells untreated (lane 1), from cells transfected with pcDNA3.1 plasmid (lane 2), from cells treated with honokiol (lane 3), from cells transfected with pcDNA3.1‐mPS (lane 4), and from cells treated with pcDNA3.1‐mPS plus honokiol (lane 5). Protein mass molecular markers are indicated on the left. (c) RT‐PCR analysis of expression of exogenous mPNAS‐4 in vivo. Total RNA samples were isolated from tumor tissues in the untreated group (lane 1), pcDNA3.1 treatment group (lane 2), honokiol treatment group (lane 3), pcDNA3.1‐mPS treatment group (lane 4), and pcDNA3.1‐mPS plus honokiol treatment group (lane 5), respectively.

Inhibition of cell proliferation in vitro by mPNAS‐4 and honokiol.  Honokiol, the structure of which is shown in Figure 2(a), can inhibit the growth of tumor cells in a dose‐ and time‐dependent manner.^{( 12} , ¹³ , ^{14 )} We first treated LL2 and CT26 cells with honokiol at different concentrations, with a 24‐h interval, and found that the IC₅₀ of honokiol to CT26 and LL2 is approximately 12 and 16 µg/mL, respectively (Fig. 2b). To assess whether mPNAS‐4 could augment the antiproliferative effect of honokiol, we decided to treat cells with honokiol at a suboptimal dose (10 µg/mL) with various schedules. After treatment, the viability of cells was determined by MTT assay. As shown in Figure 3(a), compared with the control group, either honokiol or mPNAS‐4 significantly reduced CT26 cell viability by 53% (P < 0.001) and 62% (P < 0.001), respectively. Treatment with mPNAS‐4 plus honokiol very significantly reduced CT26 cell viability by 81% (P < 0.001). When treated with mPNAS‐4 in combination with honokiol, the percentage inhibition of LL2 cells was 74% (P < 0.001) whereas the percentage inhibition of LL2 cells treated with mPNAS‐4 or honokiol alone was less than 50% (both P < 0.01), compared with the control group (Fig. 3b).

Figure 2.

Effect of honokiol on survival of LL2 and CT26 tumor cells. (a) Structure of honokiol. (b) Effect of honokiol at different concentrations on survival of LL2 and CT26 tumor cells.

Figure 3.

Inhibition of carcinoma cell proliferation in vitro by treatment with mouse PNAS‐4 (mPNAS‐4) plus honokiol. The MTT assay was carried out as described in Materials and Methods. (a) The treatment of mPNAS‐4 plus honokiol reduced CT26 cell viability more significantly than the treatment of mPNAS‐4 alone or honokiol alone did. (b) The treatment of mPNAS‐4 plus honokiol reduced LL2 cell viability more significantly than the treatment of mPNAS‐4 alone or honokiol alone did. Statistically significant differences compared with the two single‐treatment groups (^# P < 0.05; ^## P < 0.01). Significant differences compared with the control group (**P < 0.01; ***P < 0.001). Results are shown as means ± SD of six wells and triplicate experiments. In each experiment, the medium‐only treatment (untreated) indicates 100% cell viability.

Induction of apoptosis of tumor cells in vitro by mPNAS‐4 and honokiol.  The quantitative assessment of sub‐G₁ cells by flow cytometry was used to estimate the number of apoptotic cells. As shown in Figure 4(a), the combined treatment of LL2 cells with mPNAS‐4 plus honokiol increased number of sub‐G₁ cells compared with the two single‐treatment groups or the two control groups. Furthermore, agarose gel electrophoresis of mPNAS‐4 plus honokiol demonstrated a more apparent ladder‐like pattern of DNA fragments compared with the two single‐treatment groups (Fig. 4b). To further assess apoptosis by morphological changes, LL2 cells were treated. As shown in Figure 4(c), the morphological changes of cells at 48 h post‐transfection monitored by fluorescence microscopy were characteristic of apoptosis (rounded or floating). By contrast, the untreated cells showed a green normal shape. The percentage of apoptotic cells was further determined by the number of green cells with apoptotic morphology divided by the total number of green cells. As shown in Figure 4(d), either honokiol or mPNAS‐4 resulted in a significant increase in the percentage of apoptotic cells compared with that in the PBS or empty vector groups (P < 0.01; P < 0.001); the combination of mPNAS‐4 and honokiol further increased the percentage of apoptotic cells compared with that in the PBS or empty vector groups (P < 0.001), and significantly increased the percentage of apoptotic cells compared to the mPNAS‐4 or honokiol groups (P < 0.01). No significant differences were found between the two single‐treatment groups (P = 0.947) or the two control groups (P = 0.714). As expected, results obtained with apoptotic cell counting in the morphological analysis strongly correlated with those obtained with flow cytometric analysis.

Figure 4.

Induction of apoptosis of tumor cells in vitro by mouse PNAS‐4 (mPNAS‐4) and honokiol. (a) Representative DNA fluorescence histograms of propidium iodide‐stained cells. LL2 cells were treated with mPNAS‐4 for 24 h, then with 10 µg/mL honokiol for an additional 24 h. LL2 cells were (1) untreated, (2) treated with empty vector or (3) honokiol, (4) mPNAS‐4 alone or (5) mPNAS‐4 plus honokiol and groups 1, 2, 3, 4, and 5 correspond with these five treatments (the same as shown in the subsequent panels), with (1) 15.9%, (2) 23%, (3) 47.3%, (4) 40.6%, and (5) 58.6% sub‐G₁ cells (apoptotic cells), respectively, as assessed by flow cytometry. (b) Agarose gel electrophoretic patterns of DNA. LL2 cells were treated with the same conditions as described above. Lane M, DNA marker; lane 1, untreated LL2; lane 2, treated with empty vector; lane 3, treated with honokiol alone; lane 4, treated with mPNAS‐4 alone; lane 5, treated with mPNAS‐4 plus honokiol. (c) Morphology of green normal cells and apoptotic cells (rounded or floating). LL2 cells were treated with the same conditions as mentioned above. Arrow indicates an example of apoptotic cells. (d) The percentage of apoptotic cells. Statistically significant difference in the percentage of apoptotic cells treated with honokiol, or mPNAS‐4 versus PBS and pcDNA3.1 controls (**P < 0.01; ***P < 0.001); significant difference in the percentage of apoptotic cells treated with mPNAS‐4 + honokiol versus PBS and pcDNA3.1 controls (***P < 0.001); and significant difference for the combination treatment versus mPNAS‐4 or honokiol single treatment (^## P < 0.01). The percentage was determined by the number of green cells with apoptotic morphology divided by the total number of green cells (average of three individual experiments).

Antitumor efficacy of the combination of mPNAS‐4 and low‐dose honokiol.  We further examined whether PNAS‐4 could improve the antineoplastic effect of honokiol on CT26 and LL2 tumors in vivo. As shown in Figure 5(a), on day 34 after implantation, the CT26 and LL2 tumors of mice treated with PBS reached 3572.22 ± 319.34 and 4919.94 ± 660.76 mm³ in volume, respectively. The CT26 and LL2 tumors treated with honokiol were significantly (P < 0.001) smaller than those treated with PBS, reaching only 1768.71 ± 245.67 and 2178.66 ± 229.44 mm³ in volume, respectively. mPNAS‐4 also resulted in a significant reduction in tumor volume (772.7 ± 161.14 mm³ in the CT26 model, and 1212.27 ± 188.11 mm³ in the LL2 model) compared with control tumors (P < 0.001). The combination of mPNAS‐4 and honokiol further suppressed tumor growth such that the CT26 and LL2 tumors reached 412.86 ± 46.94 and 709.97 ± 65.78 mm³ in volume, respectively, which was significantly (P < 0.001) smaller than control tumors, and significantly smaller than the tumors treated with mPNAS‐4 or honokiol (P < 0.05). Significant differences in tumor volume (2899.9 ± 459.13 mm³ in the CT26 model and 3424.32 ± 663.76 mm³ in the LL2 model) were also observed in the pcDNA3.1‐treated group compared with the control group (P < 0.01).

Figure 5.

Combined effect of mouse PNAS‐4 (mPNAS‐4) and honokiol on two murine tumor models. Female mice at 6–8 weeks of age were transplanted subcutaneously with 5 × 10⁵ CT26 or 5 × 10⁵ LL2 cells. Ten days after tumor cells were transplanted, the mice were assigned randomly to five groups and treated with PBS, pcDNA3.1, honokiol, pcDNA3.1‐mPS, or pcDNA3.1‐mPS + honokiol. (a) Suppression of tumor growth in mice. The sizes (mm³) of tumors were monitored and recorded. Significant differences for tumors treated with honokiol or mPNAS‐4 versus PBS and pcDNA3.1 controls (***P < 0.001); significant difference for tumors treated with mPNAS‐4 + honokiol versus PBS and pcDNA3.1 controls (***P < 0.001); and significant difference for the combination therapy versus mPNAS‐4 or honokiol monotherapy (^# P < 0.05). (b) Survival curve of mice per treatment group. Statistically significant differences compared with PBS and pcDNA3.1 controls (*P < 0.05; **P < 0.01). Significant differences compared with the two single‐treatment groups (^# P < 0.05).

Survival curve analysis (Fig. 5b) showed that CT26 tumor‐bearing mice in the PBS or pcDNA3.1‐treated groups survived less than 43 days on average. By contrast, either mPNAS‐4 or honokiol resulted in a significant (P < 0.05) increase in life span compared with the two control groups, with the mean survival time being 52 days. The combination of mPNAS‐4 and honokiol further improved survival to a greater extent than the two control groups (P < 0.01), with the mean survival time being 57 days. Similar results were also found in the LL2 tumor model (Fig. 5b).

In addition, the mice treated with mPNAS‐4 alone, honokiol alone, or in combination have been, in particular, investigated for potential side effects. No adverse consequences were indicated in gross measures such as weight loss, ruffling of fur, life span, behavior, and feeding. Furthermore, no pathological changes in heart, liver, spleen, lung, or kidney were found by microscopic examination (data not shown).

Inhibition of tumor angiogenesis.  Angiogenesis in tumor tissues was estimated by MVD assay and quantified. As shown in Figure 6, either honokiol or mPNAS‐4 resulted in a significant reduction in tumor MVD compared with the MVD in the PBS or empty vector groups (P < 0.01). Furthermore, the combination therapy with mPNAS‐4 and honokiol more significantly reduced MVD compared with treatment with PBS or empty vector (P < 0.001), and significantly reduced MVD compared to the mPNAS‐4 or honokiol monotherapy groups (P < 0.01). Evidence for inhibition of angiogenesis was also found in the alginate‐encapsulated tumor cell assay (Fig. 7). Figure 7 shows representative images of tumor vasculature in CT26‐encapsuled alginate. The vascular plexus of the tumors in the PBS and pcDNA3.1 groups formed a ladder‐like pattern that was richly morphological and irregular (Fig. 7a,b). Whereas tumors treated with honokiol or pcDNA3.1‐mPS alone exhibited relatively little vascularity (Fig. 7c,d), the combination therapy group displayed the least amount of vascularization (Fig. 7e). Angiogenesis of implanted alginate beads was quantified by measuring FITC–dextran uptake. As expected, both honokiol and mPNAS‐4 resulted in a significant reduction in FITC–dextran uptake compared with that in the PBS or empty vector groups (P < 0.01). The combination therapy with mPNAS‐4 and honokiol further reduced FITC–dextran uptake compared with treatment with PBS or empty vector (P < 0.01), and significantly reduced FITC–dextran uptake compared to the mPNAS‐4 or honokiol monotherapy groups (P < 0.01). No significant difference was found between the two single‐treatment groups (P = 0.307).

Figure 6.

Inhibition of angiogenesis assayed by immunohistochemistry with CD31. Frozen CT26 tumor sections from the mice treated with (a) PBS, (b) pcDNA3.1, (c) honokiol, (d) mPNAS‐4, and (e) mPNAS‐4 + honokiol are shown. Tumor angiogenesis was assessed by immunohistochemical staining with anti‐CD31 antibody (brown). Microvessels in CT26 (open columns) and LL2 (solid columns) tumor sections were counted in blindly chosen random fields to record microvessel density (MVD). Significant difference in MVD for tumors treated with honokiol or mPNAS‐4 versus PBS and pcDNA3.1 controls (**P < 0.01); significant difference for tumors treated with mPNAS‐4 + honokiol versus PBS and pcDNA3.1 controls (***P < 0.001); and significant difference for the combination therapy versus mPNAS‐4 or honokiol monotherapy (^## P < 0.01).

Figure 7.

Antiangiogenesis assay by alginate bead in vivo. Alginate beads containing 1 × 10⁵ tumor cells were subcutaneously implanted into the backs of BALB/c mice (four beads per mouse). The treatments of mice (five mice/group) and the quantification of FITC–dextran uptake were described in Materials and Methods. Representative images of alginate beads after treatment with (a) PBS, (b) pcDNA3.1, (c) honokiol, (d) mouse PNAS‐4 (mPNAS‐4), and (e) mPNAS‐4 + honokiol were shown under a dissecting microscope (×10). (f) FITC–dextran of alginate beads was quantified. Statistically significant difference in the FITC–dextran uptake from the mice treated with honokiol or mPNAS‐4 versus PBS and pcDNA3.1 controls (**P < 0.01); significant difference in the FITC–dextran uptake from the mice treated with mPNAS‐4 + honokiol versus the two controls (**P < 0.01); and significant difference for the combination therapy versus mPNAS‐4 or honokiol monotherapy (^## P < 0.01).

Induction of apoptosis in tumor tissues.  TUNEL assays were carried out to detect apoptosis of tumor cells. As shown in Figure 8, both honokiol and mPNAS‐4 increased the apoptotic rate of tumor cells when compared with PBS or empty vector. The combination therapy further increased the density of apoptotic cancer cells. The apoptotic index of tumors treated with mPNAS‐4 or honokiol was significantly higher than that of tumors treated with PBS or pcDNA3.1 empty vector (P < 0.01). The apoptotic index for tumors treated with the combination of mPNAS‐4 and honokiol was significantly higher than for tumors treated with PBS or pcDNA3.1 control vector (P < 0.01), and significantly higher than the apoptotic index of tumors treated with the mPNAS‐4 and honokiol monotherapies (P < 0.01). No significant differences in the apoptotic index of tumors treated with the two monotherapies were found in the CT26 (P = 1.000) and LL2 models (P = 1.000). In addition, it was notable that no significant differences were found between the two control groups of CT26 (P = 0.166) and LL2 (P = 0.972), which suggested that the vector should be only slightly toxic to cells.

Figure 8.

TUNEL staining of tumor tissues. Tumor tissue preparation and procedure for TUNEL staining were described in Materials and Methods. Representative CT26 tumor sections from mice receiving (a) PBS, (b) pcDNA3.1, (c) honokiol, (d) mouse PNAS‐4 (mPNAS‐4), and (e) mPNAS‐4 + honokiol were prepared. (f) Apoptotic index within tissue from CT26 (open columns) and LL2 (solid columns) from 10 mice were counted. Statistically significant difference in the apoptotic index for tumors treated with honokiol or mPNAS‐4 versus PBS and pcDNA3.1 controls (**P < 0.01); significant difference for tumors treated with mPNAS‐4 + honokiol versus the two controls (**P < 0.01); and significant difference for the combination therapy versus mPNAS‐4 or honokiol monotherapy (^## P < 0.01). The apoptotic index was calculated as a ratio of the apoptotic cell number to the total cell number in each field.

Discussion

Although conventional chemotherapy drugs such as cisplatin, 5‐fluorouracil, and adriamycin have proved to be effective against some tumors, dose‐dependent toxicity and the development of drug resistance greatly limit their therapeutic efficacy. To date, none of the conventional chemotherapy drugs is curative. Therefore, more efficient molecules and more therapeutic strategies for combating cancer are required. Antiangiogenic therapy attempts to stop new vessels from forming around a tumor and break up the existing network of abnormal capillaries that feeds the cancerous mass, and has proved to be effective in inhibiting tumor growth.^{( 52 )} However, antiangiogenic therapy has not been shown to be tumoricidal in most studies.^{( 53 )} This therapeutic limitation can be overcome by combining angiogenesis inhibitors with cytotoxic therapies.^{( 53} , ⁵⁴ , ⁵⁵ , ^{56 )} Honokiol, a promising new anticancer compound, shows strong anticancer activities such as pro‐apoptotic activity, antiangiogenic activity, and antimetastatic activity.^{( 11} , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ^{18 )}

PNAS‐4 was previously identified as a putative apoptosis‐related protein in the human acute promyelocytic leukemia cell line NB4. Recent evidence has shown that PNAS‐4 may be involved in the genesis of some cancers.^{( 23} , ²⁴ , ^{25 )} hPNAS‐4 was recently identified as a novel pro‐apoptotic gene activated during the early response to DNA damage,^{( 26 )} and it could induce significant apoptosis when overexpressed in some tumor cells such as U2OS, SKOV3, and A549.^{( 26} , ³⁰ , ^{31 )} mPNAS‐4 was expressed in some normal tissues such as liver and heart, and extensively expressed in a variety of tumor cells such as LL2, CT26, MethA fibrosarcoma, and B16 melanoma cells. However, the expression levels of mPNAS‐4 in different tumor cells were different. For instance, the expression levels of mPNAS‐4 in MethA and B16 cells were higher than those in LL2 and CT26 (data not shown).

Based on the antiangiogenic, pro‐apoptotic, and antimetastatic activities of honokiol, and the observations that PNAS‐4 could induce significant apoptosis when overexpressed in some tumor cells, the present study was designed to investigate whether mPNAS‐4 could augment the apoptosis of tumor cells induced by honokiol in vitro, and whether the antiangiogenic activity of honokiol and induction of apoptosis by mPNAS‐4 could work cooperatively to enhance antitumor efficacy in vivo.

Several observations have been made in this work concerning induction of apoptosis and inhibition of angiogenesis. The in vitro experiments showed that mPNAS‐4 or honokiol alone inhibited proliferation and induced apoptosis, and that mPNAS‐4 plus honokiol augmented the induction of apoptosis in the experimental tumor cells, as evidenced by MTT assay (Fig. 3), flow cytometric analysis (Fig. 4a), DNA fragmentation (Fig. 4b), and apoptotic morphology analysis (Fig. 4c,d). Apoptosis has been described as multiple pathways converging from numerous different initiating events and insults.^{( 57 )} Although the molecular mechanism of PNAS‐4 for induction of apoptosis in cancer cells is not fully clear, our preliminary results suggest that PNAS‐4 inhibits tumor cell proliferation through the following mechanisms: (i) overexpression of PNAS‐4 causes S phase arrest by regulating the expression of cell cycle‐related proteins such as p21, E2F1, and proliferating cell nuclear antigen; and (ii) PNAS‐4 induces apoptosis through the mitochondrial apoptosis pathway, as evidenced by downregulation of Bcl‐2, upregulation of Bax, release of mitochondrial cytochrome c to the cytosol, and cleavage of caspase‐3 in PNAS‐4‐overexpressing A549 human lung adenocarcinoma cells.^{( 31 )}

Previous studies have shown that honokiol induces apoptosis and inhibits the growth of some tumor cell lines.^{( 11} , ¹² , ¹³ , ¹⁴ , ¹⁵ , ^{16 )} Similarly, in our study, inhibition of tumor cell growth (2, 3) and induction of apoptosis (Fig. 4) have been proved with a suboptimal dose of honokiol (10 µg/mL) in vitro. Recent reports have demonstrated that honokiol inhibits the growth of cancer cells through cell cycle arrest (G₀/G₁) and induction of apoptosis.^{( 13} , ⁵⁸ , ⁵⁹ , ^{60 )} The latter was associated with the modulation of Bcl‐2 family proteins such as Bad, Bcl‐2, and Bcl‐xL, release of cytochrome c from the mitochondria to the cytosol, and activation of caspases (caspase‐3, caspase‐8, and caspase‐9).^{( 13} , ⁵⁸ , ⁵⁹ , ^{60 )}

Furthermore, the in vitro enhanced antiproliferative and pro‐apoptotic activities of mPNAS‐4 plus honokiol correlates well with the in vivo improved antitumor efficacy. The enhanced antitumor efficacy in vivo was associated with the enhanced inhibition of angiogenesis and the augmented induction of apoptosis. The enhanced inhibition of angiogenesis was confirmed by the alginate assay (Fig. 7) and CD31 immunohistochemistry analysis (Fig. 6), whereas the augmented induction of apoptosis was verified by TUNEL analysis (Fig. 8).

In our study, honokiol inhibited angiogenesis in vivo. This confirmed previous observations by other authors reporting that honokiol inhibits angiogenesis through inhibition of VEGFR‐2/Flk/KDR autophosphorylation in endothelial cells.^{( 15 )} Treatment with mPNAS‐4 alone also partially inhibited angiogenesis in vivo, which was similar to our previous observation.^{( 30 )} The combined treatment with honokiol and mPNAS‐4 further enhanced the antiangiogenic efficacy. Previous studies have demonstrated that tumor cells can produce multiple highly active angiogenesis promoters that interact with one another to form an efficient angiogenic network, even residual tumor cells surviving in a dormant state can strongly promote neovascularization and tumor growth, suggesting that tumor cells have a robust ability of stimulating angiogenesis.^{( 61} , ^{62 )} Therefore, we speculate that induction of apoptosis by mPNAS‐4 causes reduction of tumor cells, which in turn leads to a decrease of active angiogenesis promoters, thus triggering inhibition of angiogenesis. However, the mechanisms of antiangiogenic generation in the treatment with mPNAS‐4 alone and those in the combined treatment remain to be elucidated.

In the in vivo experiments, it was notable that pcDNA3.1 empty vector suppressed the growth of CT26 and LL2 tumors to some extent (Fig. 5a). The pcDNA3.1 vector contains four unmethylated CpG motifs: two GACGTT motifs (one between the poly A site and the f1 ori, and the second within the neomycin resistance gene) and two AACGTT motifs (situated in the ampicillin resistance gene).^{( 63 )} Previous studies showed that unmethylated CpG DNA could enhance B cell survival, influence dendritic cell differentiation, and induce B cells, monocytes and natural killer cells to secrete cytokines such as IL‐6, IL‐12 and interferon‐γ.^{( 63} , ⁶⁴ , ^{65 )} These non‐specific immunostimulating effects might have an antitumor effect, thus contributing to the suppression of growth of LL2 in vivo.

In addition, the induction of apoptosis and the suppression of angiogenesis by mPNAS‐4 were unexpectedly not the strongest in vivo. This appeared to be inconsistent with the significant suppression of tumor growth. One possible explanation for this inconsistency is that the unmethylated CpG motifs in the pcDNA3.1‐mPS vector may induce non‐specific immunostimulating effects in vivo, thus contributing to the suppression of tumor growth.^{( 63} , ⁶⁴ , ^{65 )} The other possible explanation is that other processes induced by mPNAS‐4, such as cell cycle arrest, may be involved in tumor inhibition in vivo. Recently, we have confirmed that overexpression of hPNAS‐4 causes cell cycle arrest at S phase in A549 cells.^{( 31 )} However, other mechanisms of mPNAS‐4 for the suppression of tumor growth remain to be elucidated.

Previous studies have shown that antiangiogenic agents significantly increase the apoptosis of tumor cells.^{( 53} , ⁵⁴ , ^{55 )} Our data suggest that the enhanced induction of apoptosis may also play an important role in the suppression of tumor growth in the combination strategy. Although the mechanism for the enhanced induction of apoptosis is not known, some rational hypothesis may be offered to explain this pending question. One possibility is that the enhanced inhibition of angiogenesis decreases the diffusion of oxygen and nutrients to the solid tumors, thus causing tumor cell apoptosis. Another scenario may be that mPNAS‐4 and honokiol share a synergistic apoptotic pathway, that is, the release of mitochondrial cytochrome c to the cytosol and sequential activation of caspases, which may be one of the principal reasons for the improvement of antitumor effectiveness.

More recently, we have confirmed that hPNAS‐4 potentiates the antineoplastic effects of gemcitabine.^{( 32 )} In this work, mPNAS‐4 also enhances the antitumor efficacy of honokiol. As far as antitumor efficacy is concerned, this combination has no significant superiority over the combination of PNAS‐4 plus gemcitabine. However, this novel combination strategy is still of significance and warrants further attention. First, on account of the dose‐dependent toxicity of chemotherapy drugs and the protective effects of honokiol, such as hepatoprotective effects and neuroprotective activity,^{( 6 )} the combination may be an effective adjuvant therapy strategy of PNAS‐4 in combination with other conventional chemotherapy drugs. Second, given that honokiol can overcome drug resistance, the combined approach may have potential applications in the treatment of recrudescent or drug‐resistant cancer, although further research is necessary to explore the possible link between drug resistance and the combination strategy.

In conclusion, our data suggest that PNAS‐4, a novel pro‐apoptotic gene, can inhibit proliferation of tumor cells via apoptosis and augment the apoptosis of tumor cells induced by honokiol in vitro. Our data also demonstrate that the combination of mPNAS‐4 and honokiol enhances the antitumor efficacy in vivo, and that the improved therapeutic efficacy may be associated with the enhanced induction of apoptosis and the augmented inhibition of angiogenesis. Our results may provide a novel and effective approach to treat cancer.

Abbreviations

FBS	fetal bovine serum
hPNAS‐4	human PNAS‐4
IL	interleukin
mPNAS‐4	mouse PNAS‐4
MVD	microvessel density
NBF	neutral buffered formalin
pcDNA3.1‐mPS	pcDNA3.1 plasmid encoding mouse PNAS‐4 gene
pPNAS‐4	porcine PNAS‐4
TUNEL	terminal deoxynucleotidyltransferase‐mediated dUTP nick end labeling
VEGFR	vascular endothelial growth factor receptor

Disclosure Statement

No potential conflicts of interest were disclosed.