Honokiol activates AMP-activated protein kinase in breast cancer cells via an LKB1-dependent pathway and inhibits breast carcinogenesis

The human breast cancer cell lines, MCF7 and MDA-MB-231, were obtained from the American Type Culture Collection and maintained in DMEM supplemented with 10% fetal bovine serum (FBS) (Gemini Bioproducts, Woodland, CA, USA) and 2 μM L-glutamine (Invitrogen, Carlsbad, CA, USA). Cell-line authentication was done by analysis of known genetic markers or response (for example, expression of estrogen receptor and p53 and estrogen responsiveness) [20]. AMPK-null and AMPK-WT immortalized MEFs were kindly provided by Dr. Keith R. Laderoute (SRI International, Menlo Park, CA, USA) [21]. Honokiol is a natural product extracted from seed cone of Magnolia grandiflora, as previously described [6]. Antibodies for p-AMPK (phospho-AMPK), AMPK, ACC, p-ACC (phospho-ACC), pS6K, p-pS6K (phospho-S6K), 4EBP1, p-4EBP1 (phospho-4EBP1), p-Akt (phospho-Akt), Akt, and LKB1 (3047) were purchased from Cell Signaling Technology (Danvers, MA, USA).

Five pre-made lentiviral LKB1 short-hairpin RNA (shRNA) constructs and a negative control construct created in the same vector system (pLKO.1) were purchased from Open Biosystems (Huntsville, AL, USA). Paired LKB1 stable knockdown cells (MCF7 and MDA-MB-231) were generated by following our previously published protocol [22].

Cell-viability assay was performed by estimating the reduction of XTT (2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxyanilide), by using a commercially available kit (Roche) [23]. Breast cancer cells were treated with honokiol as indicated.

For colony-formation assay [24], MCF7 and MDA-MB-231 cells were treated with honokiol as indicated for 10 days; colonies containing > 50 normal-appearing cells were counted.

Anchorage-independent growth of MCF7 and MDA-MB-231 cells in the presence of honokiol treatment was determined by colony formation on soft agar [25]. Colonies were counted in five randomly selected fields at ×10 magnification by using Olympus IX50 inverted microscope.

Migration assay was performed according to our published protocol [26]. Cells were treated with honokiol as indicated. Plates were photographed after 24 and 48 hours at the identical location of the initial image.

Wound-healing assay was performed by using the ECIS (Applied BioPhysics, Troy, NY, USA) technology and following our previously established protocol [27].

MDA-MB-231 and MCF7 cells (1.5 × 10⁴) were seeded in 0.5% agar-coated plates and cultured on an orbital shaker (100 rpm) for 48 hours in a humidified atmosphere containing 5% CO₂ at 37°C. Intact tumor spheroids were selected and transferred to six-well plates. The spheroids were treated with honokiol, as indicated. After 48 hours of incubation, spheroids were fixed with 10% buffered formalin in PBS and stained with crystal violet. The migration of cells from spheroids was observed under a light microscope.

For an in vitro model system of metastasis, Matrigel invasion assay [28] was performed by using a Matrigel invasion chamber from BD Biocoat Cellware (San Jose, CA, USA). The slides were coded to prevent counting bias, and the number of invaded cells on representative sections of each membrane were counted under light microscope. The number of invaded cells for each experimental sample represents the average of triplicate wells.

Whole cell lysate [23] was prepared by scraping MCF7 and MDA-MB-231 cells in 250 μl of ice-cold modified RIPA buffer. An equal amount of protein was resolved on sodium dodecylsulfate polyacrylamide gel, transferred to nitrocellulose membrane, and Western blot analysis was performed. Immunodetection was performed by using enhanced chemiluminescence (ECL system; Amersham Pharmacia Biotech, Arlington Heights, IL, USA) according to manufacturer's instructions.

Immunoprecipitation of LKB1 was performed by following the previously published protocol [23] by using anti-LKB1 antibody followed by immunoblotting with anti-STRAD antibody.

Breast cancer cells (5 × 10⁵ cells/well) were plated in four-well chamber slides (Nunc, Rochester, NY, USA) followed by treatment with honokiol and subjected to immunofluorescence analysis as described [22]. Fixed and immunofluorescently stained cells were imaged by using a Zeiss LSM510 Meta (Zeiss) laser scanning confocal system configured to a Zeiss Axioplan 2 upright microscope with a 63XO (NA 1.4) plan-apochromat objective. All experiments were performed multiple times by using independent biologic replicates.

MDA-MB-231 (5 × 10⁶) cells in 0.1 ml of HBSS were injected subcutaneously into the right gluteal region of 4- to 6-week-old female athymic nude mice. Two weeks after initial implantation, the animals were placed into two experimental groups. Mice were treated with intraperitoneal injections of (a) control (saline and Intralipidor (b) honokiol, at 3 mg/mouse/day in 20% Intralipid (Baxter Healthcare, Deerfield, IL, USA), 3 times per week for the duration of the experiment. Tumors were measured by using vernier calipers, with tumor volume calculated by using the formula (V = a/2 × b2), where V is the tumor volume in cubic millimeters, and a and b are the largest and smallest diameters in millimeters, respectively. All animals were killed after 4 weeks of treatment. Tumors were collected; weighed, fixed in 10% neutral-buffered formalin; and subjected to further analysis with immunohistochemistry.

We used tumor sections to determine the effect of honokiol on expression of p-AMPK, LKB1, and Ki-67 by immunohistochemistry. Immunohistochemistry was performed essentially as described by us previously for other proteins [25, 26]. At least four nonoverlapping representative images from each tumor section from five mice of each group were captured by using ImagePro software for quantitation of p-AMPK, LKB1, and Ki-67 expression. Total cell lysates were prepared from tumor samples and subjected to immunoblot analysis. All animal studies were conducted in accordance with the guidelines of University ACUC.

All experiments were performed thrice in triplicates. Statistical analysis was performed by using Microsoft Excel software. Significant differences were analyzed by using the Student t test and two-tailed distribution. Data were considered to be statistically significant if P < 0.05. Data were expressed as mean ± SEM between triplicate experiments performed thrice.

Growth-inhibition and apoptosis-induction properties of honokiol have been reported in several cancer cell lines [3, 9‐12, 18]. In the current study, two breast cancer cell lines, MCF7 and MDA-MB-231, were treated with various concentrations ranging from 1 μM to 25 μM honokiol and subjected to clonogenicity (Figure 1a) and anchorage-independent growth assay (Figure 1b). Dose-dependent and statistically significant inhibition of clonogenicity and soft-agar colony formation was observed in the presence of honokiol. Treatment with 5 μM honokiol resulted in ~50% to 60% inhibition in clonogenicity and soft-agar colony formation, whereas higher concentrations (10 and 25 μM) were more inhibitory (Figure 1a, b). We further examined the effect of honokiol on the growth of HCC1806 breast cancer cells, which harbor an LKB1 homozygous mutation, by using clonogenicity and soft-agar colony formation assay. Our studies show that honokiol does not inhibit the growth of HCC-1806 cells (Additional file 1). These results indicate that LKB1 might be an integral molecule for honokiol-mediated growth inhibition.

Cancer progression is a multistep process that involves invasion of basement membrane by tumor cells and migration to points far from a given primary tumor mass, leading to metastasis [29]. We examined the effect of honokiol on breast cancer cell migration and invasion by using scratch migration, electric-cell-substrate impedance sensing (ECIS)-based migration, spheroid migration, and Matrigel invasion assays. Honokiol treatment resulted in inhibition of migration of breast cancer cells (Figure 2a) in comparison with untreated cells. For quantitative determination of alteration in the migration potential of breast cancer cells on treatment with honokiol, we performed a quantitative real-time impedance assay by using an ECIS-based technique. As expected, confluent cells showed high resistance values. Confluent cells were subjected to a high-voltage pulse that resulted in decrease in resistance, indicating death and detachment of cells present on the small active electrode. Cells were left untreated or treated with honokiol, and changes in resistance were recorded for 24 hours. Control untreated cells showed an increase in resistance, showing increased migration of cells surrounding the small active electrode that were not submitted to the elevated voltage pulse to reach the resistance values of the nonwounded cells at the start of the experiment. Honokiol-treated cells showed a decrease in resistance, indicating decreased migration. Notably, honokiol-treated cells never reached the values of nonwounded cells, showing significant inhibition of migration potential (Additional file 2). We examined the effect of honokiol treatment on the migratory capacity of MCF7 and MDA-MB-231 cells spheroids. Significant migration of MCF7 and MDA-MB-231 cells from the spheroids was seen under untreated conditions. Honokiol treatment resulted in inhibition of migration of cells from spheroids (Figure 2b). Next, we performed Matrigel invasion assay to examine the effect of honokiol on the invasion potential of breast carcinoma cells. As evident from Figure 2c, honokiol treatment decreased invasion of breast cancer cells through Matrigel in comparison with untreated cells. Activation of FAK has been shown to regulate cancer cell migration and invasion through distinct pathways by promoting the dynamic regulation of focal adhesion and peripheral actin structures [30‐32] and matrix metalloproteinases (MMPs)-mediated matrix degradation [33]. We examined whether honokiol treatment affects FAK activation to inhibit migration and invasion of breast cancer cells. Honokiol treatment inhibited FAK phosphorylation in breast cancer cells, indicating the involvement of FAK activation in honokiol-mediated inhibition of migration and invasion potential of breast cancer cells (Figure 2d). Collectively, these results show that honokiol treatment can effectively inhibit clonogenicity, anchorage-independent colony formation, migration, and invasion of breast carcinoma cells.

Honokiol modulates multiple pathways (nuclear factor (NF)-κB, ERK, Akt, and JNK) in a cellular process and target-tissue-dependent manner [7, 9‐11, 19]. AMP-activated protein kinase (AMPK) is a serine/threonine protein kinase that acts as a master sensor of cellular energy balance in mammalian cells by regulating glucose and lipid metabolism [34]. Recent studies have implicated AMPK as an important factor in cancer cell growth and migration [35, 36]. Thus, we sought to determine the effect of honokiol on AMPK phosphorylation and activation. Honokiol treatment stimulated phosphorylation of AMPK at Thr-172 in MCF7 and MDA-MB-231 cells. Honokiol had no effect on total AMPK protein expression levels (Figure 3a). AMPK phosphorylation at Thr-172 has been widely associated with its activation [37]. Once activated, AMPK directly phosphorylates and inactivates a number of ATP-consuming metabolic enzymes including acetyl-coenzyme A carboxylase (ACC) [38]. We examined the phosphorylation of ACC to evaluate AMPK activity with honokiol treatment. Increased phosphorylation of ACC in MCF7 and MDA-MB-231 cells was observed in response to honokiol treatment as compared with untreated cells, whereas total ACC protein levels remain unchanged (Figure 3a). Activation of AMPK leads to suppression of mammalian target of rapamycin (mTOR) signaling, and the molecular mechanisms involve phosphorylation of tuberous sclerosis complex protein TSC2 at Thr-1227 and Ser-1345 that increases the activity of the TSC1-TSC2 complex to inhibit mTOR [37, 39]. Two very well characterized and widely studied downstream effectors of mTOR are the p70 kDa ribosomal protein S6 kinase 1 (p70S6K1 or pS6K) and the eukaryotic translation initiation factor 4E (elF4E)-binding protein (4EBP1) [40]. Phosphorylation of pS6K and 4EBP1 has been widely used to assess changes in mTOR activity in response to various growth-factor pathways. We next examined the effect of honokiol on mTOR activity in breast cancer cells. Honokiol decreased phosphorylation of pS6K and 4EBP1 in both MCF7 and MDA-MB-231 cells while not affecting the total protein levels of pS6K and 4EBP1 (Figure 3b). Recent studies have shown that pS6K regulates the actin cytoskeleton by acting as an actin filament cross-linking protein and as a Rho family GTPase-activating protein [41]. It has been shown that reorganization of the actin cytoskeleton is critical for cell migration, as motile cancer cells must assemble and disassemble the actin filaments at their leading edges [42]. Depletion or inhibition of the activity of pS6K results in inhibition of actin cytoskeleton reorganization and inhibition of migration [41]. Owing to the integral role of pS6K in cancer cell migration, it is possible that honokiol-mediated inhibition of migration is mediated through pS6K inhibition.

mTOR, a key regulator of cell growth and proliferation, exists in two structurally and functionally distinct multiprotein complexes, mTORC1 and mTORC2. mTORC1 is known to activate protein synthesis and cell growth through regulating pS6K and 4E-BP1 activity, whereas mTORC2 phosphorylates Akt on Ser-473, activating cell growth, proliferation, and survival [43, 44]. We found that honokiol increases AMPK activation and inhibits mTORC1 function, as evidenced by inhibition of pS6K and 4E-BP1 phosphorylation.

We next determined whether honokiol treatment modulates mTORC2 function. mTORC2 phosphorylates Akt on Ser-473. Therefore, to determine whether mTORC2 is also inhibited by honokiol under similar conditions, breast cancer cells were treated with honokiol, and the phosphorylation of Akt was determined. Honokiol did not alter Akt phosphorylation on Ser-473 in breast cancer cells (Additional file 3). These results provide evidence that honokiol only inhibits mTORC1 in breast cancer cells. Contrasting findings have been reported previously, showing reduction in Akt phosphorylation in response to honokiol treatment. Of note, MDA-MB-231 cells were treated with much higher concentrations of honokiol (60, 80, and 100 μM) in this study [45]. Hence, the observed decrease in Akt phosphorylation may be due to the treatment with higher concentrations of honokiol. Honokiol inhibits breast cancer growth in a concentration-dependent manner, with higher concentrations much more inhibitory than lower concentrations (Figure 1).

Although our findings clearly showed the involvement of AMPK activation in the honokiol signaling network, we raised the question whether honokiol-induced inhibition of mTOR and cell migration requires AMPK protein. We used MEFs derived from AMPK-WT (WT) and AMPK knockout (AMPK-null) mice to test the potential requirement of this protein in honokiol-mediated inhibition of migration. Immunoblotting confirmed the absence of the AMPK protein in AMPK-null MEFs (Figure 4a). In agreement with the absence of AMPK protein, the AMPK-null MEFs did not show any phosphorylation of ACC, even in the presence of honokiol. AMPK-WT MEFs, conversely, exhibited honokiol-stimulated phosphorylation of ACC, indicating activation of AMPK (Figure 4b). Exposure of MEFs derived from AMPK-WT mice to honokiol resulted in inhibition of phosphorylation of pS6K, whereas the MEFs derived from the AMPK-null mice were significantly resistant to the honokiol-mediated inhibition of pS6K phosphorylation (Figure 4d).

We next asked whether AMPK is directly involved in honokiol-mediated inhibition of migration. AMPK-WT MEFs exhibited inhibition of migration in response to honokiol treatment in scratch migration as well as ECIS-based migration assay. Interestingly, honokiol treatment could not inhibit migration of AMPK-null MEFs (Figure 4c; Additional file 4). AMPK knockdown also inhibited the antiproliferative effect of honokiol (Figure 4e). These results showed that AMPK is an integral molecule in mediating the negative effects of honokiol on the mTOR axis and migration potential of cells.

The tumor suppressor LKB1 (also known as Stk11) is an evolutionarily conserved serine/threonine protein kinase that has a broad range of cellular functions, including tumor suppression, cell polarity, cell-cycle regulation, and promotion of apoptosis [46, 47]. LKB1 has recently been identified as a critical upstream kinase for AMPK, regulating its activity. Intriguingly, we found that honokiol increases expression of tumor suppressor LKB1 in MCF7 and MDA-MB-231 cells, with a significant increase at 1 hour of treatment with maximal expression at 24 hours in MCF7 cells and at 4 hours in MDA-MB-231 cells (Figure 5a). Variable expression of LKB1 in MDA-MB-231 breast cancer cells has been reported [48, 49]. We recently procured MDA-MB-231 cells from various established breast cancer research laboratories and analyzed the expression and functional status of LKB1. Our data unequivocally showed the presence of functional LKB1 in MDA-MB-231 cells [22]. Human LKB1 is both nuclear and cytoplasmic, but a mutant of LKB1 lacking the nuclear localization signal still retains the ability to suppress cell growth, suggesting that the cytosolic pool of LKB1 plays an important role in mediating its tumor-suppressor properties [50, 51]. STRAD (Ste20-related adaptor) protein has been shown to form a complex in which STRAD activates LKB1, resulting in cytoplasmic translocation of LKB1 [47]. We investigated the effect of honokiol on the formation of the LKB1-STRAD complex in breast cancer cells. To address this question, breast cancer cells were treated with honokiol followed by immunoprecipitation with LKB1 antibodies. Immunoprecipitated protein complexes were analyzed for the presence of STRAD by using Western blot analysis. Higher levels of STRAD immunoprecipitated with LKB1 in the presence of honokiol indicated increased formation of the LKB1-STRAD complex (Figure 5b). Immunostaining of honokiol-treated MCF7 and MDA-MB-231 cells revealed that honokiol treatment increases cytoplasmic accumulation of LKB1. LKB1 was localized predominantly in the nucleus in untreated breast cancer cells, although cytoplasmic LKB1 expression was also detected (Figure 5c). Control experiments with secondary antibody (results not shown) gave an extremely faint background staining that was distributed uniformly throughout the cells, irrespective of the treatment. Studies on the subcellular localization of LKB1 have indicated a wide variety of localization patterns. Mouse LKB1 was found to be predominantly nuclear, whereas Caenorhabditis elegans PAR-4 and Xenopus XEEK1 were detected exclusively in the cytoplasm [52‐54]. Human LKB1 has been detected to be both nuclear and cytoplasmic in several cell types [55, 56]. Although LKB1 expression is exclusively cytoplasmic in lung and pancreatic cancer [57, 58], gastrointestinal hamartomatous polyps from Peutz-Jeghers syndrome patients, head and neck squamous cell carcinoma, invasive lobular breast carcinoma, and solid papillary ductal carcinoma in situ breast cancer show both cytoplasmic and nuclear LKB1 expression [59‐61]. Studies in adult rat primary cardiomyocytes and C2C12 myoblasts showed that LKB1 was located predominantly in nucleus and undergoes cytoplasmic localization in various stimulations [62‐64]. In vitro studies suggest that nuclear LKB1 regulates cell-cycle progression and acts as a transcription factor [65, 66], whereas cytoplasmic LKB1 participates in controlling energy metabolism and cell polarity [67]. It is not completely understood how subcellular localization of LKB1 affects its tumor-suppressor function and activation of other signaling pathways in vivo.

We raised the question whether LKB1 plays an important regulatory role in honokiol- mediated modulation of AMPK and inhibition of migration and invasion of breast cancer cells. To address these questions, we used LKB1^shRNA lentivirus and puromycin to select for stable pools of MCF7 and MDA-MB-231 cells with LKB1 depletion. We analyzed pLKO.1 and LKB1^shRNA stable MCF7 and MDA-MB-231 cell pools for LKB1 protein expression with immunoblot analysis and found that LKB1 protein expression was significantly reduced in LKB1^shRNA cells (shRNA1 and shRNA2) as compared with pLKO.1 control cells (Figure 5d). pLKO.1 and LKB1^shRNA cells were treated with honokiol, and phosphorylation of AMPK was determined by using Western blot analysis. We found that honokiol increased phosphorylation of AMPK in pLKO.1 cells. Intriguingly, displaying a crucial role of LKB1, honokiol treatment did not change the phosphorylation levels of AMPK in LKB1^shRNA cells (Figure 5e). Invasion and migration are the key biologic features of malignant behavior of carcinoma cells [29]. In addition to examining the effect of LKB1 depletion on honokiol-induced modulation of AMPK, we also examined the requirement of LKB1 in honokiol-mediated inhibition of metastatic properties of breast cancer cells. As evident from Figure 5f, honokiol treatment efficiently inhibited migration of pLKO.1 cells, whereas untreated pLKO.1 cells showed increased migration. Our results showed that LKB1^shRNA cells exhibited increased migration in the absence of honokiol treatment. Interestingly, honokiol treatment did not inhibit the migration of LKB1^shRNA cells (Figure 5f). We next examined the effect of honokiol on invasion potential of pLKO.1 and LKB1^shRNA cells and found that honokiol inhibited invasion of pLKO.1 cells, whereas LKB1^shRNA cells were not affected by honokiol treatment (Figure 5g). These results collectively show that honokiol-induced LKB1 overexpression is indeed a crucial component of the signaling machinery used by honokiol in modulating the AMPK-S6K axis and inhibiting the metastatic properties of breast cancer cells.

We investigated the physiological relevance of our in vitro findings by evaluating whether honokiol has any suppressive effects on the development of breast carcinoma in nude mouse models and the involvement of the LKB1-AMPK axis. In the experimental group treated with honokiol, the rate of tumor growth was significantly inhibited, and the tumor size and weight were significantly reduced compared with control group (Figure 6a, b). The immunohistochemical assessment of tumor proliferation showed higher Ki-67 in the control group as compared with the honokiol-treated group (Figure 6c). In our in vitro analyses, we discovered the involvement and requirement of the LKB1-AMPK axis in biologic functions of honokiol. We examined the expression of LKB1 and p-AMPK in tumors treated with honokiol. Tumors treated with honokiol displayed higher levels of phosphorylated AMPK and LKB1 (Figure 6c). In addition, we examined the expression levels of phosphorylated and unphosphorylated AMPK, ACC as well as S6K, in honokiol-treated and vehicle-treated mice. We found higher levels of phosphorylated AMPK and ACC in honokiol-treated tumors as compared with vehicle-treated controls. Honokiol-treated tumors showed lower levels of phosphorylated S6K, whereas vehicle-treated controls exhibited high levels of phosphorylated S6K (Figure 6d). These data presented direct in vivo evidence of the involvement of LKB1-AMPK activation and the subsequent inhibition of pS6K in honokiol function (Figure 6e).