Bufalin, a Traditional Chinese Medicine Compound, Prevents Tumor Formation in Two Murine Models of Colorectal Cancer Free

Colorectal cancer has become the third leading cause of cancer-related death among all fatal cancer types worldwide (1). As in most cancers, colorectal cancer is largely related to a complex interplay between genetics and the environmental factors and typically progresses in a stepwise fashion. The accumulation of genetic mutations in oncogenes and tumor suppressor genes caused by aging and inherited mutations are considered as major risk factors of colorectal cancer (2, 3). For instance, adenomatous polyposis coli (APC) tumor suppressor gene mutation is known as the classical route and clearly proved to be a significant risk factor for the development of colorectal cancer (4). On the other hand, colonic tumorigenesis is also driven by chronic inflammation which is another key promoter of colorectal cancer. This prolonged low-grade inflammatory response involves a progressive change in the type of cells present at the site of inflammation and continuous production of DNA-damaging oxygen radical species, extracellular matrix-modifying enzymes and growth factors which can infiltrate into the local mucosal tissue in the colon, leading to a microenvironment conducive to tumorigenesis (5). Given the fact that it takes years to decades for normal mucosa developing into carcinoma, a window of opportunity, therefore, is offered to prevent the incipient morbidity of colorectal cancer by employing proper interventions (6). At this point, the concept of chemoprevention is highlighted once again in recent decades. Chemopreventive agents are many and can be classified into various modalities, including non-toxic naturally occurring compounds, dietary extracts and synthetic pharmacotherapeutic compounds.

Bufalin is an active ingredient of the traditional Chinese medicine “Chan Su,” which is an extract of dried toad venom from the skin glands of Bufo gargarizans or Bufo melanostictus. Bufalin was found to inhibit Na⁺/K⁺-ATPase and is conventionally used as a cardiac glycoside to increase the contractile force of cardiomyocytes by enhancing Ca²⁺ influx (7). Intriguingly, emerging evidence has shown that bufalin exhibits potent anticancer action in various cancer cell lines, solid tumors and hematologic malignancies including lung (8), gallbladder (9), breast (10), liver (11), prostate (12), bone (13), gastric (14), pancreatic (7), endometrial, ovarian (15), and oral cancers (16) as well as melanoma (17) and leukemia (18). The anticancer action of bufalin is attributed not only to its abilities to induce cell-cycle arrest, apoptosis, differentiation, and autophagy in cancer cells but also to its inhibitory effects on inflammation, angiogenesis, and metastasis. It has been postulated that direct inhibition of Na⁺/K⁺-ATPase is involved in these actions (19).

Previously, we elucidated the anticancer action of bufalin in colorectal cancer in vitro (20) and found that bufalin could induce G₂–M cell-cycle arrest in cultured colon cancer cells via inhibition of both hypoxia-inducible factor (HIF)-1α and NF-κB to mediate downregulation of serine/threonine-protein kinase (PIk1, also known as polo-like kinase 1). In addition, we found that bufalin could induce reactive oxygen species (ROS) to activate c-Jun NH2-terminal kinase (JNK) and thereby promoting autophagic cell death in colorectal cancer cells. The anticancer mechanisms reported by other investigators in other cancer types include activation of extrinsic and intrinsic pathways of apoptosis, inhibition of oncogenic signaling or molecules and transactivation of vitamin D receptor.

Epidemiologic studies have elucidated that the use of cardiac glycosides such as digitalis is associated with a reduced risk of prostate cancer (12). Dose-dependent relationships between plasma concentration of cardiac glycoside and a lower risk of leukaemia/lymphoma as well as cancers in the kidney/urinary tract have also been reported (21). In addition, patients with breast cancer patients on cardiac glycoside therapy tend to have reduced recurrence (22). However, before putting cardiac glycosides into clinical use for chemoprevention, more experimental evidence from animal models has to be obtained. “Huachansu,” an injectable form of “Chan Su” containing bufalin at 14.3 ± 0.03 ng/mL, has been officially approved for cancer therapy in China. Previous studies demonstrated that no significant toxicity was observed in using Huachansu injection at doses up to eight times the typical dose used in China in patients with hepatocellular carcinoma and pancreatic cancer (7). It is noteworthy that although the anticancer effects of bufalin have been demonstrated in colorectal cancer (in vitro only), it remains unclear whether bufalin could exhibit anti-colorectal cancer effects in vivo, and no profound evidence has been raised to validate the precise mechanisms of its anticancer and chemopreventive actions in in vivo models. On the basis of the relative safety of bufalin and the epidemiologic findings on the association between the use of cardiac glycosides and cancer together with our previous data on the anticancer action of bufalin in colorectal cancer, we were encouraged to further explore whether bufalin could be employed as an effective chemoprevention agent to prevent colon tumorigenesis in two murine models of colorectal cancer.

The antibodies used were β-actin, anti-cyclin A, anti-cyclin D1, anti-cyclin E, anti-CDK2, anti-CDK4, anti-p21, anti-p27, anti-Bcl-2, anti-Bcl-xL, anti-survivin, anti-Bax, anti-Bak, anti-caspase-9, anti-cleaved caspase-3, anti-COX-2, anti-p-IκBα, anti-IκBα, anti-p-IKKα, anti-IKKα, anti-p-p50, anti-p50, anti-p-p65, anti-p65, anti-p-PI3K, anti-PI3K, anti-p-Akt, anti-Akt, anti-p-mTOR, anti-mTOR, anti-p-GSK-3β, and anti-GSK-3β. All antibodies above were purchased from Cell Signaling Technology. Bufalin (Wako Pure Chemical Corporation), azoxymethane (AOM; Sigma), dextran sulphate sodium (DSS; International Lab), Alexa Fluor 555–conjugated donkey anti-rabbit secondary antibody (Life Technologies).

All mice were previously purchased from The Jackson Laboratory and maintained at the Laboratory Animal Services Centre of The Chinese University of Hong Kong. During experiments, animals were housed at the Animal Holding Core facility of the School of Biomedical Sciences. All mice were allowed to acclimate for one week before experiments commenced. All experimental procedures were designed in accordance to the U.S. NIH Guidelines for the Care and Use of Laboratory Animals and approved by the Department of Health and the Animal Experimentation Ethics Committee of The Chinese University of Hong Kong.

Colitis-associated colorectal cancer model was established as described previously (23). Briefly, 7-week-old male BALB/c mice were injected intraperitoneally with AOM (10 mg/kg). After 1 week, 3% DSS in drinking water was provided to induce colon inflammation for 1 week interrupted by a 2-week period of normal drinking water without DSS for three consecutive cycles. Mice at 5 weeks old (2 weeks before AOM injection) were started providing with bufalin (0.5 mg/kg) or vehicle through intraperitoneal injection every other day for 16 weeks. Mice without AOM/DSS treatment were injected with vehicle as the negative control. After treatment, mice were euthanized by cervical dislocation. Tissues were collected for the followed experiments.

Male Apc^Min/+ mice on C57BL/6J genetic background were obtained by mating with wild-type female mice. Heterozygous genotypes of mice were verified by allele-specific PCR using genomic DNA extracted from tail biopsies with specific primers (wild-type, 5′-GCCATCCCTTCACGTTAG-3; Apc^Min, 5′-TTCTGAGAAAGACAGAAGTTA-3′; common antisense, 5′-TTCCACTTTGGCATAAGGC-3′). Mice at 6 weeks old were injected intraperitoneally with bufalin (0.5 mg/kg) or vehicle every other day for 12 weeks. Wild-type mice injected with vehicle were included as negative control. After treatment, mice were euthanized by cervical dislocation. Colons and small intestines were excised and collected for the followed experiments.

The human colon cancer HCT116 and SW620 cells were purchased from the ATCC and were cytogenetically tested and authenticated before the cells were frozen. All experiments were performed in HCT116 and SW620 cells between passage 10 and 25, without any evidence of phenotypic change. HCT116 cells were routinely passaged in McCoy's 5A medium and SW620 cells in RPMI1640 medium. All experiments were carried out in complete medium containing 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin at 37°C in a 5% CO₂ atmosphere.

siRNAs against human Bax, Bak, Na⁺/K⁺-ATPase α3 were purchased from Genepharm. In brief, cells were transfected in McCoy 5A and RPMI-1640 mediums (respectively) with 90 nmol/L of each siRNA duplex using DharmaFECT transfection reagent according to the manufacturer's protocol. The siRNA duplexes targeting sequences are listed in Supplementary Table S1.

Total protein was extracted by lysing colon tissues in cold RIPA buffer (150 mmol/L NaCl, 50 mmol/L Tris, 1% Triton-100, 0.5% sodium deoxycholate, 0.1% SDS, pH 7.6) supplemented with PhosSTOP phosphatase inhibitors (Roche) and cocktail protease inhibitors (P8340, Sigma) and run on SDS-PAGE gels, then transferred onto a nitrocellulose membrane (Pall Corporation). The membrane was incubated with respective primary antibodies at 4°C overnight and then incubated with horseradish peroxidase (HRP)-conjugated secondary antibody as described previously (20). Chemiluminescence was developed with HRP substrate (Bio-Rad) and detected under Bio-Rad ChemiDoc Imaging Systems.

Total RNA was extracted from colon tissues using RNAiso Plus (TaKaRa) and reverse transcribed into cDNA using PrimeScript RT Reagent kit (TaKaRa) as per manufacturer's protocol. The synthesized cDNA was then amplified by quantitative PCR using ChamQ SYBR qPCR Master Mix (Vazyme) on ABI Quantstudio 7 Flex Real Time PCR System. Expression of target genes was normalized against Actb using 2^−ΔΔC_t method. Primer sequences are listed in Supplementary Table S2.

Formalin-fixed paraffin-embedded colon tissues were sectioned at 5 μm thickness. IHC was performed using IHC Select Immunoperoxidase Secondary Detection System (Merck Millipore) following the manufacturer's protocol. Images were taken under a Carl Zeiss Axiophot 2 Upright Microscope. Histology score (H-score) for bromodeoxyuridine (BrdUrd) and cyclooxygenase-2 (COX-2) staining was evaluated on the basis of the percentage and intensity of immunoreactivity in tumor cells. Percentage of immunoreactivity was graded as: 1 (<20%); 2 (21%–40%); 3 (>40%). Intensity was graded as: 0 (none); 1 (weak); 2 (moderate); 3 (strong). H-score was calculated as the average of the products of grade for percentage and intensity of immunoreactivity from four random fields.

Formalin-fixed paraffin-embedded colon tissue sections were deparaffinized, rehydrated, undergone antigen retrieval process following a standard protocol. For cleaved caspase-3 staining, sections were incubated with primary antibody against cleaved caspase-3 at 4°C overnight, then incubated with Alexa Fluor 555–conjugated secondary antibody. For in situ TUNEL assay, procedures were performed using In Situ Cell Death Detection Kit, Fluorescein (Roche) according to the manufacturer's protocol. Sections were counter-stained with DAPI and mounted using DAPI Fluoromount-G (Southern Biotech). Images were taken under an Olympus FV1200 Confocal Microscope. Mucosal TUNEL-positive and cleaved caspase-3-positive cells were evaluated in five random fields and the average was calculated for each sample.

Student t test was performed when comparing two groups and one-way ANOVA was performed when comparing more than two groups. Kaplan–Meier plot and log-rank test were used to assess and compare the survival outcomes of mice. Mann–Whitney U test was used to compare histology scores in IHC analysis. All statistical tests were conducted using GraphPad Prism 6. Data were expressed as mean ± SD from at least three independent experiments. A two-sided P value less than 0.05 was considered statistically significant.

To investigate the role of bufalin in colorectal cancer, two murine models of colon tumorigenesis had been adopted, viz colitis-associated colorectal cancer and colonic tumors induced by Apc germline mutation, which resemble inflammatory and genetically predisposed pathways of colorectal cancer in humans, respectively. Colitis-associated colorectal cancer often appears in patients suffering from ulcerative colitis, a chronic inflammatory bowel disease. By using a protocol that combines the carcinogen AOM with repeated administration of the irritant DSS to induce colitis (Fig. 1A), we generated mice that developed colon tumors mainly in the distal to the middle colon (Fig. 1B), which is the predominant location of human colorectal tumors. The body weights of mice were reduced substantially during each exposure cycle to the 3% DSS and regained after withdrawal of DSS. Mice treated with bufalin, however, did not show extensive alterations in body weight loss compared to the DSS-vehicle group (Supplementary Fig. S1A). In line with this observation, survival curves showed no significant difference between these two groups (Supplementary Fig. S1B), suggesting that administration of bufalin at 0.5 mg/kg is relatively safe.

The severity of DSS-induced colorectal inflammation can be reflected by a significant visual indicator, colon shortening (24). In our study, we observed that the colorectal length of the group of mice treated with DSS was shorter than that of the vehicle-treated group. Interestingly, DSS-induced colon shortening was markedly prevented in bufalin-treated mice (Fig. 1B and C). By analyzing the growth patterns of colon tumors, we observed that upon long-term (16 weeks) and low-dose (0.5 mg/kg, every other day) administration of bufalin, mice had decreased tumor incidence with significantly lower average tumor load (Fig. 1D) and less macroscopic tumors than that of the DSS-vehicle group (Fig. 1B). We also noticed that bufalin-treated mice had lower average tumor size (Fig. 1E). Statistical analysis showed that bufalin-treated mice had globally reduced tumors across all different sizes (Fig. 1F), not only the macroscopic tumors but also the microscopic ones, suggesting that bufalin might have strong effect in suppressing tumor growth in the initiation and promotion stage. These results indicate that bufalin could significantly inhibit DSS-induced colitis and exhibit the chemopreventive effect on carcinogen-triggered early tumorigenesis in colon.

To further investigate the chemopreventive action of bufalin in colorectal cancer, we established another mouse model, Apc germline mutant mouse model. The autosomal dominant mutation of multiple intestinal neoplasia (Min) causes a loss of function mutation in the murine adenomatous polyposis coli (Apc) locus, which encodes a nonsense mutation at codon 850 (25). Mutations of APC gene predispose individuals to familial adenomatous polyposis (FAP), which is characterized by multiple adenomas in intestine and a small number of polyps in colon (26). Like humans with germline mutations in APC, Apc^Min/+ mice are predisposed to intestinal adenomas and colonic polyps as well, and thus provide us a good animal model for studying the chemopreventive role of bufalin in genetically predisposed pathways of colorectal cancer. Using a protocol that treated mice with bufalin (0.5 mg/kg) or vehicle every other day for a total 12 weeks starting from 6 weeks old (Fig. 2A), we found no significant changes in colon length (Fig. 2B).

Similar to the colitis-associated colorectal cancer model, we found that long-term and low-dose administration of bufalin ameliorated the development of tumors in terms of number and size in both colon and intestine in Apc^Min/+ mice. Bufalin-treated Apc^Min/+ mice showed markedly reduced average tumor number in colon (Fig. 2C) and intestine (Fig. 2D). Meanwhile, our results indicate that Apc^Min/+ mice had less microscopic tumors in both colon (Fig. 2E) and intestine (Fig. 2F) upon bufalin administration. In addition, the average number of macroscopic tumors in colon tissues from Apc^Min/+ mice definitely showed a trend between the vehicle and bufalin groups, although no significance has been reached (Fig. 2E). Given the large reduction in tumor incidence, these findings suggest that treatment of bufalin might have a stronger effect in early stages of tumor initiation and/or promotion. Similar to the colitis-associated colorectal cancer model, Apc^Min/+ mice treated with bufalin did not show significant alterations in body weight loss compared with the Apc^Min/+ vehicle group (Supplementary Fig. S2A). And no significant difference was found between these two groups in mice survival outcomes as well (Supplementary Fig. S2B).

To evaluate the action of bufalin on colonic cancer cell proliferation, BrdUrd staining assay was performed in both AOM/DSS-induced tumors and Apc germline mutation–developed tumors. In brief, mice were given a single intraperitoneal injection of BrdUrd (100 mg/kg) prior to sacrifice, and the BrdUrd staining was then carried out following a standard protocol. With the induction of AOM/DSS, abundant BrdUrd-positive cells were detected throughout the crypts in colonic tumor mucosal area as compared with mice treated with vehicle alone. Upon bufalin treatment, a reduced number of BrdUrd-positive cells was observed in tumor cells from the crypt-villus junction down to the crypt base (Fig. 3A). Similar results were obtained using the same cellular analysis in the colonic crypts from Apc^Min/+ mice (Fig. 3B). Evidence of reduced BrdUrd incorporation from IHC was followed by statistical analysis, by which we found significant decreases in histology scores in bufalin-treated groups in both murine models.

To further confirm this observation, several cell proliferation markers were assessed. We analyzed the whole colonic tumor tissues by Western blot analysis. Notably, a striking attenuation of cyclin A, cyclin D1, cyclin E, CDK2, and CDK4 protein levels were detected respectively from AOM/DSS–treated mice (Fig. 3C) and Apc^Min/+ mice (Fig. 3D) after long-term administration of bufalin. For this, we also assessed the expressions of two universal cyclin-CDK inhibitors, p21 and p27, which consistently showed enhanced protein levels upon bufalin treatment (Fig. 3C and D). On the basis of these findings, we clearly revealed that bufalin indeed prevented the colonic tumor cell proliferation in both colitis-associated and Apc mutant–developed colorectal cancer.

Normal mucosa in colon tissue maintains its homeostasis by a dynamic balance between the proliferation and apoptosis of colonic epithelial cells, where the proliferative activities are localized to the lower part of the crypt and apoptotic frequencies are found at the top of the crypt (27). However, these gradients are reversed in tumor cells. Increased levels of proliferation are found toward the upper part of the crypt, whereas promoted apoptosis is localized at the bottom of the crypt; thus, a greater overall rate of apoptosis is found in tumor cells than in normal crypt (28).

To determine whether bufalin shed its chemopreventive action on apoptotic activity in two murine models, we performed in situ TUNEL assay and stained cleaved caspase-3 in parallel. With the induction of AOM/DSS, an increased number of TUNEL-positive cells and cleaved caspase-3–positive cells, which mainly localized on the top of crypts, were detected from the mice of AOM/DSS-vehicle group compared with vehicle group (Fig. 4A). Interestingly, significantly higher numbers of TUNEL-positive cells and cleaved caspase-3–positive cells were observed from the top down to the bottom base of the crypt in tumor cells from bufalin-treated mice (Fig. 4A), indicating an overall elevated apoptosis in tumors upon bufalin administration. Similar results were obtained from Apc^Min/+ mice (Fig. 4B).

To further explore the molecular basis of these observations, the protein levels of several apoptotic markers were then determined by Western blot analysis. We noticed that decreased levels of antiapoptotic proteins Bcl-2, Bcl-xL, and survivin (Fig. 4C and E) were concomitant with enhanced levels of proapoptotic proteins Bax, Bak, and cleaved caspase-9 and caspase-3 (Fig. 4D and F) in tumor cells from bufalin-treated groups, respectively, suggesting that bufalin might promote apoptosis through intrinsic pathway in both murine models. To further validate the roles of Bax and Bak in bufalin-induced apoptosis, we used specific siRNAs against Bax and Bak on two human colon cancer cell lines, HCT116 and SW620, to assess their effects on a downstream apoptotic marker (i.e., cleaved caspase-3). We observed that bufalin increased the level of cleaved caspase-3 in scramble siRNA-transfected cells but not in Bax-knockdown cells (Supplementary Fig. S3A and S3B) or in Bak-knockdown cells (Supplementary Fig. S3C and S3D). Taken together, these results indicate that bufalin induces apoptosis via intrinsic pathway in colorectal cancer.

Because it is well established that inflammation is essential to carcinogen-induced tumorigenesis and colonic cancer, and DSS delivered in drinking water is believed to induce colonic inflammation through physical disruption of the mucosal barrier, therefore exposing innate immune cells in the lamina propria to bacteria and/or bacterial products (29), we assessed whether bufalin could inhibit colonic inflammation and thereby preventing colorectal cancer development. We noticed that chronic inflammation induced by 3% DSS was concomitant with notable colon shortening. Interestingly, this colon shortening was restored to the blank vehicle levels upon bufalin treatment (Fig. 1B), which indicated that bufalin might have acted as a powerful anti-inflammatory agent against DSS-induced colonic inflammation. Meanwhile, we observed a significant decrease in spleen weights, which reflects an attenuated inflammatory response in bufalin-treated mice (Fig. 5A).

To validate the molecular mechanisms underlying these observations, a panel of proinflammatory markers were measured by real-time PCR. An important inflammation-associated enzyme COX-2, which is overexpressed in 40% of adenomas and 85% of colorectal cancers and strongly participated in colonic tumorigenesis (30), was significantly downregulated in mRNA level from DSS-treated mice upon long-term administration of bufalin (Fig. 5B). Meanwhile, Western blotting analysis of protein levels of COX-2 in both tumor and tumor adjacent tissues (Fig. 5C) and COX-2 staining assay in tumor tissues (Fig. 5D) further validated this finding, indicating that bufalin might exert its chemopreventive action through ameliorating colonic inflammation and limiting the initial rate in the very early stage of carcinogenesis, thus leading to the reduced tumor burden in mice. This is consistent with the known importance of COX-2 in commonly triggered chronic inflammatory diseases and colitis-associated colon tumor development (30). Moreover, a series of colorectal inflammatory cytokines TNFα, IL6, IL1β, CXCL-1, CXCL-2, CXCL-5, and IL10 were also measured. In agreement with the known importance of these cytokines for colon tumorigenesis (31), we found that the mRNA levels of these cytokines significantly decreased in tumor cells from bufalin-treated mice (Fig. 5E), whereas analysis of IL10 levels failed to achieve statistical significance (Supplementary Fig. S1C).

Because bufalin is a natural inhibitor of Na⁺/K⁺-ATPase, we used two specific siRNA duplexes against Na⁺/K⁺-ATPase isoform α3, which is highly expressed in colorectal cancer, to confirm the role of bufalin in regulating COX-2 expression in HCT116 and SW620 cells (Supplementary Fig. S4A). We found that knockdown of Na⁺/K⁺-ATPase α3 downregulated the protein level of COX-2 in both cell lines (Supplementary Fig. S4B). In parallel, bufalin treatment also attenuated COX-2 expression in both cell lines (Supplementary Fig. S4C). Thus, these findings suggest that bufalin could suppress inflammation-driven tumorigenesis in colitis-associated colorectal cancer, at least in part, via inhibiting Na⁺/K⁺-ATPase.

Many tumor-promoting cytokines activate NF-κB signaling or these cytokines are activated via NF-κB. NF-κB, a set of transcription factors active and prominent in most tumors including colorectal cancer and colitis-associated tumorigenesis, constitute a link between inflammation and cancer (32). NF-κB targets inflammatory-based genes, IL8, TNFα, and COX-2. In this respect, we further investigated the molecular mechanisms behind the reduced proinflammatory cytokines in tumor cells from AOM/DSS–treated mice. The immunoblotting results showed that bufalin inhibited NF-κB activation in tumors by enhancing the activity of NF-κB inhibitor IκBα (nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor) and attenuating the phosphorylation of NF-κB activator IKKα (inhibitor of nuclear factor kappa-B kinase subunit alpha), which led to the suppressed activities of two NF-κB subunits, p50 and p65 (Fig. 6A). The treatment of bufalin (Fig. 6B) and two specific siRNAs against Na⁺/K⁺-ATPase α3 (Fig. 6C) on two human cancer cell lines further validated this finding.

Moreover, studies conducted over the past several years have pointed out that glycogen synthase kinase-3β (GSK-3β), a multi-tasking serine/threonine kinase, plays a vital role in inflammation and cancer (33, 34). In addition, the PI3K/AKT/mTOR/GSK-3β signaling pathway is broadly involved in the development and progression of chronic inflammatory diseases and tumorigenesis (35). Therefore, we assessed whether bufalin participates in the regulation of PI3K/AKT pathway in colitis-associated colorectal cancer. In our study, protein expression analysis revealed that phosphorylation levels of PI3K and Akt were significantly reduced in the bufalin-treated group, which led to the subsequently decreased phosphorylation of mTOR and enhanced the activity of GSK-3β (Fig. 6D). The treatment of bufalin (Fig. 6E) and two specific siRNAs against Na⁺/K⁺-ATPase α3 (Fig. 6F) on HCT116 and SW620 cells further confirmed this finding. Taken together, these results reveal that bufalin could strongly inhibit colonic inflammation by multi-tasking on several signaling pathways in tumor cells and thereby preventing colorectal cancer development (Fig. 6G).

Colorectal cancer is the third most common cause of cancer-related death. Surgery undoubtedly remains the most effective curative treatment for colorectal cancer. However, the risk of recurrence remains high (36). Chemoprevention is of significant importance to reduce cancerous morbidity and mortality with the use of nontoxic, naturally occurring or synthetic pharmacologic compounds to either block the initiation of cancers, reverse or arrest the progression and development of cancers. Therefore, the search for chemoprophylactic agents for the prevention of colorectal cancer is highly warranted. To our knowledge, several chemopreventive agents are commonly used in practice in recent decades, including aspirin (37), NSAIDs (38), metformin (39), calcium, vitamin D (40), and magnesium (41). Although drugs such as aspirin and COX-2 inhibitors have been proved to have astonishing effect in reducing the occurrence of colorectal cancer, long-term use of these agents was associated with increased risk of cardiovascular events, which raises the concern of high risk-benefit ratio to recommend these agents as colorectal cancer chemoprophylaxis.

In the current study, the chemopreventive action of bufalin in colorectal cancer development has been further investigated. Given to the inadequate dissolution of bufalin in saline solution and long-term and relatively frequent administration, we used intraperitoneal injection route to ensure the accurate dosage for each mouse during the administration period. Mice treated with AOM/DSS/bufalin had a large decrease in tumor incidence and much lower average tumor load with reduced average tumor size in colon compared with the mice treated with AOM/DSS alone. The findings from Apc^Min/+ mice had confirmed these observations. Although the tumors that arise from Apc^Min/+ mice were distinguishable from those that induced by AOM/DSS, two murine models still showed a remarkable significance of attenuated tumor growth upon bufalin treatment. This points to a possibility that bufalin modulates the initiation and promotion stages of tumorigenesis in an effective way and shed its potent tumor-suppressive action on both colitis-associated and genetically predisposed pathways of colorectal cancer. Congruently, these observations lead to the findings of suppressed proliferation and promoted apoptosis of tumor cells upon bufalin administration.

On the other hand, the link between chronic inflammation and colorectal cancer has long been appreciated. Patients with ulcerative colitis are at higher risk for developing colorectal cancer. It has been considered that cytokines generated by activated immune cells are important components in orchestrating the relationship between inflammation and cancer (31). A key player in inflammation is transcription factor NF-κB whose activity could be regulated by infectious agents and proinflammatory cytokines via IκB kinase (IKK) complex. Sustained activation of the NF-κB pathway also contributes to tumor development by increasing expression of growth factors (42). Although it remains unclear whether NF-κB is instrumental for tumor initiation, it is well accepted that inhibition of NF-κB activity in tumor cells can lead to increased sensitivity to chemo- and radiotherapy (43). In our study, bufalin was found to attenuate the activation of NF-κB signaling. This is consistent with the observation of reduced proinflammatory cytokines production and NF-κB downstream inflammatory factor COX-2 expression upon bufalin administration. Together, this evidence highlights a vital role of bufalin in suppressing colorectal tumorigenesis through alleviating chronic inflammation.

Mechanistic investigations into the potential interaction of proinflammatory and protumorigenic signals have further highlighted the relevance of COX-2 to alteration of bax/bcl-2 ratio, which has been reported to favor apoptosis in colon cancer (44). Regarding colorectal cancer, induction of apoptosis in tumor promotion represents a critical step for chemoprevention. Therefore, downregulation of Bcl-2 and Bcl-xL is of chemoprophylactic importance (45). Furthermore, proapoptotic members Bax and Bak act as gateways to the intrinsic pathway (46, 47). Together, they serve as upstream sentinels responding to specific death signals, leading to the activation of upstream initiator caspases, such as caspase-9, which is capable of autocatalytic activation and subsequently activates the downstream caspase-3. In this study, bufalin was found to repress the levels of Bcl-2 and Bcl-xL and enhance the levels of Bax, Bak, cleaved caspase-9, and 3. Together with knockdown experiments for Bax and Bak on two human colon cancer cell lines, HCT116 and SW620, we revealed that bufalin induces apoptosis via intrinsic pathway in colorectal cancer.

Overall, for any colorectal cancer prevention strategy to deliver significant effects, it has to be executed at the primary care level. Understanding the underlying mechanism of programmed cell death is the primary criterion for developing chemopreventive strategies. Therefore, by the virtue of two murine models of colon tumorigenesis, our study provides solid animal experimental evidence and characterizes the chemopreventive action of bufalin in primary health care settings and the resulting mechanistic insights into its significant proliferation-attenuating, apoptosis-promoting, and anti-inflammatory actions will ultimately provide a better understanding of the signaling pathways that can be therapeutically modulated. Moreover, solid preclinical evidence raised in this study reveals the relative safety of bufalin in in vivo settings and demonstrates for the first time of its chemopreventive role not only in inflammation-associated tumor promotion of colorectal cancer but also in genetically predisposed pathways of colorectal cancer in addition to its conventional role as a cardiac glycoside. Taken together, bufalin could be a promising chemopreventive agent that warrants further clinical evaluation.

For the further human clinical trial, the equivalent dose for human can be calculated by using Equation 1: HED (mg/kg = Animal NOAEL mg/kg) × (Weight_animal [kg]/Weight_human [kg])^(1–0.67). As to bufalin, the no-observed-adverse-effect-level (NOAEL) value for mice weighing approximately 20 g is 0.5 mg/kg. To calculate the starting dose for human studies, we used Equation 1. HED (mg/kg = 0.5 × (0.02/75)(0.33) = 0.033 mg/kg. Thus, for a 75 kg patient, the dose is 2.475 mg. This HED value should be further divided by a safety factor value of 10; therefore, the suggested initial dose in entry into human studies is 0.2475 mg.

No potential conflicts of interest were disclosed.

Conception and design: X. Sun, W.K.K. Wu, C.H.K. Cheng

Development of methodology: X. Sun, L. Zhang

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): X. Sun, T.T.H. Ng

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): X. Sun, T.T.H. Ng, K.W.Y. Sham, M.T.V. Chan, C.H.K. Cheng

Writing, review, and/or revision of the manuscript: X. Sun, W.K.K. Wu, C.H.K. Cheng

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): K.W.Y. Sham

Study supervision: W.K.K. Wu, C.H.K. Cheng

This work was supported by Health and Medical Research Fund (HMRF, 13140731, Hong Kong) and by the Direct Grant For Research (2014.1.076) from The Chinese University of Hong Kong. We would like to thank Core lab technicians from School of Biomedical Sciences of The Chinese University of Hong Kong for their excellent technical assistance.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.