Abstract
Ursolic acid is a key active compound present in many medicinal herbs that have been widely used in traditional Chinese medicine for the clinical treatment of various cancers. However, the precise mechanisms of its antitumor activity have been poorly understood. To identify the cellular targets of ursolic acid, two-dimensional gel electrophoresis combined with mass spectrometry was performed in this study, which identified 15 proteins with significantly altered levels in protein expression. This demonstrated that ursolic acid–induced cytotoxicity in colorectal cancer cells involves dysregulation in protein folding, signal transduction, cell proliferation, cell cycle, and apoptosis. Corresponding protein regulation was also confirmed by Western blotting. Furthermore, the study of functional association between these 15 proteins revealed that 10 were closely related in a protein–protein interaction network, whereby the proteins either had a direct interaction with each other or were associated via only one intermediary protein. In this instance, the ATP5B/CALR/HSP90B1/HSPB1/HSPD1-signaling network was revealed as the predominant target which was associated with the majority of the observed protein–protein interactions. As a result, the identified targets may be useful in explaining the anticancer mechanisms of ursolic acid and as potential targets for colorectal cancer therapy.
Introduction
Colorectal cancer (CRC) is one of the most prevalent and lethal cancers worldwide, and its incidence and mortality rates have sharply increased in recent years. 1 Currently, the main conventional methods of treatment for CRC include surgery, radiotherapy, and chemotherapy, in combination with 5-fluorouracil (5-FU), vincristine, and doxorubicin-based drugs. Unfortunately, 30%–50% of all patients with CRC have developed metastases, resulting in a very poor prognosis and a 5-year survival rate of less than 10%.2,3 The main reasons for the high failure rates during treatment of patients with CRC include severe toxicity during anticancer chemotherapy and multi-drug resistance. Therefore, the development of novel and more effective anticancer agents which can be utilized in the treatment of CRC is urgently required.4,5
In recent years, the use of Chinese medicinal herbs for cancer treatment has received greater recognition due to its antitumor activity, providing new therapeutic strategies for the treatment of cancer. Ursolic acid (UA), belonging to the class of pentacyclic triterpenoids, is a common chemical component that is found within a variety of medicinal herbs. Previous studies have shown that UA has inhibitive effects in many cancers, including CRC, breast cancer, lung cancer, prostate cancer, endometrial cancer, melanoma, and leukemia cells due to its effect on a variety of biological pathways, such as during oncogene expression, 6 cell cycle regulation,7–9 apoptosis,7,8,10–12 tumorigenesis,13–15 metastasis, 16 and drug resistance. 17 These properties give UA the potential to become an effective anticancer agent. However, because traditional analytical means have provided only a limited understanding of the underlying mechanisms, the anticancer properties of UA and its interactions with target-related protein during CRC are still not fully understood.
Following recent high-throughput advances in technology, two-dimensional polyacrylamide gel electrophoresis (2-DE) has been commonly used to determine the heterogeneity of protein expressions in cells under different conditions. 2-DE combined with matrix-assisted laser desorption ionization–time-of-flight mass spectrometry (MALDI-TOF-MS) is often applied in comparative proteomics in order to identify potential anticancer mechanisms. In the present study, we analyzed the UA-mediated response on the proliferation and apoptosis of human colorectal cells. Subsequently, we performed a comprehensive analysis of the molecular targets of UA using a proteomics approach, in order to identify specific protein expressions following exposure to UA in colorectal cells. We conducted 2-DE analysis, where differentially expressed proteins were identified using MALDI-TOF-MS/MS, and the corresponding protein regulation was further confirmed by Western blot analysis. Moreover, a comprehensive network analysis was conducted to analyze the functional protein–protein associations involved.
Materials and methods
Chemicals and reagents
Leibovitz’s L-15 medium, RPMI-1640 medium, Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS), penicillin–streptomycin, trypsin–ethylenediaminetetraacetic acid (EDTA), and TRIzol reagent were purchased from Life Technologies Corp. (Grand Island, NY, USA). BCA Protein Assay kit was purchased from Pierce (Rockford, IL, USA). Reagents for 2-DE were purchased from Bio-Rad (Bio-Rad, Hercules, CA, USA). Antibodies against calreticulin (CALR), cytokeratin 19, 14-3-3 sigma, heat shock protein beta-1 (HSPB1), prohibitin, and β-actin were purchased from Abcam (Hong Kong, China). All other chemicals, unless otherwise stated, were obtained from Sigma-Aldrich (St Louis, MO, USA).
Cell culture
Human colon carcinoma HT-29, HCT-8, and SW620 cells were obtained from the Cell Bank of Chinese Academy of Sciences (Shanghai, China). HT-29 cells were grown in DMEM medium; HCT-8 cells were cultured in RPMI-1640 medium and SW620 were grown in L-15 medium. All cells were supplemented with 10% FBS, 100 U/mL penicillin, and 100 mg/mL streptomycin, and maintained in a humidified incubator at 37°C with 5% CO2.
Evaluation of cell viability by Cell Counting Kit 8 assay
Cell viability was assessed by the Cell Counting Kit 8 (CCK-8) assay. HT-29, HCT-8, and SW620 cells were seeded into 96-well plates at a density of 1 × 104 cells/well in 0.1 mL medium. The cells were treated with various concentrations (10, 20, 40, 80, 100 µM) of UA, with five replicate wells each. After treatment with UA for 24, 48, or 72 h, 10 µL of CCK-8 reagents (Dojindo Laboratories, Japan) was added to each well and incubated at 37°C for 2.5 h. Absorbance was measured at 450 nm in a ELISA reader (BioTek, Model ELX800, USA). Each experiment was repeated a total of three times.
Cell apoptosis assay
The cell apoptosis was determined with Hoechst 33258 staining. HT-29 cells were treated with 20, 40, and 80 µM UA. After 24 h, cells were washed three times with phosphate buffered saline (PBS) and stained with 1 µg/mL of Hoechst 33258 nuclear dye (Beyotime, China) for 30 min at 37°C. Cells were imaged under fluorescence microscopy (Nikon Ti-S, Japan) to assess chromatin condensation and segregation.
Sample preparation and 2-DE
Cells were cultured at a density of 1 × 106 cells per 10 cm plate overnight prior to treatment with dimethyl sulfoxide (DMSO) or 40 µM UA for 24 h. Subsequently, cells were washed three times with ice-cold PBS, collected using cell scrapers, and subsequently centrifuged for 10 min at 2000g. The resulting supernatant was discarded and cell pellets were solubilized in 500 µL lysis buffer containing 7 M urea, 2 M thiourea, 4% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), 1% dithiothreitol (DTT), 0.2% biolyte 3/10, and 1% protease inhibitor cocktail. The cells were homogenized using ultrasonication (10 strokes, low amplitude) on ice. The lysed cells were centrifuged at 15,000g for 30 min at 4°C and the supernatant containing the solubilized proteins was stored at 80°C until further use. Protein samples from at least three independent experiments were collected for the 2-DE assay and the total protein concentration was measured using 2-D Quant Kit (GE Healthcare). In the first dimension of the 2-DE, protein samples (800 µg) were applied to ReadyStrip immobilized pH gradient (IPG) strip (17 cm, pH 4-7; Bio-Rad). The strips were placed into a Protean isoelectric focusing (IEF) cell (Bio-Rad), rehydrated at 50 V for 12 h and subsequently the proteins were separated based on their pI under the following conditions: (1) 250 V, 0.5 h; (2) 500 V, 0.5 h; (2) 1000 V, 1 h; (3) 4000 V, 1 h; (4) 8000 V, 3 h; and (5) 65000 VH. Following IEF separation, equilibration was performed by incubating with equilibration buffers I and II (37.5 mM Tris-Cl, pH 8.8, 20% glycerol, 2% sodium dodecyl sulfate (SDS), and 6 M urea, with 2% DTT in buffer I and 2.5% iodoacetamide in buffer II, respectively). The second dimension electrophoresis was run on 12% SDS-PAGE gels in a Protean II xi cell system (Bio-Rad) until the bromophenol blue dye reached the bottom of the gels. Proteins were visualized using Coomassie Brilliant Blue R-250 stain.
Gel analysis
The stained 2-DE gels were scanned with UMAX PowerLook 2100XL scanner. Image analysis and 2-DE gel proteome database management were performed using PDQuest software 8.0 (Bio-Rad). The intensity of each protein spot was normalized to the total intensity of the entire gel image. Quantitative analysis was performed using the Student’s t test. Protein spots with significant differential expression (P < 0.05), essentially representing 1.5-fold or greater change in intensity between the control and UA-treated groups, were selected for further identification by MALDI-TOF MS/MS.
In-gel digestion
Protein spots were automatically excised from the 2-DE gels using EXQuest spot cutter (Bio-Rad). Gel pieces were destained in 100 mM NH4HCO3/30% acrylonitrile (can) for 10 min, dehydrated in 100% ACN for 10 min, and completely dried in a SpeedVac device (Concentrator Plus, Eppendorf, Hamburg, Germany). Dried pieces were allowed to swell for 30 min on ice in a digestion buffer containing 25 mM ammonium bicarbonate and 2.5–10 ng/µL trypsin (Promega). Following 30 min incubation, the gels were digested for 12 h at 37°C. Peptides were then extracted twice using 0.1% trifluoroacetic acid (TFA) in 50% ACN. The resulting extracts were concentrated and stored at −20°C until further analysis.
MALDI-TOF/TOF MS and database searches
Samples were spotted onto a MALDI target plate with an equal volume of matrix solution, containing 5 mg/mL 4-hydroxy-α-cyanocinnamic acid in 50% ACN and 0.1% TFA. An AutoFlex speed MALDI-TOF/TOF MS (Bruker Daltonics) was used with a mass accuracy of 50 ppm after external calibration. The samples were analyzed in MS mode (for generation of peptide mass fingerprints) as well as in TOF/TOF mode (for fragmentation analysis of the highest intensity peaks). MS spectra were transformed into peak lists using the software flexAnalysis version 3.0 (Bruker Daltonics). The peak lists of the MS and MS/MS spectra were merged using BioTools version 3.0 software (Bruker Daltonics). Proteins were identified using MASCOT (Matrix Science) protein identification software against the Swiss-Prot database (Swiss Institute of Bioinformatics). During data search, maximum allowance was set at one missed cleavage per peptide. MS/MS tolerance of 0.4 Da and a mass tolerance of 50 ppm were also used according to the predefined optimization protocol. Carbamidomethylation for cysteine, oxidation for methionine, and other variants were also taken into consideration. Acceptance criterion for protein identification was a score above the MASCOT identity threshold (set at 95% confidence level) for peptide mass fingerprints.
Western blot analysis
HT-29 cells were seeded into 6-well plates at a density of 2 × 105 cells per well prior to treatment with different concentrations of UA for 24 h. Subsequently, cells were washed twice with ice-cold PBS and lysed in RIPA buffer containing protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany). Lysates were centrifuged for 15 min at 12,000g and supernatants were collected for further analysis. The protein concentrations of the cell lysates were measured using BCA quantification assay kit (Pierce, Rockford, IL, USA). Proteins (50 µg) were separated using 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes with a 0.45-µm pore size (IPVH00010; Millipore, Billerica, MA, USA). The membranes were incubated with primary antibodies overnight at 4°C. Primary antibodies used include Survivin, p53, CALR, 14-3-3 sigma, HSPB1, prohibitin, and β-actin diluted to 1:1000 in immunoblot buffer (tris-buffered saline (TBS) containing 0.05% Tween-20 (TBST) and 5% non-fat dry milk). Membranes were washed three times with TBST and incubated with secondary antibody horseradish peroxidase (HRP)-conjugated anti-mouse or anti-rabbit IgG (1:1,000 dilution) for 1 h at real time (RT). After washing, the immunoblots were incubated with Clarity Western ECL Substrate (Bio-Rad, Hercules, CA, USA) for 1 min prior to band detection using the ChemiDoc XRS+ System (Bio-Rad, Hercules, CA, USA). The pixel intensities of the immunoreactive bands were quantified using ImageLab software 8.0 (Bio-Rad). β-actin was used as the internal control.
Interaction network
Functional partnerships between proteins are a fundamental aspect of cell regulation. A proteome-scale interaction network of the differentially expressed proteins that were identified in the present search was derived from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database.
Results
UA-induced apoptosis in CRC cells
We treated various human colon carcinoma cell lines (HT-29, HCT-8, and SW620 cells) with different concentrations (0, 10, 20, 40, 80, and 100 µM) of UA for 24, 48, or 72 h and measured the cell viability using CCK-8 assay. UA inhibited the viability of all three CRC cell lines in a dose-dependent and time-dependent manner, although with varying degrees of sensitivity (Figure 1). It exhibited the strongest antitumor activity in HT-29 cells, with an IC50 of ~45 µM after 24 h. Therefore, HT-29 cells were selected for subsequent analysis to further understand the tumoricidal mechanisms of UA.
The growth-inhibitory effects of UA on colorectal cancer cells determined by CCK-8 assay. (a) Structure of UA. Effects of UA on viability of different cancer cells. (b) HT-29. (c) HCT-8. (d) SW620 cells were treated with UA for 24 h, 48 h, and 72 h, respectively. The data were normalized to the viability of untreated cells. Data were expressed as the average ± standard deviation.
Cell apoptosis was evaluated by observing nuclear morphological changes following staining of cell nuclei with DNA-binding dye Hoechst33258. UA-treated cells showed condensed chromatin and fragmented nuclear morphology that are typical of apoptosis morphology, whereas the untreated cell nuclei showed homogenous and less intense staining than UA-treated cells, suggesting that UA promoted HT-29 cell apoptosis (Figure 2(a)).
UA promotes the apoptosis of HT-29 cells. (a) HT-29 cells were treated with various concentrations of UA for 24 h. Cell apoptosis was determined by Hoechst staining. (b) The expression of survivin and p53 was detected by Western blot. Images are representative of three independent experiments.
In order to verify UA-induced HT-29 cell toxicity, we also examined changes in cell survival and apoptosis markers in HT-29 cells treated with 0–40 µM of UA for 24 h. As shown in Figure 2(b), the apoptotic marker p53 was up-regulated and that of survivin was down-regulated in a dose-dependent manner by UA treatment.
2-DE and MS analysis of UA-induced changes in the protein expression of CRC cells
To further investigate the underlying mechanisms of UA-induced cytotoxicity in CRC cells, we examined the protein profiles of UA-treated and control cells using comparative proteomics analysis. Representative two-dimensional gel images of UA-treated and control cells were shown in Figure 3, and their respective proteome maps were compared with PDQuest software to identify any variations in protein spots. Following UA treatment, protein spots which showed significant differential expression (P < 0.05), essentially representing 1.5-fold or greater change in intensity from the three replicate gels, were scored. We identified a total of 15 proteins which showed significant changes between UA-treated and control groups. Among these, a total of 3 proteins had up-regulated expression, and the remaining 12 proteins had down-regulated expression following UA treatment as determined by spot densitometry. These differentially expressed spots were then excised using EXQuest spot cutter and subjected to trypsin digestion in a 96-well plate, and the resulting peptides were analyzed using MALDI-TOF/TOF mass spectrometry. Differentially expressed protein spots were analyzed by peptide mass fingerprinting (PMF) with up to three MS/MS performed on each protein spot using automatic mode. The MS and MS/MS spectra of representative proteins were shown in Figure 4. Using the data collected from MS, peptide searches were performed using MASCOT protein identification software and National Center for Biotechnology Information (NCBI) database for identification of each protein spot. The results of MS/MS analysis, including the Mr/PI and sequence coverage, were summarized in Table 1.
Results of 2-DE analysis of cellular protein expression profiles of control cells and cells treated with 40 µM UA. Representative 2-DE gel images of (a) control and (b) UA-treated groups. Differentially expressed spots are shown by the boxes.
Peptide mass fingerprint of the tryptic digest of differential protein spot (a) 1 (calreticulin), (b) 2 (HSP90B1 endoplasmin), and (c) 11 (14-3-3 sigma protein) of 40 µM UA-treated HT-29 cells.
MS identification of differentially expressed protein spots in UA-treated HT-29 cells.
Confirmation of differentially expressed proteins by Western blotting
To verify the changes in protein expressions following UA treatment, Western blot assays were performed to assess the expression levels of the five protein targets. Consistent with the proteomics results, we determined that CALR and HSPB1 had down-regulated expression, whereas, 14-3-3 sigma and prohibitin had up-regulated expression following UA treatment in HT-29 and HCT-8 CRC cells (Figure 5).
The candidate proteins (calreticulin, 14-3-3 sigma, HSPB1, prohibitin) were validated with the effect of UA (40 µM) on HT-29 and HCT-8 cells using a Western blot analysis.
Possible protein–protein interaction network of UA
We established a protein–protein interaction (PPI) network based on the 15 possible target-related proteins of UA as determined previously using 2-DE analysis. Of these 15 proteins, a total of 10 can be linked into a protein interaction network, whereby the proteins either had a direct interaction with each other or were associated via only one intermediary protein (Figure 6). In this case, we revealed that the ATP5B/CALR/HSP90B1/HSPB1/HSPD1-signaling network was associated with the majority of the observed PPIs, as illustrated in Table 2. These results suggested that ATP5B/CALR/HSP90B1/HSPB1/HSPD1 might be potential targets involved in the anticancer mechanisms of UA.
The constructed minimum PPI network. The red dots illustrate up-regulated proteins; the blue dots are down-regulated proteins identified from experiments. Proteins in the network are interacting with each other via intermediate partners (shown in black) from known PPI information. The full lines mean inhibition interaction. The imaginary lines mean activation interaction. The expanded network constructed by 31 identified proteins.
Betweenness centrality and degree of the key proteins in the interaction network.
Discussion
The anti-proliferative activity of UA has previously been demonstrated in various cancer cell lines. Consistent with previous reports, our study showed that UA induced cell-growth inhibition and apoptosis in CRC cells. We had previously demonstrated that UA could inhibit cell proliferation and CRC angiogenesis, as well as promote CRC cell apoptosis via modulation of multiple signaling pathways, such as STAT3, ERK, JNK, and p38 MAPK.15,18 Although progress has been made in defining the signaling and biochemical events involved in UA-treated cancer cells, the precise mechanisms underlying the cellular targets of UA remain largely unknown. Therefore, this study provides a detailed analysis into the growth inhibitory effects of UA on cell proliferation, as well as elucidating the possible underlying mechanism of its cytotoxicity in human CRC cells. CCK-8 assays showed that UA inhibited cell growth in both a dose- and time-dependent manner in CRC cells. Colony formation assays, hoechst 33258 staining, and the expression of apoptosis-related protein survivin and p53 demonstrated that UA inhibited proliferation and induced apoptosis in HT-29 cells. Taken together, these results demonstrated that UA can act as a powerful anti-proliferative agent on CRC cells.
To further understand the pharmacological effects of UA and investigate the molecular targets involved, we performed proteomics analysis of UA-treated HT-29 cells in order to identify the differentially expressed proteins using 2-DE and MALDI-TOF/MS. We revealed a total of 15 protein spots with significant differential expression following UA treatment. We identified that the ATP5B/CALR/HSP90B1/HSPB1/HSPD1 network was associated with the majority of the observed PPIs, suggesting that these might be the critical targets involved in the anticancer mechanisms of UA, as well as being potential targets for CRC therapy.
The ability of UA-induced cytotoxicity in CRC cells can be attributed to the regulation of proteins involved in signal transduction (14-3-3 protein sigma), gene expression (CALR, NPM1, prohibitin), cell proliferation, cell cycle, and apoptosis (cytokeratin 19, HSPB1, HSPA9, HSPD1), as well as nucleotide metabolism (ATP5B, TUBB2C).
Heat shock proteins (HSPs) are a group of ubiquitous molecules in cells which act as molecular chaperones during protein folding, assembly, and trafficking under cellular stress, such as carcinogenesis. 19 HSPs are over-expressed in tumor tissues and their expression is often correlated with increased tumor aggressiveness and poorer disease prognosis.20,21 There is also a significant correlation between the expression of HSPs and the increased resistance of malignant cells to chemotherapy. Therefore, a group of highly abundant HSPs can be utilized as potential markers to evaluate the progression of carcinomas. Likewise, the inhibition of HSP90, HSP70, HSP60, and/or HSP27 has emerged as a novel strategy for cancer therapy. In this study, we identified four members of the HSP family (HSP90B1, HSPB1, HSPD1, and HSPA9), which were down-regulated following treatment with UA. HSPB1 (heat shock protein beta-1) also known as HSP27, plays important roles in cytoprotection, apoptosis, and maintaining normal protein structure during environmental stress. 22 A high level of HSPB1 is a common indicator in CRC, and is usually associated with poorer disease prognosis.22–24 It is also involved in the regulation of cellular stress response, and as an inhibitor of apoptosis. In fact, the switch between survival and apoptosis in adenocarcinoma cells has been attributed to HSPB1/HSP27. 24 HSPA9, another member of the HSP family, is composed of 679 amino acids and is localized in plasma membranes, cytoplasm, or mitochondria.25,26 HSPA9 is well recognized as an anti-apoptotic factor due to its interaction with the tumor-suppressor protein p53, and its association with various carcinomas has also been reported.26–28 In addition, HSPD1, also known as HSP60, is also involved in cell proliferation and apoptosis. When HSP60 is down-regulated, Bax is transferred from the cytoplasm to the mitochondria, resulting in the release of cytochrome c, thereby leading to cell apoptosis.29,30 Consistent with these findings, our study demonstrated a significant reduction in HSPB1 expression following UA treatment for 24 h, which was associated with cellular apoptosis.
CALR is a multi-functional endoplasmic reticulum protein, involved in a variety of cellular processes including calcium homeostasis, cell adhesion, cancer formation, and progression. 31 CALR is involved in the regulation of cell proliferation, migration, and invasion of prostate cancer cells via MEK/ERK pathway. Knockdown of CALR can significantly decrease p-ERK expression and cell chemoresistance, independent of activated p53 and caspase-3-related apoptosis. The interactions between CALR and MEK/ERK pathway could provide a basis for gene-targeted cancer chemotherapy.31,32 Our present study demonstrated the potential role of CALR in cancer development and progression as a potential tumor suppressor. The increase in CALR expression in UA-treated CRC cells is a likely contributor to the anticancer activity of UA.
The protein 14-3-3 sigma belongs to the 14-3-3 protein family,33–35 a class of highly conserved proteins involved in the regulation of signal transduction pathways, cellular proliferation, apoptosis, adhesion, differentiation, and survival. Among these, 14-3-3 sigma is the isoform most directly linked to cancer. There is evidence that 14-3-3 sigma acts as a tumor-suppressor gene and that its inactivation is crucial for tumorigenesis.36,37 Consistent with previous studies, we demonstrated that up-regulation of 14-3-3 sigma results in the inhibition of tumor cell growth.
Prohibitin is an intracellular protein with anti-proliferative activity, which was also significantly up-regulated following treatment with UA. Previous studies have demonstrated that prohibitin obstructs DNA synthesis and inhibits cell growth. 38 Therefore, our study showed that up-regulation of prohibitin was likely responsible for UA-induced cytotoxicity in CRC cells.
Based on the results of the present study, we can predict the possible signaling network by which UA induced apoptosis, as shown in Figure 6. In addition, we showed that ATP5B/CALR/HSP90B1/HSPB1/HSPD1 might be potential targets involved in the anticancer mechanisms of UA. As a multitarget compound, UA is likely involved in the activation of various signaling cascades within cells. Therefore, finding these potential factors can contribute to understanding the anticancer mechanisms of UA and form the basis for improved CRC therapy.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The present study was sponsored by the National Natural Science Foundation of China (81403390) and the University Distinguished Young Research Talent Training Program of Fujian Province.