This work was performed at Texas A&M University as a collaborative work within the Department of Small Animal Clinical Sciences, Department of Veterinary Physiology & Pharmacology (both part of the College of Veterinary Medicine), and the Center for Environmental, and Genetic Medicine in the Texas A&M Health Science Center.
Data from this project were presented at the 30th annual Veterinary Cancer Society Meeting, October 31, 2010 in San Diego, California
Corresponding author: Department of Small Animal Clinical Sciences, College of Veterinary Medicine, Texas A&M University, College Station, TX. e-mail:firstname.lastname@example.org
Tolfenamic acid (TA) is an NSAID currently under investigation as an anticancer agent in humans. TA induces proteosome-dependent degradation of transcription factors Sp 1, 3, and 4. These proteins are known to be overexpressed in many human cancers.
To evaluate the protein expression of Sps in canine tissue, and efficacy of TA against several canine tumor cell lines.
Six canine cell lines (2 osteosarcoma, 2 mammary carcinoma, 2 melanoma) were evaluated. Protein levels of Sp 1–4 and their downstream targets were evaluated using Western Blots. Cell survival and TUNEL assays were performed on cell lines, and Sp1 expression was evaluated on histologic samples from archived canine cases.
Six immortalized canine cancer cell lines derived from dogs were used. Archived tissue samples were also used.
Sps were highly expressed in all 6 cell lines and variably expressed in histologic tissues. TA decreased expression of Sps 1–4 in all cell lines. All of the downstream targets of Sps were inhibited in the cell lines. Variable Sp1 expression was identified in all histologic samples examined. TA significantly inhibited cell survival in all cell lines in a dose dependant fashion. The number of cells undergoing apoptosis was significantly increased (P < .05) in all cell lines after exposure to TA in a dose-dependent fashion.
Conclusions, and Clinical Importance
Tolfenamic acid is a potential anticancer NSAID and further investigation is needed to determine its usefulness in a clinical setting.
Specificity protein 1 (Sp1) is a sequence-specific DNA-binding protein that activates a diverse spectrum of mammalian, and viral genes. Sp1 is a nuclear protein that modulates gene transcription. There is evidence that Sp proteins could play a critical role in tumor development, growth, and metastasis. Sp1 protein is highly expressed in the nuclei of gastric tumor cells, whereas minimal levels are found in stromal or normal glandular cells within or surrounding the tumor. Survival times of patients with high Sp1 expression is significantly decreased compared with patients with weak to nondetectable Sp1 expression. Moreover, promoter analysis studies confirm that constitutive expression of many kinases and proteins is regulated by Sp1 interaction with GC-rich promoters in these genes. These data link increased Sp protein expression in tumors to upregulation of genes that are involved in tumor growth and metastasis.
Immortalized canine cell lines of osteosarcoma, mammary carcinoma, and melanoma were evaluated for Sp expression. These tumor types were evaluated because of their translational appeal from veterinary to human medicine and availability of archived tissue samples for comparison.
The importance of Sp transcription factors as drug targets is because of the pro-oncogenic activity of Sp-regulated proteins, which include VEGF, VEGF-receptor 1 (VEGFR1), c-MET, epidermal growth factor receptor (EGFR), bcl-2, and survivin.[3-6] Several of these Sp-regulated proteins are also an important oncogenic factors in canine tumors. Therefore, this study has focused on determining the expression of Sp1, Sp3, Sp4, and Sp-regulated proteins in canine tumor cell lines and the anticancer activity of the anti-inflammatory drug tolfenamic acid,1,2 which has been used for treatment of inflammation and cancer in humans and inflammation, pain, and fever in domestic animals.[7-13]
Tolfenamic acid inhibits cox2 with an affinity for cox2 approximately 15 times that of its affinity for cox1. However, TA appears to have a novel anticancer mechanism that differs from other traditional anticancer NSAIDs. A panel of NSAIDs (including ibuprofen, ketoprofen, naproxen, and others) and the cox2 selective drug celecoxib were compared with TA in their ability to inhibit Sp proteins. Ampiroxicam (a prodrug of piroxicam with fewer gastrointestinal adverse effects) did not inhibit Sp proteins, making it a good control drug for differentiating mechanisms that are cox related from mechanisms that are Sp related. TA was the best inhibitor of Sp proteins and was therefore chosen for further investigation. The growth inhibitor effects of TA on human pancreatic cancer cell lines is far superior to that of ampiroxicam, which was not significantly different from the DMSO control in 1 study.
Materials and Methods
The 2 osteosarcoma cell lines were established from clinical cases seen at the University of Wisconsin. The melanoma cell line 17CM98 was established at the University of Wisconsin as well. The CML-6M melanoma cell line and the CMT-12 mammary carcinoma cell line were established from clinical cases by Dr Lauren Wolfe, Auburn University, Auburn, AL.[16, 17] The REM mammary carcinoma cell line was provided by Dr David Argyle, University of Edinburgh, Edinburgh, UK.
Cells were cultured in adherence 75 cm flasks in R-10 complete media (RPMI 1640 media supplemented with fetal bovine serum, l-glutamine, sodium pyruvate, penicillin, and streptomycin) until cells reached 70% confluence. After washing the flasks with PBS the media was replaced with media containing DMSO (vehicle) or 25, 50, or 75 μM concentrations of tolfenamic acid in R-10. Cells were incubated for 48 hours at 37°C in 5% C02. Cells were then trypsinized and collected for protein extraction.
Total RNA was isolated from each cell line using Trizol3 according to the manufacturer's protocol and quantified using a biophotometer.4 RNA was cleaned up using the RNeasy Mini kit5 according to the manufacturer's protocol. RNA was reverse transcribed using Superscript II reverse transcriptase3 according to the manufacturer's protocol. PCR was carried out with the SYBR Green PCR Master Mix from PE Applied Biosystems on an ABI Prism 7700 Sequence Detection System6 using 0.5 Amol/L of each primer and 2 μL AL cDNA template in each 25 μL AL reaction. TATA binding protein (TBP) was used as an exogenous control to compare the relative amount of target gene in different samples.
The PCR profile was as follows: 1 cycle of 95°C for 10 minutes, then 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. The comparative CT method was used for relative quantitation of samples. Primers were purchased from Integrated DNA Technologies. The primers used are listed in Table 1.
Table 1. PCR primers for Sp genes
F 5′-TGCACAGGAGCCAAGAGTGAA-3′ R 5′-CACATCACAGCTCCCCACCA-3′
100% Homology with Canis familiaris
F 5′-ATGGGGGCAATGGTAATGGTGG-3′ R 5′-TCAGAACTTGCTGGTTCTGTAAG-3′
95% Homology with predicted Canis familiaris sequence
F 5′-ATGACTGCAGGCATTAATGCCG-3′ R 5′-TGTCTCTTCAGAAACAGGCGAC-3′
100% Homology with predicted Canis familiaris sequence
F 5′-ATGGCTACAGAAGGAGGGAAAAC-3′ R 5′-TTGACCAGGGGGTGGAAGAATTAC-3′
100% Homology with predicted Canis familiaris sequence
Cells were cultured as described previously until the flasks had reached 80% confluence. Cells were trypinized and collected. Cells were replaced at 10,000 cells per well in a 96 well plate with R-10 for 24 hours. The next day the media was replaced by R-10 with either DMSO (vehicle) or 25, 50, or 75 μM concentrations of tolfenamic acid. All samples were repeated in triplicate. Cells were incubated for an additional 72 hours at 37°C according to the manufacturer's instructions. MTS salts were added to each well according to the manufacturer's instructions and incubated at 37°C for 3 hours, the plate was imaged at 490 nm, and the absorbances for all wells were recorded.
Cells were grown to confluence in flasks as described above. Standard media was replaced by R-10 media containing either 25 or 50 μM concentrations of tolfenamic acid. Control flasks contained R-10 plus DMSO (vehicle). The authors elected not to use 75 μM concentrations because of the high rate of cell death seen at this concentration in the cell survival assays and the fact that this concentration does not likely relate to a safely achievable serum concentration in animals. Cells were incubated at 37°C for 24 hours, trypinized, and collected. Cells were suspended in PBS and 1–2 drops were placed on l-lysine coated slides and left under the cell culture hood to dry. All experiments were done in duplicate. Slides were prepared according to the manufacturer's protocol and visualized using a light microscope. Cells demonstrating nuclear staining were counted. Up to 500 cells per slide (1,000 cells per flask) were counted in at least 5 randomly chosen high power fields (40×) per slide and the numbers were recorded for statistical analysis and expressed as means ± SE for each treatment group.
Antibodies and Reagents
Cyclin D1, Sp3, Sp4, VEGF, bcl-2, erbB2, and EGFR antibodies were purchased from Santa Cruz Biotechnology.9 NFkB-p65 antibody was obtained from Abcam.10 Monoclonal β-actin antibody was purchased from Sigma. C-met antibody was purchased from Cell Signaling Technology.11 Sp1 antibody and Horseradish peroxidase substrate for Western blot analysis was obtained from Millipore.12 Survivin antibody was purchased from R&D Systems.13
Protein lysates were incubated for 3 minutes at 100°C before electrophoresis, and then separated on 7.5–12% SDS-PAGE at 120 V for 3–4 hours. Proteins were transferred onto polyvinylidene difluoride membranes by wet electroblotting in a buffer containing 25 mmol/L of Tris, 192 mmol/L of glycine, and 20% methanol for 1.5 hours at 180 mA. Membranes were blocked for 30 minutes with 5% TBST-Blotto (10 mmol/L of Tris-HCI, 150 mmol/L of NaCI (pH 8.0), 0.05% Triton X-100, and 5% nonfat dry milk) and incubated in fresh 5% TBST-Blotto with 1 : 500 primary antibody overnight with gentle shaking at 4°C. After washing with TBST for 10 minutes, the polyvinylidene difluoride membranes were incubated with secondary antibody (1 : 5000) in 5% TBST-Blotto for 3 hours with gentle shaking. The membrane was washed with TBST for 10 minutes, incubated with 4 mL of chemiluminescence substrate for 1 minute, and exposed to Kodak image station 4,000 mm Pro.14
Tissue sections were deparaffinized in xylene and treated with a graded series of alcohol and rehydrated in phosphate-buffered saline. Antigen retrieval was done using 10 mM sodium citrate (pH 6.0–6.2) and endogenous peroxidase was blocked by 3% hydrogen peroxide in methanol for 6 minutes. Slides were then incubated with blocking serum15 for 45 minutes. Samples were then incubated overnight with Sp1 antibody at 4°C. Sections were then washed in phosphate-buffered saline/Tween and then incubated with biotinylated secondary antibody followed by streptavidin. The brown staining specific for antibody binding was developed by exposing the avidin and biotinylated peroxidase complex to diaminobenzidine reagent15 and sections were then counterstained with hematoxylin.15
Data for the DeadEnd TUNEL and Cell Titer 96 assays were recorded and reported as mean ± SEM. One-way ANOVA16 was utilized to analyze the differences between groups. Dunnett's posthoc multiple comparison test was performed on all data sets to determine differences among control and tolfenamic acid treated groups. P values < .05 were considered as an indicator for the significant difference between study groups.
Isolated RNA from each cell line was examined for the presence of SP 1, 3, and 4 transcription factors. All SP transcription factors were identified in all cell lines (Fig 1). Ct values were recorded for all samples and the ΔCt was plotted in a graph as seen in Figure 1. The Ct value represents the 1st cycle at which the gene product was detected; therefore, the Ct value is inversely related to the expression levels of the gene analyzed. The ΔCt represents the log change difference between the expression of the housekeeping gene TBP and the SP gene expression. TBP is a housekeeping gene that was used as a control with a mean Ct value of 22.9. All SP genes were highly expressed, with SP1 being the highest expressed (mean Ct values between 21–31 cycles) in all cell lines. The mean number of cycles until detection for SP1 in all cell lines was similar to that of TBP, a house keeping gene. Mean Ct levels for SP3 ranged from 26 to 32 cycles in all cell lines and from 24 to 34 cycles for SP4. All PCR products were run out on a gel to confirm DNA presence in the correct location. All bands were present in all samples tested (see Fig 1 – SP1 and 3 gel images shown).
Figure 2A illustrates the effects of TA on growth of 2 canine osteosarcoma cell lines (UWOS1 and UWOS2) and significant growth inhibition was observed at concentrations of 25, 50, and 75 μM TA (P <.0001). TA also inhibited growth of CMT 12 (P <.02) and REM (P <.0001) canine mammary carcinoma cells (Fig 2B) and canine 17CM98 (P <.01) and CML6M (P <.0001) canine melanoma cells (Fig 2C). The cell growth inhibition experiments were determined after treatment for 72 hours and it was apparent from these data that the growth inhibitory effects of TA were variable in these canine cancer cell lines: UWOS1, CMT12, and CML6M cells were more responsive than the UWOS2, REM, and 17C98 cells. Nevertheless, with the exception of REM cells, significant growth inhibition by 25 μM TA was observed in all other cell lines.
Western Blot Analysis
Lysates from the canine cancer cell lines were analyzed by western immunoblots for expression of Sp1, Sp3, and Sp4 proteins. In untreated cells, the levels of Sp1, Sp3, and Sp4 relative to β-actin as a loading control were variable. The 2 osteosarcoma cell lines (UWOS1 and UWOS2) exhibited the highest relative expression of Sp1, Sp3, and Sp4; the mammary (CMT12 and REM) and melanoma (17CM98 and CML6M) cancer cells expressed similar levels of Sp proteins. After treatment of the 6 cell lines with 50 μM TA for 48 hours, there was a decrease in expression of Sp1, Sp3, and Sp4 proteins (Fig 3). There were some differences in the effects of TA on downregulation of Sp1, Sp3, and Sp4 proteins: the UWOS2 and REM24 cells exhibited the most resistance, particularly with respect to downregulation of the low molecular weight band for Sp3 (Fig 3).
RNA interference studies using individual small inhibitory RNAs against Sp1, Sp3, and Sp4 have confirmed that in human cancer cell lines several pro-oncogenic proteins are Sp-regulated.[3, 4] The effects of TA on their expression in canine cancer cell lines were also investigated. The results (Fig 4) demonstrate both similarities and differences in basal expression of Sp-regulated proteins among the 6 cell lines; relatively high immunostaining was observed for VEGF and the p65 subunit of NFkβ in all cell lines; relatively low staining intensity was observed for survivin, and cyclin D1, bcl-2, and erbB-2. Staining for EGFR was both relatively high (UWOS1, CMT12, and REM) and low (17CM98, UWOS2, and CML6M0), and high c-MET staining was observed in REM cells with relatively low to nondetectable staining in the other cell lines. In cells treated with TA, the effects of this compound on down-regulation of Sp1, Sp3, and Sp4 were similar to those observed for Sp-regulated proteins although there were some cell context-dependent differences. For example, VEGF and the p65 subunit of NFkβ were decreased by TA in all cell lines; however, differences in treated versus untreated were the lowest in UWOS2 cells and this correlated with the relative-resistance of this cell line to TA-induced Sp down-regulation at the concentrations used (Fig 3). TA decreased the high c-MET, p65, and EGFR immunostaining in REM cells after treatment with TA for 48 hours.
The western blot data confirmed that TA decreased expression of protein products required for cell growth, angiogenesis, and survival and the effects of TA on survival were further investigated using a TUNEL assay to measure apoptosis. The results show that 25 and/or 50 μM TA induced increased TUNEL staining in canine osteosarcoma (UWOS1 and UWOS2- P <.0001 at 50 μM) (Fig 5A), melanoma (CMCL6 and 17CM98- P <.0001 at 25 and 50 μM) (Fig 5B), and mammary (CMT12- P <.0001 at 25 and 50 μM and REM P <.0001 at 50 μM) (Fig 5C) cancer cells.
The relative staining of Sp1 protein in a series of canine tumors was also determined using sections which contain both tumor and nontumor tissue (Fig 6). Sp1 was chosen because it was determined to be the most reliably expressed Sp transcription factor and was expressed at the highest level in the cell lines examined. Immunostaining of these sections from canine osteosarcomas, melanomas, and mammary tumors showed increased Sp1 staining in tumor versus nontumor tissues (Fig 6). Sp1 staining was primarily nuclear and the differences in staining in tumor and nontumor tissues were more evident for the osteosarcomas and melanomas than for the mammary tumors (Fig 6). These data showing high levels of Sp1 staining in canine tumors are consistent with previous reports in human tumors and cancer cell lines.[4, 15]
The potential for development of drugs that repress Sp1, Sp3, and Sp4 transcription factors for treating cancer in dogs depends on the relative expression of these proteins in canine tumors. Results illustrated in Fig 3 show that in canine melanoma, osteosarcoma, and mammary cancer cell lines, Sp1, Sp3, and Sp4 are expressed with the lowest levels in 17CM98 and CML6M melanoma cells, the highest expression was observed in the 2 osteosarcoma cell lines (UWOS1 and UWOS2) and moderate expression in the canine mammary cancer cells (CMT12 and REM). With the exception of 17CM98 cells, staining of Sp1 and Sp3 bands were higher than observed for Sp4; however, the absolute levels of all 3 Sp transcription factors remain unknown. A direct comparison of Sp1 expression in cancer versus noncancer tissues was determined for several canine tumors (Fig 6) and there was enhanced immunostaining of Sp1 in tumor versus nontumor tissue; this was particularly evident for the osteosarcomas and melanomas.
The relatively high expression of Sp1, Sp3, and Sp4 in the canine cancer cells and tumors (Figs 3 and 6) is comparable to that observed in human tumors and cancer cell lines. Therefore, we investigated the anticancer activity of TA, an anti-inflammatory drug used for treatment of migraine headaches in humans and for arthritis and inflammatory conditions in humans and domestic animals.[7-13] The results provided in this study demonstrate a potential clinical use for TA in veterinary oncology among various tumors.
Previous studies with canine mammary tumors have identified VEGF and p65 expression as biomarkers; VEGF overexpression has also been observed in malignant melanomas and osteosarcomas.[20, 21] In this study, we examined expression of these and other Sp-regulated proteins, including cyclin D1, EGFR, bcl-2, erbB2, c-MET, and survivin in the canine cancer cell lines. RNA interference studies in human cell lines showed that knockdown of Sp1, Sp3, Sp4, or their combination resulted in decreased expression of these protein products.[1-6] The relative expression of the Sp-regulated proteins in the canine cancer cell lines was variable; however, with few exceptions TA-mediated downregulation of Sp1, Sp3, and Sp4 was accompanied by decreased expression of Sp-regulated protein products. The effect of TA on expression of VEGF and p65 in UWOS2 cells was less than observed in other cell lines. This correlated the effects of 50 μM TA on expression of Sp1, Sp3, and Sp4 proteins which was also lower than observed in the other cancer cells.
Results of this study demonstrate that Sp1, Sp3, and Sp4 transcription factors are highly expressed in canine cancer cells and using Sp1 as a marker, immunostaining of canine tumors confirmed higher expression of Sp1 in tumor versus nontumor tissue. The anti-inflammatory drug TA, which is a highly effective anticancer agent in human cancer cells and tumors (xenografts), appears to also have efficacy in canine cancer cell lines and the anticancer activity in human and canine cancer cells may be due in part to the downregulation of Sp transcription factors and Sp-regulated proteins. TA is approved for dogs in Canada, South America, and Australia for analgesia and fever relief. Based on published pharmacokinetic data and our benchtop results, we anticipate that oral doses exceeding 4mg/kg will be necessary to create serum levels similar to the 25–50 μm concentrations used here.[10, 12, 13] Further in vivo data will be needed to determine what anti-neoplastic doses will be effective. Anti-inflammatory doses of TA in canine patients have an average serum level of approximately 4 μg/mL. Safety studies have demonstrated no adverse events at twice this dose for a total of 10 consecutive days; however, no long-term studies at higher doses have been performed in the dog.[10, 12, 13] Further animal studies are needed to determine if this dose level will be tolerated for long-term administration in the dog. Currently, we are planning clinical studies to determine the effective anticancer dose of TA as novel mechanism-based drug for treating canine tumors.
The main limitations of this study are related to the in vitro nature of its design. Cell culture techniques may predict, but do not mirror, events that occur in vivo in the cancer patient. In addition, the Sp antibodies used in this experiment were developed as anti-human antibodies. Anti-Sp antibodies are not commercially available for the dog. However, based on the NCBI Blast database the predicted protein sequences for the dog are 92% (Sp3), 99% (Sp1), and 100% (Sp4) similar to those known for the human. The authors feel that this combined with the clean bands visualized on the western blot analysis make it unlikely that the antibodies are binding nonspecific proteins. Finally, the authors have only evaluated the Sp transcription factor pathway as a means for downregulation of specific proteins; however, this is only 1 mechanism of TA. Other factors, such as inhibition of heat shock proteins, could also play a role in the downregulation of specific protein products mentioned in this article.
This work was not supported by any grants or outside funding.
Tofedine is approved in Europe, Canada, Australia, and South America for multiple veterinary species as a NSAID
Crystalline TA (>99%) was purchased from LKT Laboratories, St Paul MN
Invitrogen, Grand Island, NY
Eppendorff, Hamberg, Germany
Qiagen, Valencia, CA
AB Applied Biosystems, Carlsbad, CA
Cell Titer 96 Assay, Promega, Madison, WI
DeadEnd TUNEL Assay, Promega
Santa Cruz, CA
Carestreamhealth, Woodbridge, CT
Vecstatin ABC Elite kit, Vector Laboratories, Burlingame, CA