BB-Cl-Amidine as a novel therapeutic for canine and feline mammary cancer via activation of the endoplasmic reticulum stress pathway
Mammary cancer is highly prevalent in dogs and cats and results in a poor prognosis due to critically lacking viable treatment options. Recent human and mouse studies have suggested that inhibiting peptidyl arginine deiminase enzymes (PAD) may be a novel breast cancer therapy. Based on the similarities between human breast cancer and mammary cancer in dogs and cats, we hypothesized that PAD inhibitors would also be an effective treatment for mammary cancer in these animals.
Canine and feline mammary cancer cell lines were treated with BB-Cl-Amidine (BB-CLA) and evaluated for viability and tumorigenicity. Endoplasmic reticulum stress was tested by western blot, immunofluorescence, and quantitative reverse transcriptase polymerase chain reaction (qRT-PCR). Canine and feline mammary cancer xenograft models were created using NOD scid gamma (NSG) mice, and were treated with BB-CLA for two weeks.
We found that BB-CLA reduced viability and tumorigenicity of canine and feline mammary cancer cell lines in vitro. Additionally, we demonstrated that BB-CLA activates the endoplasmic reticulum stress pathway in these cells by downregulating 78 kDa Glucose-regulated Protein (GRP78), a potential target in breast cancer for molecular therapy, and upregulating the downstream target gene DNA Damage Inducible Transcript 3 (DDIT3). Finally, we established a mouse xenograft model of both canine and feline mammary cancer in which we preliminarily tested the effects of BB-CLA in vivo.
We propose that our established mouse xenograft models will be useful for the study of mammary cancer in dogs and cats, and furthermore, that BB-CLA has potential as a novel therapeutic for mammary cancer in these species.
KeywordsPeptidyl arginine deiminase Endoplasmic reticulum stress Mammary cancer BB-cl-Amidine
Activating transcription factor 4
Canine mammary epithelial cells
DNA Damage Inducible Transcript 3
DNA Damage Inducible Transcript 4
Estrogen Receptor alpha
Feline mammary epithelial cells
78 kDa Glucose-regulated Protein
Peptyidyl Arginine Deiminase enzymes
Quantitative reverse transcriptase polymerase chain reaction
Ribosomal protein 17
Ribosomal protein 32
Terminal deoxynucleotidyl transferase (TdT) dUTP Nick-End Labeling
Mammary cancer is highly prevalent in dogs and cats, representing the most common cancer in intact bitches and the third most common neoplasm in queens [1, 2]. There are five basic classes of mammary cancer treatments in dogs and cats: surgery, radiation therapy, chemotherapy, hormone therapy and immunotherapy. Because the effectiveness of the latter four therapies is uncertain, surgery remains the treatment of choice . Unfortunately, tumor recurrence after surgery is common in both species and adjuvant treatment with chemotherapeutics often fails [4, 5, 6]. Therefore, there is an urgent need to develop more effective treatments for feline and canine mammary cancer.
Recent human and mouse studies have suggested that peptidyl arginine deiminase enzymes (PAD) could represent a novel target for epigenetic cancer therapy. The PAD family is made up of 5 (PAD1–4 and PAD6) calcium-dependent enzymes that convert positively charged arginine residues to neutrally charged citrulline target proteins, such as histones, in a process called citrullination . PAD are involved in cell differentiation, apoptosis, embryonic development and gene regulation, and each PAD isozyme has a unique tissue distribution pattern and substrate preference, which most likely confers biological specificity . PAD2 represents the ancestral PAD family member and is abundantly represented in female reproductive tissues, including the mammary gland . PAD2 activity was found to be significantly increased during human breast cancer progression and to be predictive of human breast cancer recurrence [9, 10]. Interestingly, PAD2 was recently identified in feline and canine carcinomas, suggesting that this oncogene may also play an important role in mammary cancer of small companion animals . PAD4 was initially identified in hematopoietic cells, and elevated expression of PAD4 has been observed in various malignant tumors [12, 13]. More recently, PAD4 was also found to be highly expressed in tumorigenic human breast cancer cell lines that are resistant to conventional therapeutics . However, the expression of PAD4 in feline or canine mammary cells has not been previously studied. A broad PAD inhibitor, Cl-Amidine, has been shown to have remarkable effects on human breast cancer in vitro and in vivo in a mouse xenograft model . A more recent study by Wang et al. showed that the PAD4 inhibitor, YW3–56, triggers breast cancer cell death by inducing the endoplasmic reticulum (ER) stress response [15, 16]. Cancer cells are frequently exposed to exogenous and endogenous insults that induce ER stress. ER stress normally activates an intricate pathway with the goal of restoring homeostasis; however, if this stress is too prolonged or severe, cell death will occur [17, 18, 19].
Mammary cancer in dogs and cats is similar to human breast cancer in histology, gene expression, and metastatic pattern. Given these similarities, PAD inhibitors could potentially serve as novel therapeutics for mammary cancer in these species as well . To begin exploring this possibility, we evaluated the effects of BB-Cl-Amidine (BB-CLA), a more potent derivative of Cl-Amidine , on normal and tumoral canine and feline mammary cells in vitro and in vivo, and studied the potential mechanisms underlying these effects.
The immortalized, non-malignant canine mammary epithelial cell line (CMEC), established in-house as previously described , the canine mammary carcinoma cell line REM134 , the immortalized, non-malignant feline mammary epithelial cell line (FMEC), a kind gift from Dr. John Parker (Baker Institute for Animal Health, Cornell University, Ithaca, NY) , and the feline mammary adenocarcinoma cell line K12–72.1, a kind gift from William Hardey Jr. (School of Veterinary Medicine, Cornell University, Ithaca, NY) , were cultured in DMEM supplemented with 10% fetal bovine serum (Atlanta Biologicals, Lawrenceville, GA) and 1% penicillin/streptomycin. All cell lines were maintained at 37 °C in a humidified environment with 5% CO2.
Antibodies and chemicals
The rabbit anti-human PAD2 and PAD4 antibodies were obtained from Proteintech (12110–1-AP; Rosemont, IL) and Sigma-Aldrich (P4749; St. Louis, MO), respectively. The rabbit anti-human caspase-3, activating transcription factor 4 (ATF4), 78 kDa Glucose-regulated Protein (GRP78) and beta actin antibodies were obtained from Abcam (ab4051; Cambridge, UK), Santa Cruz Biotechnology (sc-200; Dallas, TX), LS Bio (LS-C312961; Seattle, WA), and Abcam (ab8227), respectively. The mouse anti-dog/cat estrogen receptor alpha (ERα) antibody was obtained from Thermo Fisher Scientific (MA5–13065; Waltham, MA). The human anti-modified citrulline antibody was obtained from Millipore (MABS487; Darmstadt, Germany). Alexa 488-conjugated goat anti-rabbit antibodies, HRP-conjugated goat anti-mouse antibodies, HRP-conjugated goat anti-rabbit and HRP-conjugated goat anti-human antibodies were all obtained from Jackson ImmunoResearch Labs (West Grove, PA). The PAD inhibitor, BB-CLA, was synthesized as previously described and diluted in dimethyl sulfoxide (DMSO) .
Cell viability assays
After 48 h of treatment with either BB-CLA (0 to 20 μM) or DMSO (control), a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) in vitro toxicology assay (Sigma-Aldrich) was carried out per manufacturer’s instructions and absorbance was measured at 570 nm on a Multiskan EX plate reader (Thermo Fisher Scientific). Optical densities of wells treated with BB-CLA were compared with those treated with equivalent amounts of DMSO to determine cell viability. Values were expressed relative to DMSO treated wells.
Soft agar assays
Cells were detached using (0.25%) Trypsin-EDTA (Corning Life Sciences, Manassas, VA) and counted. Soft agar assays were performed as previously described . To this end, 2 mL of 0.6% 2-hydroxyethylagarose melted in appropriate culture medium were pipetted into wells of 6-well culture plates and plates were held at 4 °C for 15 min until the agarose solidified. Ten thousand cells per well were gently resuspended in 1.5 mL 0.3% 2-hydroxyethylagarose melted in appropriate culture medium with either DMSO or 10 μM BB-CLA, and layered over the base agarose. Cells were cultured in soft agar for 14 days at 37 °C with 5% CO2. Every 3 days, cultures were fed with 1 mL 0.3% 2-hydroxyethylagarose melted in appropriate culture medium with either DMSO or BB-CLA. Cultures were photographed at 10× using a Nikon Diaphot-TMD inverted light microscope with an attached Cohu CCD camera (Nikon, Melville, NY). The number of spheres, defined as clusters of cells increasing in size due to cell division , in 30 fields were counted.
Gene expression analyses
Primers used in the study
Forward Primer (5′ → 3′)
Reverse Primer (5′ → 3′)
Flow cytometry analyses
Cells were collected using accutase (Innovative Cell Technologies, San Diego, CA), fixed for 15 min in 4% paraformaldehyde (PFA), permeabilized and blocked in 0.1% Triton-X with 1% normal goat serum in PBS with 1% bovine serum albumin BSA for 1 h, and stained with PAD2 or PAD4 antibodies, diluted 1:50 in PBS with 1% BSA, for 1 h. Staining with an isotype control antibody (Abcam), diluted 1:50 in PBS with 1% BSA, or no antibody stain were included as controls. Cells were washed three times in PBS and incubated with Alexa 488-conjugated secondary goat anti-rabbit antibodies diluted 1:500 in PBS with 1% BSA, for 30 min. After washing, 10,000 cells were analyzed on a Gallios flow cytometer controlled by Kaluza for Gallios software (Beckman Coulter, Indianapolis, IN). Data analysis was conducted using Kaluza Analysis software (Beckman Coulter, Indianapolis, IN).
Western blot (WB) analyses
Cells were lysed in RIPA buffer containing a 1X general protease inhibitor. Protein concentration was determined with a BCA protein assay (Thermo Fisher Scientific) prior to gel loading to ensure loading of an equal amount of protein (40 μg). 6X reducing sample buffer was added to yield a final concentration of 1X and lysates were boiled for 10 min at 95 °C. Samples were subjected to SDS polyacrylamide gel electrophoresis on a 10% gel and transferred to Immobilin PVDF membranes (Millipore, Billerica, MA) using a transblot turbo system (Biorad, Hercules, CA). Membranes were blocked in 5% BSA diluted in Tris buffered saline (TBS) and incubated with PAD2 or PAD4 antibodies diluted 1:1000, GRP78 antibodies diluted 1:500, ERα antibodies diluted 1:200, or modified-citrulline antibodies diluted 1:1000 in TBS for 1 h room temperature (RT) on a rotating platform. Blots were washed for 50 min (10 × 5 min) with TBS-Tween, then incubated with HRP conjugated secondary antibodies for 1 h at RT. All blots were washed for 50 min (10 × 5 min) with TBS-Tween and then visualized by chemiluminescence using Clarity Western ECL (BioRad, Hercules, CA). Membranes were stripped and probed with beta actin antibodies, diluted 1:5000, as a loading control. Gels were imaged on a BioRad ChemiDoc MP system (BioRad) and band intensities were determined using Image Lab software. Intensities of the bands of interest were divided by the intensities of loading control bands to calculate relative band intensity.
Cells for IF were grown in 24 well culture dishes fitted with 35 mm coverslips, rinsed with PBS and fixed in 4% PFA for 10 min. Following 3 rinses with PBS, cells were permeabilized using PBS + 1% Triton-X 100 + 1% BSA for 30 min at RT. Primary antibodies against PAD2, PAD4, ATF4, all at 1:50 dilution in PBS, or Caspase-3 at 1:100 dilution, were added to the wells for 1 h at 37 °C. Wells were rinsed 3 times with PBS and Alexa 488-conjugated secondary antibodies were added to appropriate wells. After 30 min at RT, cells were washed 3 times with PBS and DAPI was added for 5 min. After a PBS wash, coverslips were removed from wells and mounted on slides using mounting medium (Dako, Carpenteria, CA). Cells were examined with a Zeiss LSM confocal microscope (Oberkochen, Germany) and images were captured with a camera controlled by ZEN imaging software. Nine images per sample, approximately 250 cells, were analyzed and DAPI-PAD co-localization was quantified using Image J Software.
Generation of REM134 and K12.72.1 xenografts
Both REM134 and K12.72.1 xenograft tumors were generated by injecting 1 × 106 cells in 0.1 mL Matrigel (1:1) (BD Biosciences, San Jose, CA, USA) subcutaneously near the nipple of gland #4 in 6-week old female NOD scid gamma (NSG) mice (Taconic, Germantown, NY, USA), as previously reported for mammary xenograft models . After two weeks, intraperitoneal injections of BB-CLA (1 mg/kg/day) or vehicle control (PBS) were initiated and carried out for 14 days. Tumor diameter was measured daily by digital caliper and the volume (mm3) was calculated using the formula: a2 x b/2, where “a” is the shortest diameter and “b” is the longest diameter of the tumor. Results are reported as mean ± SD. After 14 days, tumors were removed and added to 10% buffered formalin. Parrafin sections were stained with hematoxylin and eosin to determine for mitotic rate and necrosis. A rodent pathologist estimated the percent of necrosis by looking at the entire tumor at low power. For mitotic rate, the number of mitotic figures were counted in the area with the highest mitotic rate in a 400× field. To measure apoptosis, a Terminal deoxynucleotidyl transferase (TdT) dUTP Nick-End Labeling (TUNEL) assay (Sigma-Aldrich) was used according to manufacturer’s instructions and images were quantified using Image J Software. Three mice per group were used for each treatment. All mouse experiments were reviewed and approved by the Institutional Animal Care and Use Committees (IACUC) at Cornell University (#2013–0022).
All experiments were repeated at least three times and results were expressed as the mean ± STDEV from three independent experiments. Data were analyzed by the Student’s T-test and MTT assays were analyzed by one-way ANOVA and p values < 0.05 were considered statistically significant. * indicates P < .05, ** P < .01, *** P < .001, and **** P < .0001.
BB-CLA selectively inhibits growth of canine and feline mammary tumor cells compared to normal mammary epithelial cells
Normal and tumoral feline and canine mammary cells express similar levels of PAD2 and PAD4
BB-CLA activates the endoplasmatic reticulum (ER) stress pathway in canine and feline tumoral mammary cells
Generation of canine and feline mammary cancer xenograft mice
This study was initiated to evaluate the efficacy of the peptidyl arginine deiminase (PAD) inhibitor BB-Cl-Amidine (BB-CLA) on canine and feline mammary cancer cell lines in vitro and in vivo. Our salient findings were that BB-CLA reduces the viability and tumorigenicity of canine and feline mammary cancer cell lines at low concentrations, via the activation of the endoplasmic reticulum (ER) stress pathway, with no to minimal effects on the viability of normal mammary cells. This study also provided the first description of PAD4, and a more in-depth description of PAD2, expression in normal and tumoral mammary cells from canine and feline origin. In addition, we report on the generation of xenograft models using the cell lines REM134 and K12.72.1, and the preliminary evaluation of BB-CLA efficacy in these models.
In contrast to what has been described previously for human breast cell lines, we found a similar PAD2 and PAD4 expression between normal and tumoral mammary canine and feline cell lines. One explanation for this discrepancy between species is based on the source of the cells that were used. McElwee et al. found higher PAD2 expression in human breast cancer cells compared to normal mammary epithelial cells using the MCF10AT mammary cell line series in which the tumor cells were developed from the transformed normal mammary MCF10a cell line, and thus all cells have the same cell of origin [10, 38, 39]. In our study, the normal and tumoral feline and canine cell lines being compared were developed from different individuals for both species. Therefore, inter-animal variations in PAD2 and PAD4 levels could explain why no differences were detected between normal and tumoral cells in the present study. It would be interesting in future work to look at PAD2 and PAD4 expression in primary tumor and healthy tissue samples instead of cell lines and/or to use cell line series originating from the same animal, similar to the MCF10AT cell lines series, to determine if there are indeed inter-animal variations. Another explanation for the observed discrepancy could be the receptor status of the cell lines used. Previous data on PAD2 and PAD4 expression in human breast cells were primarily generated using estrogen receptor alpha (ERα)-positive breast cancer cells , whereas the canine and feline cell lines used in this study were ERα negative. This explanation is further supported by previous studies showing that PAD2 is more abundantly expressed in the luminal breast cancer subtype, which is ERα positive . Additional studies should evaluate PAD expression in ERα-positive feline and canine mammary cancer cells to determine if ERα expression is indeed correlated with PAD expression.
Our results showed that treatment with the PAD inhibitor BB-CLA in canine and feline mammary cancer cells resulted in an activation of the ER stress pathway. This is in line with a recent study from Wang et al., showing that YW3–56, another PAD inhibitor, initiated the ER stress pathway through activation of ATF4 in human breast cancer cells, leading to autophagy in these cells [1, 2]. When evaluating the different genes and proteins involved in the ER stress pathway in our canine and feline cell lines, some interesting observations were made. First, we observed a difference in ATF4 localization between feline and canine mammary cancer cells, irrespective of BB-CLA treatment, and so we like to postulate that these inherent differences may be reflective of the tumoral malignancy in vivo. It is known that feline mammary cancer is on average more aggressive than canine mammary cancer, with 80% of mammary tumors being malignant in cats versus 50% being malignant in dogs, and consequently, that mammary cancer in cats has a worse prognosis [40, 41]. Also, we observed that the feline tumoral cells were more responsive to BB-CLA compared to canine tumoral cells and in parallel, we found that untreated feline cells have inherently more nuclear ATF4 and thus are at a higher baseline ER stress level. Since BB-CLA stimulates the ER stress pathway, this could explain why a greater effect was seen with this drug in the feline cells. Second, we observed a significant upregulation of the gene DDIT3, but not DDIT4, after BB-CLA treatment in both feline and canine mammary cancer cell lines, which could indicate that BB-CLA is promoting activation of the apoptosis pathway rather than the autophagy pathway in response to ER stress [18, 42, 43, 44]. Indeed, cells were positive for Caspase-3, a marker of apoptosis, after treatment with BB-CLA, indicating this is likely the predominant method for cell death after ER stress in these cells.
Despite the use of REM134 and K12.72.1 cell lines in canine and feline mammary cancer research, there were no reports of mouse orthotopic xenografts utilizing these cell lines prior to this study. REM134 has previously been used in xenografts after subcutaneous inoculation [22, 45], however orthotopic xenografts offer the advantage of a tumor microenvironment that leads to the development of tumor cells more biologically similar to clinical cases and with metastatic potential [46, 47, 48]. We found that tumors can be induced in the mammary gland reproducibly in NOD scid gamma (NSG) mice using these cell lines. Unfortunately, when we preliminary analyzed the efficacy of BB-CLA in these xenograft models, we did not find robust differences in tumor size, necrosis or mitotic rates in treated versus untreated animals. A potential explanation could be the dosage of BB-CLA used for treatment, which was based on a previous in vivo study where BB-CLA was used in a murine arthritis model . It is plausible that the dosage needed for effective treatment of mammary cancer may be higher than the dose needed for the treatment of arthritis. Additionally, a longer dosage period or more frequent doses may have increased effectiveness. Future experiments in xenograft models of mammary cancer are needed to define the optimal effective BB-CLA dose. Unexpectedly, there was an increase in mitotic rate following treatment with BB-CLA in the K12–72.1 xenograft model. This would normally be associated with a poorer prognosis, however, the tumor volumes were similar in both groups. BB-CLA may have had additional effects on the tumor microenvironment that allowed for increased aggressiveness of the tumor cells, but this is unlikely as the mitotic rate was only increased in the K12–72.1 tumors and not REM134 tumors. Lastly, this finding could also be a product of low sample numbers, which should be addressed in future studies. We did, however, observe that the tumors in the BB-CLA-treated mice were crusty and that the skin showed local hair loss. A potential explanation for this observation is that other PAD, like PAD1, PAD2, and PAD3, are expressed in the skin of humans and rodents , and could have been affected by the PAD-inhibitor BB-CLA.
Taken together, our in vitro study provides the first data on the potential of BB-CLA as a novel therapeutic to treat mammary cancer in dogs and cats. This is an exciting finding given the lack of available treatments for mammary cancer in these species. Canine and feline mammary cancer cell lines were successfully xenografted in the mouse mammary gland and will be beneficial in future research aimed at the development of mammary cancer therapeutics.
We are very grateful for the excellent technical assistance of Allison Seebald and Lynne Anguish.
This work was supported by the Morris Animal Foundation (grant #D14CA-063), the Feline Health Center, and the President’s Council of Cornell Women, to GRV; and the National Science Foundation (grant #DGE-1144153) to MML.
Availability of data and materials
MML carried out the majority of the laboratory procedures, was involved in conception and design, and manuscript writing; RA and RH assisted with the laboratory procedures; AM and PRT provided expertise with the BB-CLA experiments; SC and GVdW were involved in conception and design; GVdW was involved in data analyses and manuscript writing. All authors read and approved the final manuscript.
All mouse experiments were reviewed and approved by the Institutional Animal Care and Use Committees (IACUC) at Cornell University (#2013–0022).
Consent for publication
The authors declare they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 5.Simon D, Schoenrock D, Baumgärtner W, Nolte I. Postoperative adjuvant treatment of invasive malignant mammary gland tumors in dogs with doxorubicin and docetaxel. J Vet Intern Med. 20:1184–90.Google Scholar
- 26.Horibata S, Vo TV, Subramanian V, Thompson PR, Coonrod SA. Utilization of the soft agar colony formation assay to identify inhibitors of tumorigenicity in breast cancer cells. J Vis Exp. 2015;99:e52727.Google Scholar
- 41.Li B, Chen R, Chen L, Qiu P, Ai X, Huang E, et al. Effects of DDIT4 in methamphetamine-induced autophagy and apoptosis in dopaminergic neurons. Mol Neurobiol. 2016;54(3):1642–60.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.