Background
Breast cancer is a heterogeneous disease that varies in etiology, pathophysiology and response to therapy. As a result, patients with disease of similar stage and grade often respond differently to therapy leading to disparate clinical outcomes. Molecular profiles characterizing the various intrinsic breast cancer subtypes, as per gene expression signatures, have been successful for predicting overall survival, relapse, and response to chemotherapy [
1‐
4]. Luminal subtypes are defined by expression of estrogen receptor α (ERα) and cell cytokeratins (CKs) 8 and 18 [
5,
6]. Basal-like tumors are typically triple-negative (i.e. lacking expression of ERα, progesterone receptor, and human epidermal growth factor receptor 2 (HER2)), yet express basal CKs 5, 14, and/or 17 [
5,
7,
8]. The claudin-low subtype is characterized by low gene expression of junction and adhesion proteins that include claudins 3, 4 and 7, as well as E-cadherin [
3]. While these tumors initially respond to chemotherapy, there is a high risk of recurrence and disease progression, consequently leading to poor patient survival [
9‐
11].
Abnormal protein regulation of cell-surface receptors promotes cancer development/progression, and is widely used to determine patient prognosis and dictate therapeutic regime. CD44 and lipolysis stimulated lipoprotein receptor (LSR) are both cell-surface, transmembrane proteins that mediate cellular responses towards their microenvironment. These molecules participate in cell-cell and cell-matrix interactions, as well as regulate cell growth, survival, differentiation, and motility [
12‐
14]. High CD44 levels are a marker for tumor initiating and chemotherapeutic-resistant cells in many cancers, including breast [
15,
16]. High CD44-expressing cells have heightened tumorigenicity, self-renewal
in vivo, and give rise to functional as well as molecular heterogeneity: properties directly linked to chemotherapeutic-resistant, aggressive cancers [
15]. It has also been reported that basal-like tumors contain the highest percentage of CD44-positive cells [
17], while high CD44 expression correlates to a basal-like phenotype, increased metastases, and unfavorable prognosis in breast cancer patients [
18‐
20]. Similar to high CD44 levels, increased expression of LSR has been associated with altered gene expression of pathways involved in transformation and tumorigenesis, enhanced proliferation, survival in anchorage independent conditions and promotion of collective cell migration in breast cancer cells [
21]. High LSR levels have also been identified as a marker for tumor-initiating and chemotherapeutic-resistant cells [
14]. Collectively, these studies highlight a direct role for LSR in driving aggressive breast cancer behavior.
The use of bacterial toxins for selective and efficient cancer therapeutics has been gaining attention due to recent successes
in vitro and
in vivo[
22,
23]. Bacterial toxins possess efficient cytotoxic capabilities, making them suitable candidates for gene therapeutic applications towards various cancers [
24‐
31].
Clostridium perfringens iota toxin has various properties that make it a potential candidate for targeted cancer therapy. For instance, like many of the “classic” AB exotoxins, iota toxin is secreted by the bacterium and contains two functionally distinct, subunits not linked in solution [
32]. The B subunit (Ib) binds to a cell-surface receptor, facilitating docking and uptake of the enzymatic A subunit (Ia) through receptor-mediated endocytosis. Ib forms heptamers on the cell surface and creates pores within an acidified endosome membrane enabling release of Ia into the cytosol. The Ia molecule mono-ADP-ribosylates G actin, subsequently preventing F actin assembly that leads to overt rounding of cells and death [
32‐
34].
Recent studies have implicated LSR and CD44 as functional receptors, or co-receptors, mediating iota toxin binding to host cells [
35,
36]. However, the relationship between these cell-surface proteins with respect to promotion of iota cytotoxicity is still unclear, given the various model systems used during these investigations. In the present study, we investigated the role of LSR and CD44 in a tissue-specific manner by identifying which protein mediates the cytotoxic processes specific to breast cancer. Ultimately, our future goal is to evaluate the potential of using iota toxin as an adjuvant, targeted therapy for breast cancers that may be less toxic than current available treatments.
Discussion
The objective of the current study was to further characterize the roles of LSR and CD44 during
C. perfringens iota cytotoxicity on breast cancer cells. Two studies have indicated that iota toxin has the ability to bind to membrane-bound proteins, CD44 and LSR [
35,
36]. Complementary to the study by Papatheodorou
et al.[
35], in HAP-1 and HeLa cells, treatment of 14 breast cancer cell lines in our study show that those expressing LSR were sensitive to iota toxin. Additionally, consistent with reports in fibroblasts and hepatocytes demonstrating enhanced LSR-mediated endocytosis in the presence of oleic acid, treatment of LSR-expressing breast cancer cells with oleic acid increased sensitivity to iota toxin. However, our data presents an interesting mechanism in breast tissue that is contrary to reports in other tissues. Remarkably, breast cancer cells expressing CD44 displayed varying levels of toxin resistance. Specifically, LSR+/CD44- lines were highly sensitive, LSR+/CD44+ lines were slightly sensitive, and LSR-/CD44+ lines were resistant to the cytotoxic effects of iota toxin. Consistent with this observation, toxin sensitivity was highest in the luminal cell lines, median in basal-like, and lowest in claudin-low lines, corresponding to their reciprocal expression levels of LSR and CD44. Toxin sensitivity among the various basal-like cell lines was heterogeneous which can be attributed, in part, to the individual cell lines diverse expression of both CD44 and LSR levels. It is important to note that while our data are unlike those found in the study by Wigelsworth
et al. [
36], where they found CD44 expression promotes iota intoxication in Vero (African green monkey kidney) and human melanoma (RPM) cell cultures (
in vitro), as well as in mice (
in vivo lethality), the authors state that the cells they used contained LSR. They further investigated interaction between LSR and CD44 via co-precipitation experiments, showing no interaction between the proteins; however, they acknowledge that CD44 and LSR may co-facilitate entry of iota toxin into cells via an unknown mechanism. This is highlighted by the fact that the authors were unable to completely block intoxication by anti-CD44 and high amounts of toxin still cause cytotoxicity in CD44- cells [
36].
Our current study suggests that the cellular response to iota toxin is tissue-specific, and for reasons not totally understood at this time. As CD44 serves many roles for a cell, and appears as many different isotypes, there clearly needs to be further study to determine more definitively the role(s) played by CD44 during iota intoxication. In fact, this current study reveals that CD44 prevents endocytosis of iota toxin and conveys cytotoxic resistance in breast cancer cells. Other studies have shown that post-translational modifications of CD44 may also affect CD44 function and endocytosis [
12,
40‐
42]. In our current study, we show that treatment with deglycosylation agents did not appreciably affect iota toxin cytotoxicity; however, other modifications unknown to us may be involved. For example, inhibition of CD44 palmitylation has no effect on CD44 binding to hyaluronan, yet there is inhibition of hyaluronan internalization [
52]. Moreover, the posttranslational modifications and variants of LSR may also play a role in tissue-specific toxin endocytosis. While currently little is known about these variants and post-translational modifications of LSR, one study identifies at least one phosphorylated site (Ser
435 within RPRARpS
435VDAL) that affects binding of 14-3-3, a cytosolic adaptor protein involved in mediating signaling pathways by binding to phosphoserine-containing proteins [
53].
Identifying the precise mechanisms of interaction between iota toxin and cell-surface proteins, as well as the subsequent downstream intracellular pathways that lead to cytotoxicity, are necessary steps in developing targeted therapies for breast cancer. Earlier studies show that in Vero cells, iota toxin enters through clathrin-independent endocytosis mechanisms, is dependent upon dynamin, and regulated by Rho-GDI [
54]. Similarly, both CD44 and LSR have been proposed to be internalized via clathrin-independent mechanisms in non-breast tissues [
36]. Central to our studies, breast cancer stem/tumor-initiating cells have a significantly higher rate of clathrin-independent endocytosis [
55]. We have also previously shown that cells expressing high levels of LSR have enhanced cancer stem cell-like properties [
21], thus collectively suggesting that iota toxin may have heightened effects upon breast cancer stem cells.
Nagahama
et al.[
56] describe the dynamics of intracellular trafficking of the Ib component of iota toxin. Through their study of MDCK cells, they found that post-internalization involves Ia escape from early endosomes into the cytosol and subsequent ADP-ribosylation of α- and β-actin. The majority of Ib goes through the endocytotic pathway into lysosomes and is degraded. A small percentage of Ib reportedly recycles back to the plasma membrane, which they suggest extends Ia entry into the target cell [
56,
57]. In line with this mechanism, our study revealed Ib within the lysates of LSR + cells treated with iota toxin and enhanced lysosome formation, demonstrating Ib internalization. While we did not presently analyze whether cells treated with low levels of toxin had a percentage of Ib recycled to the cell-surface, this strengthens the potential of using iota toxin as a therapeutic. Recycled Ib to the plasma membrane may further sensitize the cancer cell to a secondary toxin (Ia) treatment, thereby potentially eradicating any remaining cells.
Tamoxifen and aromatase inhibitors (AI) are commonly used to treat ERα-positive breast cancers as these therapies inhibit estrogenic signaling, ultimately leading to inhibition of cell proliferation and survival involving activated apoptotic pathways [
58‐
61]. Tamoxifen and AI resistance are emergent clinical problems that induce phenotypic changes in tumor cells, including decreased apoptosis as well as increased proliferation and invasion [
58‐
61]; however, the molecular mechanisms behind resistance are largely unknown. As we show in this current study, iota toxin interacts with LSR to induce cell death and LSR expression is correlated with ERα-positive breast cancers [
21]. Our laboratory is currently testing the potential of iota toxin as an adjuvant therapy for women with ERα-positive, tamoxifen- and AI-resistant breast cancers. We have previously shown a multifaceted role of LSR in directing breast cancer cell behavior. For example, over-expression of LSR enhances cell proliferation and migration, as well as stimulates cancer-stem cell related properties such as survival in anchorage-independent conditions [
21]. We also show that LSR expression significantly correlates with ERα expression in primary breast cancer biopsies [
21]. In the current study, we show that tamoxifen-resistant breast cancer cell lines also express LSR and are sensitive to iota toxin-induced cytotoxicity. Iota toxin evidently circumvents the pro-survival mechanisms employed in anti-hormone resistant breast cancers by exploiting necrotic pathways.
Bacterial immunotoxins have been used in clinical trials to successfully treat hematological malignancies and solid tumors, as well as used as an adjuvant therapy targeting mesothelin-expressing mesothelioma, ovarian, or pancreatic cancer [
62‐
66]. Immunotoxins derived from
Pseudomonas exotoxin A, or plant-based ricin, subunits attached to antibody fragments have been evaluated in Phase I and II clinical trials for treating solid tumors. These trials revealed that immunotoxins could specifically target cell-surface antigens expressed at high levels in tumors. Another Phase I clinical trial, conducted by von Minckwiz and colleagues, used a single-chain immunotoxin targeting Her2 from eighteen Her2-expressing cancer patients. Intratumoral injections of immunotoxin successfully reduced tumor size [
64,
65]. Additionally, studies in nude mice with mesothelin-expressing tumor xenografts reveal enhanced therapeutic responses when taxon, or other cancer drugs, are administered in combination with SS1P, a high-affinity immunotoxin that targets mesothelin [
66]. A Phase I trial was subsequently initiated, where mesothelin-positive and recurrent or unresectable mesothelioma, ovarian, or pancreatic cancer patients were infused with SS1P continually for ten days. The recombinant immunotoxin was well tolerated by patients and showed modest clinical benefit. Our current study revealed that iota toxin specifically targets LSR-expressing breast cancer cells and exerts a rapid cytotoxic effect. This gives iota toxin great potential to be utilized as an immunotoxin for (i) transporting foreign proteins into targeted cells, (ii) modification to increase specificity to a specific cell type, and (iii) increasing drug absorption of chemotherapeutic drugs [
32,
67].
Two studies with another bacterial toxin,
Clostridium perfringens enterotoxin (CPE), reveal that CPE specifically targets claudin-overexpressing mouse NT6 fibroblasts, human colorectal adenocarcinoma (Caco-2), colon (HCT116) and mammary (MCF-7) cell lines [
22,
23]. The study by Walther
et al. utilized non-viral, intratumoral
in vivo gene transfer of CPE into mice with MCF-7 and HCT116 xenografts, resulting in reduced tumor growth compared to control groups [
22]. Translational explorations of
C. perfringens iota toxin as a chemotherapeutic are yet to be exploited to date. A study by Sakurai and Kobayashi evaluated the role of Ia and Ib subunits in guinea pigs [
68]. When Ib was injected intradermally and Ia intrapertoneally, Ia was able to specifically target the Ib component in the skin resulting in localized dermonecrosis without other side effects. These studies support the feasibility of iota toxin as a specific, localized tool for drug therapy. Importantly, unlike CPE, which has the risk of eliciting its toxic effects to normal claudin expressing cells, iota toxin has had no reported effects in humans [
68,
69]. When combined with aforementioned results from the Sakurai and Kobayashi study, targeted therapeutics against breast cancer derived from iota toxin may provide a well-tolerated, effective alternative with lower off target effects compared to current targeted therapies.
Methods
Cell culture
MCF-7, T47D, ZR-75-1, HCC1937, MDA-MB-468, BT-20, HCC1143, MDA-MB-231, Hs578t, BT-549, AU565, SKBR3, M99005, MCF-10AI, MCF-10AIII, and MCF-10AIV cells were obtained from American Type Culture Collection (ATCC; Manassas, VA). SUM159, SUM149, Sum190, and SUM1315mo cells were obtained from Asterand (Detroit, MI, USA). TMX2-4, TMX2-11 and TMX2-28 cells were a kind gift from Dr. Kathleen Arcaro (University of Massachusetts, Amherst). Cells were cultured according to manufacturer’s recommendations and passaged via trypsinization when approximately 80% confluent.
Iota toxin, reagents, and toxicity testing
Iota toxin components Ia and Ib were purified as described previously [
70]. For toxin sensitivity assays, cells were seeded at 1 - 3×10
4 concentrations, enabling confluency 48 h later. Cells were then treated as either a control (10 ng Ia/ml) or with varying concentrations of iota toxin (Ia 10 ng/ml + Ib 20 ng/ml; Ia 25 ng/ml + Ib 50 ng/ml; Ia 50 ng/ml + Ib 100 ng/ml; Ia 100 ng/ml + Ib 200 ng/ml) and cultured under normal growth conditions. Observations to determine toxin sensitivity, indicated by a rounded morphology indicative of cell death, were made at 0, 1, 2, 4, 6, and 8 h post treatment. For oleic acid assays, cells were treated in the presence or absence of 0.8 mM oleic acid at 37°C, 30 min prior to added toxin. Images were obtained by Spot Advanced version 4.5 (Sterling Heights, Michigan).
Generation of knockdown and overexpressing cell lines
RFP containing HuSH shRNA plasmids containing Homo sapiens LSR specific shRNA and Myc-DDK-tagged TrueORF clones of Homo sapiens LSR and CD44 were obtained from OriGene Technologies (cat# TF303412, RC223636, and RC221771; Rockville, MD). Cells were transfected using TurboFectin 8.0 (Thermo Scientific, Rockford, IL) according to manufacturer’s instructions. For stable transfection, cells were passaged at a 1:10 dilution into fresh growth medium containing 2.5 μg/ml Puromycin or 500–900 μg/ml of G418 (Life Technologies, Grand Island, NY). Control cells were simultaneously transfected with an empty plasmid vector and selected in antibiotic-containing medium as described above.
Western blot analysis
Cells were lysed in RIPA Buffer (50 mM Tris Base, 150 mM NaCl, 1 mM EDTA, 1% NP40, 0.25% sodium deoxycholate) supplemented with protease and phosphatase inhibitors (Halt™ Thermo Scientific, Rockford, IL). Equal protein concentrations of total cell lysates, as determined by the Coomassie Plus Protein Assay (Thermo Scientific, Rockford, IL), were separated by SDS-PAGE. Proteins were transferred to nitrocellulose membranes (BioExpress, Kaysville, UT). Membranes were blocked in 5% non-fat milk in TBST (1.0 M Tris–HCl, 5.0 M NaCl, 0.1% Tween) for 1 h at room temperature, then incubated with primary antibody against LSR (1:750; sc-133765), HCAM (CD44; 1:500; sc-7297), E-cadherin (1:500; sc-7870; Santa Cruz Biotechnology, Santa Cruz, CA), or anti-Ib (1:1000) overnight at 4°C in TBST containing 5% BSA. Membranes were then washed and incubated with the appropriate secondary antibody conjugated to horseradish peroxidase (GE Healthcare, Piscataway, NJ) in TBST with 5% milk for 1 h at room temperature. Mouse monoclonal α-tubulin antibody was used to evaluate equal protein loading across all lanes at a 1:5000 dilution (T6199; Sigma Aldrich, St. Louis, MO). WesternBright ECL Kit (Bioexpress, Kaysville, UT) was used to detect peroxidase activity. NIH Image J64 software was used to quantify western blots.
Immunocytofluorescence
Immunocytofluorescence was performed as previously described [
71]. Briefly, cells were grown on 8-well chamber slides (Research Products International, Mt. Prospect, IL,) and fixed/permeabilized in ice-cold methanol:acetone. Following fixation, cells were blocked with 1% BSA and 5% normal horse serum in PBS, stained with the indicated primary antibody (1:100 dilution of anti-LSR, sc-133765 or anti-HCAM (CD44), sc-7297) for 1 h at 4°, washed, and then incubated for 30 min with an anti-rabbit or anti-mouse Alexa Fluor 488 secondary antibody (1:1000 dilution, Invitrogen). Coverslips were applied with ProLong® Gold Antifade Reagent and DAPI (Life Technologies). Imaging was performed on a Nikon DiaPhot microscope with digital camera and NIS-Elements 4.11.00 (Nikon Instruments Inc., Melville, NY). All cell lines and samples were obtained in compliance with the Helsinki Declaration and performed in accordance with the guidelines of the North Carolina Central University Institutional Review Board, approval 1201027.
Glycosylation analysis
Cells were grown under normal growth conditions until approximately 70% confluent, and then serum starved overnight. Cells were subsequently treated with 25 μg/ml of either Tunicamycin (MP Biomedical LLC, Solon, OH) or Swainsonine (Calbiochem, San Diego, CA), or vehicle control. Twenty-four hours post treatment, cell lysates were collected and analyzed via western blot analysis to determine glycosylation status of LSR.
Lysosome detection assay
Cells were grown under normal growth conditions until approximately 70% confluent. Cells were then treated as either a control (10 ng Ia/ml), or with iota toxin consisting of Ia (100 ng/ml) plus Ib (200 ng/ml), and cultured under normal growth conditions for 30 min. Following treatment, LysoTracker Green DND-26 (Cell Signaling Technology, Danvers, MA) was diluted 1:20,000 (50 nM) and added directly into growth medium, followed by imaging on a NIS-Elements 4.11.00. A minimum of three replicate wells was plated for each independent experiment, with a minimum of five fields imaged per well.
Cell death assays
Cells were seeded at 1- 3×104 concentrations, obtaining confluency 48 h later. Cells were then treated as controls (10 ng Ia/ml) or with iota toxin at low (Ia 10 ng/ml + Ib 20 ng/ml) or high (Ia 100 ng/ml + Ib 200 ng/ml) concentrations followed by culturing under normal growth conditions for 0, 2, 4, 6, and 8 h. Post treatment, cytotoxicity was determined using a CytoTox-Fluor™ Cytotoxicity Assay (Promega, Madison, WI) per manufacturer’s instructions.
CD44 variant analysis
RT-qPCR amplification reactions were conducted in duplicate using 1X Brilliant II SYBR® Green QPCR MasterMix (Agilent Technologies, Cary, NC) in the presence of variant specific primers (800 nM each; Additional file
3: Table S2) and 40 ng of cDNA (based on total RNA) in 20 μl. A non-template reaction was used as negative control. PCR conditions consisted of denaturation at 95°C for 10 min, activation of the DNA polymerase, followed by 40 cycles of 95°C for 15 seconds and specific annealing temperatures for each splicing variant for 1 min. Melting curves were generated after amplification at 95°C for 15 seconds, 60°C for 30 seconds and 95°C for 15 seconds. All reactions were conducted in a Stratagene Mx3005P detection system (Stratagene, La Jolla, CA). Amplification efficiency of each pair of primers was calculated using standard curve dilutions and incorporated into the calculation for relative expression differences as previously described [
72]. The optimal normalization factor was calculated as the geometric mean of the reference targets
B2M,
SDHA,
UBC and
YWHAZ.
LSR variant analysis
RT-qPCR amplification reactions were conducted in duplicate using 1X Brilliant II SYBR® Green QPCR MasterMix (Agilent Technologies, Cary, NC) in the presence of optimized variant specific primers (800 nM; Supplementary Table
1) and 100 ng of cDNA (based on total RNA) in 20 μl. A non-template reaction was used as negative control. PCR conditions were the same as those used for CD44 analysis. All reactions were conducted in an Applied Biosystems’ 7500 Real Time PCR system (Grand Island, NY). Raw data were normalized to
GAPDH and analyzed via the comparative CT (
ΔΔCT) method.
Statistical analysis
To evaluate toxin sensitivity, relative percent of cell rounding was determined via visualization using light microscopy for each treatment group and statistical significances between treatment groups and across cell lines was determined by student t-test using GraphPad Prism version 3.02 software (GraphPad Software Inc., San Diego, CA), significance was set at P < 0.05. Differences in statistical significance between cell lines regarding CD44 and LSR expression levels (total and variants), was determined by ANOVA and post-hoc two-tailed comparisons. Significance was set at P < 0.05 and a Bonferonni correction was used to adjust the P-value of t-tests. Graphs were plotted in Microsoft Excel as mean ± S.D.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
JMF and BGS conceived and designed the study; KFS, DKR, MCR, and JMF performed the experiments; BGS, and JMF analyzed and interpreted data; MRP supervised the study; KFS, BGS, JMF wrote the paper. All authors read and approved the final manuscript.