Clofazimine inhibits small-cell lung cancer progression by modulating the kynurenine/aryl hydrocarbon receptor axis

Abstract

Small cell lung cancer (SCLC) is one of the highly metastatic malignancies that contributes to ∼15 % of all lung cancers. Most SCLC patients (50-60 %) develop osteolytic bone metastases, significantly affecting their quality of life. Among several factors, environmental pollutant 2,3,7,8-Tetrachlorodibenzodioxin (TCDD) and kynurenine (Kyn), an endogenous ligand derived from tryptophan (Trp) metabolism, activate the aryl hydrocarbon receptor (AhR) and are responsible for SCLC progression and metastasis. Further, elevated AhR expression in bone cells intensifies bone resorption, making the Kyn/AhR axis a potential target for the bone metastatic propensity of SCLC. We first assessed the expression profile of AhR in human SCLC cell lines and found a significantly increased expression compared to normal lung cells. Additionally, we also evaluated the clinical significance of AhR expression in the patient samples of SCLC along with the relevance of the same in the Rb1^fl/fl; Trp53^fl/fl; Myc^LSL/LSL (RPM) mouse model using immunohistochemistry and found the higher AhR expression in the patient samples and RPM mouse tumor tissues. Using computational simulations, we found that clofazimine (CLF) binds at the activator (Kyn) binding site by forming a stable complex with AhR. The CLF binding with AhR was favored by Van der Waals and hydrophobic forces, and the proteins retained their secondary structure. Furthermore, we found that Kyn treatment potentiates the migration and clonogenic ability of SCLC cell lines by activating Erk/Akt oncogenic signaling. Treatment with CLF reduces AhR expression, which inhibits Kyn-mediated proliferation of SCLC cells, induces apoptosis, and cell cycle arrest in the G2/M phase. Further, our examination indicates that Kyn treatment also promotes osteoblast-mediated osteoclast differentiation through RANKL. The treatment with CLF impedes RANKL expression and osteoclastogenesis, suggesting that CLF has the potential to be used as the therapeutic for SCLC patients having bone metastasis.

Keywords: Aryl hydrocarbon receptor; Bone metastasis; Clofazimine; Kynurenin; Small cell lung cancer.

Figure 1.. Enhanced expression profile of AhR is associated with an aggressive tumor phenotype in SCLC patients and in spontaneous mouse models.

A. The expression profile of AhR at the protein level has been shown in three SCLC cell lines (SBC3, SBC5, NCI-H69), and BEAS-2B, a normal bronchoalveolar epithelial cell line, was used as a normal control. β-actin served as a loading control. B. Representative images of IHC for AhR staining on tissue microarray with samples from SCLC patients and the normal lung tissue. Magnification x10 and x20 (left panel). Quantification analysis of IHC scoring of AhR in the tumor and normal lung tissue (right panel). C. Illustration showing strong staining of AhR in the lung of wild type and Rb1/Trp53/MycT58A mice (spontaneous mouse model for SCLC) (right panel) and quantification of IHC scores in the same (left panel). Results are represented as means ± SEM. Statistical significance has been calculated using one-way ANOVA and unpaired t-test. *p<0.05, ***p<0.0003, ****p<0.0001.

Figure 2.. CLF interacts at the inhibitor/Kyn binding site in AhR.

A. Overlap of cryo-EM AhR-INDR complex with docked AhR-INDR complex. AhR is shown in rainbow ribbons, cryo-EM INDR is shown in red sticks, and docked INDR is shown in magenta sticks. B. Overlap of the cryo-EM INDR to docked CLF and Kyn in AhR. AhR is shown in rainbow ribbons, INDR is shown in red sticks, CLF is shown in green sticks, and Kyn is shown in magenta sticks. C. Two-dimensional representation of docked AhR-CLF complex. D. Two-dimensional representation of the docked AhR-Kyn complex. CLF is clofazimine, INDR is indirubin, and Kyn is kynurenine.

Figure 3.. CLF forms a stable complex with AhR.

A. Root-mean square deviations (RMSD) of backbone atoms of apo AhR, AhR-CLF complex, AhR-INDR complex, and AhR-Kyn complex as a function of time. B. Root-mean square fluctuations (RMSF) of C_α residues of AhR in the absence and presence of INDR, CLF, or Kyn. C. Radius of gyration (R_g) of backbone atoms of apo AhR, AhR-CLF complex, AhR-INDR complex, and AhR-Kyn complex as a function of time. D. Solvent accessible surface area (SASA) of apo AhR, AhR-CLF complex, AhR-INDR complex, and AhR-Kyn complex as a function of time. CLF is clofazimine, INDR is indirubin, and Kyn is kynurenine.

Figure 4.. CLF and Kyn form consistent hydrogen bonds with AhR throughout the trajectory.

A. The number of hydrogen bonds formed by CLF, INDR, or Kyn with AhR as a function of simulation time. B. Hydrogen bonds existence map of AhR-CLF complex as a function of time. C. Hydrogen bonds existence map of AhR-INDR complex as a function of time. D. Hydrogen bonds existence map of AhR-Kyn complex as a function of time. E. Average percentage of secondary structures in AhR in the absence and presence of CLF, INDR, or Kyn. CLF is clofazimine, INDR is indirubin, and Kyn is kynurenine.

Figure 5.. Free energy landscapes of CLF and Kyn in complex with AhR.

A. Two-dimensional projection of eigenvectors of apo AhR, AhR-CLF complex, AhR-INDR complex, and AhR-Kyn complex. B. Three-dimensional plot of the free energy landscape of apo AhR. C. Three-dimensional plot of the free energy landscape of AhR-CLF complex. D. Three-dimensional plot of the free energy landscape of AhR-INDR complex. E. Three-dimensional plot of free energy landscape of AhR-Kyn complex. F. MM-PBSA binding energies for the interaction of CLF or INDR with AhR. CLF is clofazimine, INDR is indirubin, and Kyn is kynurenine.

Figure 6.. CLF treatment blocked the Kyn-mediated cellular proliferation, migration, and colonization in SBC3 and SBC5 cell lines.

A. MTT assay was used to measure the effect of Kyn on cell proliferation. 50μM of Kyn treatments showed a significant increase in the proliferation of SCLC cell lines. B&C. Cytotoxicity profile of CLF has been shown through MTT assay on SBC3 and SBC5 cell lines in the presence and absence of 50μM Kyn. CLF substantially inhibits the Kyn-mediated proliferation in both the SCLC cell lines without showing any significant increase in the IC₅₀ values. The IC₅₀ value of the drug was calculated using GraphPad Prism^®7 software. D&E. Migration and colonization inhibiting capability of CLF in the presence of 50 μM Kyn has been studied by performing trans well migration and colony assays on SBC3 and SBC5 cell lines. Representative images (left panel) illustrating enhanced migration and colonization of SBC3 and SBC5 cells in the response to Kyn treatment. However, CLF treatment strongly inhibits both. Crystal violet was used to stain migrated cells with the imaging of five different areas at a magnification of 10x. Scale bar = 100μm. Representative bar graph for the quantification of cell numbers showing migration and colonization in control, 50 μM Kyn treated and 8 μM CLF with 50 μM Kyn treated samples (left panel). Data are shown as means ± SEM. F. The expression levels of oncogenic proteins involved in Akt/Erk signaling have been elucidated with western blotting of SBC3 and SBC5 cells after 48h treatment of 50 μM Kyn in the presence or absence of 8 μM CLF. β-Actin was used as a loading control. ****p<0.0001, ####p<0.0001 (compared to Kyn’s treatment).

Figure 7.. Effects of CLF in the presence of Kyn on inducing apoptosis and G2/M cell cycle arrest in small cell lung cancer cells.

A. The number of apoptotic cells was analyzed through flow cytometry analysis in SBC3 and SBC5 cell lines after their treatment with Kyn alone or in the presence of CLF for 48h (left panel). Quantification of apoptotic cells in control, 50μM Kyn treated and co-treatment of 8μM of CLF with Kyn for 48h. Student t-test has been used to identify statistical significance ****p<0.0001, and ####p<0.0001 (compared to Kyn’s treatment).B. SBC3 and SBC5 cell lines were subjected to DNA content analysis through FACS in various cell cycle phases by propidium iodide staining. The representative FACS histogram from three independent experiments shows the potential of CLF in the presence of Kyn to arrest the SCLC cells in the G2/M phase (left panel). A representative bar graph of quantitative analysis of DNA content was shown as mean ±SD (n=3) (right panel). The presence of CLF significantly arrests the cell cycle progression in the G2/M phase. C. Western blot analysis of Bcl-xl and cleaved caspase 7 shows that treatment of CLF in the presence of Kyn increased the expression of apoptotic marker protein with a significant reduction in the expression of antiapoptotic protein. Β-Actin was used as a loading control. One way ANOVA was employed for statistical significance **p<0.001, ****p<0.0001, ##p<0.001, ####p<0.0001 (compared to Kyn’s treatment).

Figure 8.. CLF inhibits the Kyn-mediated differentiation of osteoblasts.

A. Schematic representation of OBs culture, treatment, and collection of conditioned media. B & C. Kyn treatment in undifferentiated MCOs promotes osteoblast differentiation, which was inhibited by the treatment of 8 μM CLF. Undifferentiated MCOs were treated with 50 μM of Kyn in the presence and absence of 8μM CLF for 2 and 10 days and were checked for alkaline phosphatase activity and staining. D. Undifferentiated MCOs were treated with 50 μM of Kyn in the presence and absence of 8μM CLF for 48h in the differentiated medium of osteoblast, followed by the extraction of total RNA and RT-PCR analysis for osteogenic genes (ALP, Runx-2, Col-1a, and osteocalcin), and gene expression was normalized to that of β-actin. E. Representative images of Alizarin staining showing strong mineralization in the MCOs treated with 50 μM of Kyn. However, the presence of 8 μM CLF significantly inhibits the same. F. Quantification of alizarin staining in the form of OD values. G. Western blot analysis of RANKL expression in the MCOs treated with the 50 μM of Kyn in the presence and absence of 8 μM of CLF in the differentiated osteoblast medium for 48h. β-Actin was used as a loading control. Results are expressed as mean ±SD. Unpaired t-test and two-way ANOVA were used to measure statistical significance. ****p<0.0001, ####p<0.0001 (compared to Kyn’s treatment).

Figure 9.. CLF treatment inhibits the osteoblast-mediated osteoclast differentiation.

A. Schematic representation of OCs culture and treatment with OBs-CM. B. BMMs derived from C57BL/6J mice were subjected to culture in the presence of 40% CM derived from osteoblasts treated with 50 μM of Kyn alone or in the presence of 8 μM of CLF for 48h with MCSF. RNA was extracted, followed by an RT-PCR analysis of osteolytic genes like TRAP, Cathepsin K, NFATc1, c-Anhydrase, and c-fos. β-actin was used to normalize the gene expression. C. Representative images of TRAP after 7 days showing strong staining on BMMs treated with 40% condition medium of only 50 μM of Kyn treated cells as compared to control, while the treatment of condition medium derived from osteoblasts received of 8 μM of CLF along with 50 μM of Kyn significantly reduced the staining (left panel). Quantification of TRAP^+VE cells in the form of a bar graph. Data are shown as means ± SD. Two-way ANOVA and unpaired t-tests have been used for statistical significance. ***p<0.0004, ****p<0.0001, ####p<0.00001(compared to Kyn’s treatment).