Small molecule nAS‐E targeting cAMP response element binding protein (CREB) and CREB‐binding protein interaction inhibits breast cancer bone metastasis

Abstract Bone is the most common metastatic site for breast cancer. The excessive osteoclast activity in the metastatic bone lesions often produces osteolysis. The cyclic‐AMP (cAMP)‐response element binding protein (CREB) serves a variety of biological functions including the transformation and immortalization of breast cancer cells. In addition, evidence has shown that CREB plays a key role in osteoclastgenesis and bone resorption. Small organic molecules with good pharmacokinetic properties and specificity, targeting CREB‐CBP (CREB‐binding protein) interaction to inhibit CREB‐mediated gene transcription have attracted more considerations as cancer therapeutics. We recently identified naphthol AS‐E (nAS‐E) as a cell‐permeable inhibitor of CREB‐mediated gene transcription through inhibiting CREB‐CBP interaction. In this study, we tested the effect of nAS‐E on breast cancer cell proliferation, survival, migration as well as osteoclast formation and bone resorption in vitro for the first time. Our results demonstrated that nAS‐E inhibited breast cancer cell proliferation, migration, survival and suppressed osteoclast differentiation as well as bone resorption through inhibiting CREB‐CBP interaction. In addition, the in vivo effect of nAS‐E in protecting against breast cancer‐induced osteolysis was evaluated. Our results indicated that nAS‐E could reverse bone loss induced by MDA‐MB‐231 tumour. These results suggest that small molecules targeting CREB‐CBP interaction to inhibit CREB‐mediated gene transcription might be a potential approach for the treatment of breast cancer bone metastasis.


| INTRODUCTION
Breast cancer is the most commonly diagnosed cancer among females. 1 At least 80% of the breast cancer patients will develop bone metastases during the course of their diseases. 2,3 In the metastatic bone lesions, there are complex cross-talks between metastatic breast cancer cells and bone cells. [4][5][6][7] For example, breast cancer cells produce various secreted factors [8][9][10][11] in the bone microenvironment, which enhance osteoclastogenesis and inhibit osteoblastogenesis, Patients who develop bone metastases suffer from skeletal-related events (SREs) 12 such as pathological fracture, spinal cord compression, bone pain and hypercalcemia. Because breast cancer bone metastasis has a generally osteolytic nature, 13,14 inhibitors of osteoclastic bone resorption including bisphosphonates and denosumab have become the most commonly used medications in the treatment of breast cancer-induced osteolysis. [15][16][17] However, their effects are relatively moderate in clinical studies. 18,19 It is also reported that bisphosphonates have serious adverse effects such as osteonecrosis of the jaw and subtrochanteric fractures. 20 Denosumab, a monoclonal antibody against receptor activator of NF-kB ligand (RANKL), has been shown to increase risk of pancreatitis and serious infections including endocarditis, erysipelas and infectious arthritis. 21 Therefore, novel agents have been pursued to mitigate osteolysis induced by breast cancer metastasis. Cathepsin K inhibitor odanacatib has been reported to be an agent against osteoporosis and breast cancer-induced bone metastasis. 22,23 However, odanacatib was withdrawn from the clinical trials for safety reasons. In addition, clinical studies of c-Src inhibitors have been initiated in patients with bone metastases, since c-Src plays multiple roles in the bone resorption and in the proliferation, survival, metastasis of breast cancer cells. 24,25 Unfortunately, the efficiency of c-Src inhibitors has not been generally encouraging in clinical trials to date. Hence, other safe and efficient targets and new strategies in the treatment of breast cancer bone metastasis are needed.
The cyclic-AMP (cAMP)-response element binding protein (CREB) is a nuclear transcription factor belonging to a family of basic leucine zipper (bZIP)-containing transcription factors that serves a variety of biological functions including cellular proliferation and differentiation. [26][27][28] Recently, accumulating evidence has revealed that CREB participates in the immortalization and transformation of cancer cells.
It is also reported that CREB and phosphorylated CREB (p-CREB) have been shown to be consistently over-expressed in breast cancer tissues. [29][30][31][32] CREB regulates a number of critical genes involved in cellular proliferation, anti-apoptosis and metastasis of breast cancer. 33,34 These include b-cell lymphoma 2 (Bcl-2), vascular endothelial growth factor and type IV collagenase matrix metalloproteinase 2.
On the other hand, evidence has also shown that CREB plays a key role in the differentiation of osteoclast and bone resorption. 35,36 CREB activation is required for the inductions of c-Fos and nuclear factor of activated T-cells cytoplasmic 1 (NFATc1), which control the expressions of bone resorption enzymes such as tartrate-resistant acid phosphatase (TRAP), matrix metalloprotein-9 and cathepsin K. In the early-phase of osteoclast differentiation, the expression of c-Fos is induced after CREB activated is phosphorylated and activated by Calcium/calmodulin-dependent protein kinase IV, which is induced by Ca 2+ signalling. In the late-phase of osteoclast differentiation, NFATc1 is dephosphorylated by calcineurin which is also induced by Ca 2+ signalling pathway, translocates into the nucleus and interacts with p-CREB to activate the transcription of osteoclast-specific genes. 37 When A-CREB, a dominant-negative form of CREB, was introduced into osteoclast precursor cells, the formation of TRAPpositive osteoclasts and the expression of NFATc1 were significantly reduced. 37 Drugs serving as "dual inhibitors" for the prevention of breast cancer-induced bone metastasis and osteolysis are expected to improve therapeutic efficacy. Considering the key roles of CREB in breast cancer and osteoclastogenesis, we proposed inhibition of CREB activity as an intriguing strategy for the development of novel breast cancer bone metastases therapeutics. Biological approaches were pursued to inhibit CREB function in breast cancer cells such as the utilization of dominant-negative CREB mutants, CREB "decoy" oligonucleotides and RNA interference. [38][39][40][41] However, clinical applications of these approaches are rather limited since gene therapy techniques would also be required. 42 Therefore, utilizing small organic molecules to prevent breast cancer-induced bone metastasis and osteolysis could be of great interest due to their better pharmacokinetic properties. There are three potential intervention points for small molecules as chemical inhibitors of CREB-mediated gene transcription. The first approach, targeting CREB-related kinases involves in the use of kinase inhibitors to inhibit CREB phosphorylation and its transcription. Unfortunately, this approach elicits many off-target effects. 43 The second one is to inhibit CREB-CRE interaction. It is reported that NSC 12155 and NSC 45576 were identified as inhibitors of CREB-CRE interaction by high-throughput screening assay from the NCI-diversity set of 1900 compounds. However, these compounds are not specific in inhibiting CREB-CRE interaction. 44 The third strategy is to target CREB-binding protein (CREB-CBP) interaction to inhibit CREB-mediated gene transcription in breast cancer cells and osteoclasts.
CREB is not activated until it is phosphorylated at Ser133 and its subsequent binding to CBP through kinase-inducible domain (KID) in CREB and KID-interacting (KIX) domain in CBP. The binding interface between the KID-KIX interaction is structurally wellcharacterized by NMR spectroscopy. 45 In our previous study, we described naphthol AS-E (nAS-E) as a cell-permeable small molecule inhibitors of KIX-KID interaction to inhibit CREB-mediated gene transcription based on a renilla luciferase complementation assay in HEK293T cells. 46,47 In this study, we further verified the binding modes of nAS-E in KIX-KID interface by molecular docking and investigated the inhibition of nAS-E on CREB-CBP interaction in osteoclasts. We also studied its inhibition effects on the prolifera-

| Isolation of BMMs
Bone marrow monocytes for osteoclast differentiation were isolated from 4-week-old male C57/BL6 mouse as previously described. 48 Briefly, all the bone marrow cells in the femur and tibiae of mouse

| Molecular modelling
Molecular docking was operated with Autoock Vina with default settings. The protein structure was prepared using Chimera (https:// www.rbvi.ucsf.edu/chimera) 49 and molecular docking was performed with Autoock Vina with default settings. 50 The docking results were viewed using Pymol0.99rc6. The NMR structures of CREB-CBP complex are downloaded from www.RCSB.org (PDB code: 1KDX). Chain A is KIX domain of CBP and chain B is KID. By analysing the interaction interface of KIX-KID, a relatively deep pocket was found on the surface of KIX structure (state 8 of NMR structures) close to residues ILE137 and Leu141 of KID.

| Western blotting
Cells were treated as indicated, then washed and lysed. The lysates were subsequently subjected to SDS-PAGE. Immunoblot analyses were performed with primary antibodies purchased from CST. The secondary antibodies were visualized using an Immobilon Western kit (Millipore).
The cell lysates were then centrifuged at 16 000 g for 30 minutes at 4°C and the supernatants were incubated with p-CREB antibodies overnight, followed by incubation with Protein A/G-coated agarose beads (Merck) for another 4 hours at 4°C. Then samples were washed with cold IP buffer three times and the supernatants were removed by centrifugation at 2000 g for 1 minute. The proteins were then separated from the beads using IB loading buffer for 5 minutes at 95°C. The supernatants were collected and were detected with indicated antibodies by blotting and re-blotting.

| MTT assay
The proliferation effects of nAS-E on different cells were determined by MTT assay. Cells were plated to 40%-50% confluence per well in 96-well plates in full medium and cultured overnight. Then, the cells were treated with indicated concentrations of nAS-E in triplicate for indicated hours. After that, 100 μL MTT was added to each well and the plates were incubated at 37°C for 2 hours. Then, the supernatant was removed and the crystals were dissolved in 100 μL DMSO. Optical density (OD) was measured with an Infinite F200 PRO absorbance microplate reader (Tecan) at 570 nm. Cell viability was calculated relative to the control.

| Cell migration assay
To assess cell migration potential, 5 × 10 4 cells in 100 μL of serumfree medium were plated in the upper chamber of transwell migration chambers (0.8 μm pore; Millipore). The lower chamber was filled with 500 μL of serum-free medium with the indicated nAS-E.

| Tartrate-resistant acid phosphatase colorimetric assay
Bone marrow monocytes were seeded in 96-well plates at a density of 3 × 10 3 cells/well. Cells were exposed to in full medium plus 30 ng/mL M-CSF and 50 ng/mL RANKL in combinations with indicated concentrations of compounds in triplicate for 2 days. Cells were lysed and TRAP activity was measured by TRAP activity assay kit according to the manufacturer's instructions. Briefly, the cell culture media were removed and cells were washed by PBS for three times.
Then, cells were lysed by passive lysis buffer (Promega) for 15 minutes at 37°C and the supernatant were collected and incubated with para-nitrophenylphosphate (p-NPP) in the presence of disodium tartrate for 45 minutes. The reaction was subsequently stopped with the addition of NaOH. The TRAP activity was then quantified by measuring OD at 405 nm with Tecan absorbance microplate reader.

| In vitro osteoclastogenesis assay
Bone marrow monocytes were cultured in 96-well plates in full medium containing M-CSF and allowed to adhere overnight. The medium was replaced and the cells were treated with 30 ng/mL M-CSF and 50 ng/ mL RANKL for additional 5 days. The medium was replaced every 2 days. TRAP staining was then performed with a leucocyte acid phosphatase kit (Sigma) according to the manufacturer's instructions. TRAPpositive multinucleated cells (≥3 nuclei) were scored as osteoclasts.

| Bone resorption assay
For the bone resorption assay, BMMs were seeded on bovine femur bone slices in 96-well plates and the next day cells were induced with 30 ng/mL M-CSF and 50 ng/mL RANKL in combination with indicated concentrations of compound for 5 days. The medium was replaced for each 2 days. Cells were then fixed with 2.5% glutaraldehyde. Bone slices were stained by 0.5% toluidine blue and then imaged using an Olympus microscope with 200× magnification.
Three fields were randomly selected for each bone slice for further pit area analyses which were quantified using Image J software.

| RNA isolation and quantitative real-time PCR
Total RNA extraction and mRNA expression analysis by quantitative real-time (qRT)-PCR were performed as described previously. mRNA levels of TRAP, V-ATP, NFATc1, cathepsin K and β-actin were quantified by qRT-PCR using specific primers. The primer sequences were summarized below ( Table 1). All values were reported as mean ± SD of triple measurements of each cDNA sample. mRNA levels were normalized to β-actin mRNA.

| Statistical analysis
All data were expressed as mean ± SD and presented as the mean of triplicate points. One-way ANOVA and two-tailed non-paired Student's t test were used to compare differences, and statistical significance was displayed as *(#) P < 0.05 **(##) P < 0.01 or ***(###) P < 0.001.

KIX-KID interface identified by molecular docking
The chemical structure of nAS-E was shown in Figure 1A. In our previous study, we had shown nAS-E as an inhibitor of the KIX-KID interaction and CREB-mediated gene transcription. 47 To understand the basis of the inhibition of KIX-KID interaction by nAS-E, a molecular docking simulation against KID binding interface on KIX was performed to identify the binding mode of nAS-E ( Figure 1B,C). In the final docked structure, the naphthol moiety of nAS-E was tightly confined within a hydrophobic pocket lined by residues His602, Lys606, Arg646 and Tyr650, and formed T-shape π-π interaction with His602 and Tyr650 as shown in Figure 1C. N-phenylacetamide interacted with Lys606 through cation-π and hydrogen bond interaction. Electrostatic interaction was observed between chlorine and Gln661.

| nAS-E regulated the proliferation, apoptosis and migration of MDA-MB-231 breast cancer cells in vitro
To determine the inhibitory effects of nAS-E on breast cancer cells,

| nAS-E inhibited RANKL-induced osteoclastogenesis and bone resorption
Tartrate-resistant acid phosphatase (TRAP) is an established marker for osteoclastogenesis and its activity was highly related with osteoclast differentiation and bone resorption. TRAP activity assay was then conducted to investigate the activity of nAS-E in BMM osteoclastgenesis induced by M-CSF and RANKL. As shown in Figure 4A, nAS-E strongly inhibited TRAP activity in a dose-dependent manner at the concentrations of 1, 5 and 10 μmol/L without any cytotoxicity as illustrated in MTT assay ( Figure 4B). Then, we examined the

| nAS-E inhibits osteoclast-specific genes expression and c-Fos/NFATc1 signalling pathway in vitro
RT-PCR was used to assess the effect of nAS-E on RANKL-induced osteoclast-specific gene expressions during osteoclastogenesis. In the assay, BMMs were treated with M-CSF and/or RANKL alone or with nAS-E for 3 days. As shown in Figure 5A

| nAS-E inhibited breast cancer cell-induced osteolysis in vivo
Up to now, we have established that nAS-E had the potential to inhibit osteoclastogenesis and prevent breast cancer-induced osteoclastgenesis through interrupting CREB-CBP interaction. Next, we

| DISCUSSION
Breast cancer is one of the most common cancers. Most patients with advanced breast cancer normally develop osteolytic bone metastasis, which are a common cause of morbidity and mortality.
There are several agents used to delay the occurrence of bone metastases and reduce the risk of SREs, but they have many limitations and their effects are palliative rather than curative. 18 Figure 1B,C). The binding mode of nAS-E could be also used to guide further structure optimization. To further verify its mechanism, we used Co-IP assay to find out whether nAS-

E interrupted p-CREB and CBP interaction in BMMs. Our results
showed that nAS-E significantly interrupted p-CREB and CBP interaction without altering the levels of p-CREB, which indicated that nAS-E might not exert off-target effects like the CREB-related kinase inhibitors.
We also provided evidence that nAS-E has no cytotoxic effect to AS-E effectively inhibited osteoclastogenesis and prevented breast cancer-induced osteoclastgenesis through interrupting CREB-CBP interaction. Our study provided the first demonstration that small molecules targeting CREB-CBP interaction represents a novel and promising treatment for the breast cancer-induced bone loss. It is expected that our results may help for developing novel CREB-CBP interaction therapeutic agents to treat breast cancer bone metastasis.