The selective inhibitor of nuclear export (SINE) verdinexor exhibits biologic activity against canine osteosarcoma cell lines

Abstract Verdinexor (KPT‐335) is a novel orally bioavailable selective inhibitor of nuclear export (SINE) compound that inhibits the function of the nuclear export protein Exportin 1 (XPO1/CRM1). In the present study, we sought to characterize the expression of XPO1 in primary canine osteosarcoma (OS) tumour samples, OS cell lines and normal osteoblasts and evaluate the in vitro activity of verdinexor alone or in combination with doxorubicin. Canine OS cell lines and a subset of primary OS tumours showed increased XPO1 transcript and protein expression as compared with normal canine osteoblast cells. All canine OS cell lines exhibited dose‐dependent growth inhibition and increased caspase 3,7 activity in response to low nanomolar concentrations of verdinexor (IC50 concentrations ranging from 21 to 74 nM). Notably, growth inhibition of normal canine osteoblast cell lines treated with verdinexor was observed at high micromolar concentrations (IC50 = 21 μM). The combination of verdinexor and doxorubicin resulted in potent inhibition of cell viability and demonstrated synergetic activity in three canine OS cell lines. Concordantly, OS cell lines showed increased γH2A.X foci following treatment with doxorubicin and recovery in verdinexor compared with cells treated with doxorubicin and recovered in normal media for 24 hours. These findings demonstrate that verdinexor has biologic activity against canine OS cell lines at physiologically relevant doses and suggest that XPO1 inhibition in combination with standard doxorubicin treatment offers promising potential for chemotherapeutic intervention in canine OS.


| INTRODUCTION
With an estimated incidence of >10 000 new cases per year, osteosarcoma (OS) is the most common primary bone tumour in dogs. 1 Despite aggressive treatment involving limb amputation and systemic chemotherapy, 90% of dogs succumb to chemotherapeutic resistant metastatic disease. 2 Similar to canine OS, 30% to 40% of humans with OS die from metastatic disease despite treatment. [3][4][5] Targeted therapeutics and immunotherapy have revolutionized survival outcomes in a variety of cancers; however, to date, these have not translated in improved outcomes in either human or canine OS patients. [6][7][8][9][10][11][12][13] As systemic chemotherapy remains the backbone for treatment of OS metastases, the development of combinational treatments with novel and more effective agents is necessary for this disease.
The exportin-1 protein (XPO1 or CRM1) functions in mammals to regulate the transport of RNA and protein via the nuclear-cytoplasmic transport system. 14,15 XPO1 is the major nuclear export receptor from the Karyopherin β protein family that interacts with the leucine-rich nuclear export signal (NES) of cargo proteins for nuclear-cytoplasmic export. Transport of the XPO1-cargo protein complex is mediated by RanGTP which is bound with the XPO1-cargo protein complex and passively diffuses into the cytoplasm through the nuclear pore complex down a RanGTP gradient. 16 Dysregulation of this process has been implicated in tumorigenesis as aberrant trafficking of tumour suppressor and regulatory proteins modifies apoptosis and cell proliferation pathways. Of the~200 NES containing cargo proteins shuttled by XPO1, the tumour suppressor and regulatory growth proteins p53, p21, RB, survivin and FOXO are several involved in this imbalance. 17 Increased concentrations of XPO1 have been identified in a number of solid and hematologic malignancies and are associated with poor prognosis. [18][19][20][21][22] As such, restoration of nuclear-cytoplasmic protein homeostasis via XPO1 inhibition may represent a viable therapeutic strategy for human and canine cancers.
Selinexor and verdinexor, two selective inhibitors of nuclear export (SINE) compounds that specifically inhibit the binding of cargo proteins to XPO1, have shown clinical benefit in human and canine cancer patients, respectively. 17,23 Selinexor (tradename XPOVIO) was recently granted approval by the FDA for the treatment of patients with relapsed or refractory multiple myeloma in combination with dexamethasone. 24 In addition to its efficacy in the treatment of hematologic malignancies, selinexor has shown in-vitro biological activity in human OS cell lines and robust in vivo antitumor activity was observed in a variety of preclinical sarcoma animal models. 25,26 In a phase IB clinical trial evaluating single-agent treatment with selinexor in patients with refractory bone or soft tissue sarcomas, 33% of patients experienced stable disease for 4 or more months. 27 Increased expression of XPO1 in OS tissues was correlated with a reduced progression-free survival and overall survival, suggesting that therapeutics targeting XPO1 may be effective in OS treatment. 19 Prior studies investigating the in vitro activity of verdinexor in canine cancer cells demonstrate that verdinexor inhibits cell proliferation in canine lymphoma, melanoma, transitional cell carcinoma, and mammary carcinoma cell lines. [28][29][30] A phase I clinical trial in dogs with spontaneous cancers established safe dosing regimens at physiological achievable concentrations of verdinexor. 29 Subsequent phase II trials evaluating treatment with verdinexor in dogs with non-Hodgkin lymphoma found that verdinexor exhibited biologic activity with an overall objective response rate of 37%. 23 While these studies with verdinexor show promise for its use in the treatment of canine lymphoma and have provided important pharmacokinetic data in canines, the potential efficacy of verdinexor in the treatment of canine OS remains unclear.
The purpose of this study was to confirm the relevance of XPO1 as therapeutic target in canine OS and evaluate the activity of verdinexor alone and in combination with the cytotoxic chemotherapeutic drug, doxorubicin in canine OS cells. These data will serve as the foundation for potential future clinical studies investigating verdinexor in dogs with OS.

| Cell line validation statement
Six cell lines were used in this study, and all cell lines were previously established and validated. Two cell lines were purchased directly from commercial culture collections, as outlined above. Cells were not subjected to species verification testing; however, morphology under light microscopy and growth kinetics remained consistent throughout the experiments.
2.4 | RNA isolation, cDNA synthesis, and qRT-PCR RNA was extracted from canine OS cell lines and tumours using TRIzol reagent (Invitrogen) and purified using the RNeasy Mini Kit (Qiagen, Cat.#74 104) according to manufacturer's instructions. cDNA was made from 2 μg of total RNA using Superscript III (Invitrogen).
Published canine XPO1 primers and 18S primers were used with Fast SYBR Green Master Mix (Applied Biosciences) and qRT-PCR was performed using the Applied Biosystems StepOne Plus Detection System (Applied Biosystems, Foster City, California). 30 Relative expression of XPO1 was determined using the comparative threshold cycle method and normalized to the endogenous housekeeping gene 18S. 31 All reactions were performed in triplicate and three independent experiments were performed in OS cell lines. Drug combination studies were completed with verdinexor and doxorubicin in the D17, OSCA-40, and Abrams cell lines using the fixed-ratio method. Combination index (CI) was calculated with the CompuSyn software (ComboSyn, Inc.) to determine synergistic effects using the median-effect method of Chou and Talalay. 32 CI = 1, additivity; CI < 1, synergism; CI > 1, antagonism. γH2A.X fluorescence was detected by immunofluorescence microscopy using a Nikon Ti2E microscope fitted with a Prime 95B camera.

| Statistics
Data analysis was performed using the software program GraphPad  Concordantly, canine OS cells displayed enhanced sensitivity to verdinexor when compared with normal osteoblasts (Figures 2 and 3).   I G U R E 3 Verdinexor induces apoptosis in canine osteosarcoma (OS) cell lines. Canine OS cell lines and normal canine osteoblasts were treated with 0.1% DMSO, 1 μM, or 10 μM for 48 hours and caspase 3/7 activity was measured using the SensoLyte Homogeneous AMC Caspase -3/7 Assay Kit. Presented data are an average of four replicates from four independent experiments ±SD. Data were analysed using a one-way ANOVA with Dunnett's multiple comparisons test. *P < .05, **P < .01, ***P < .001, ****P < .0001 calculated using the Chou-Talalay method. 32

| Verdinexor impairs DNA damage repair caused by doxorubicin treatment
To further elucidate a mechanism by which verdinexor and doxorubicin work synergistically to reduce OS cell proliferation, the canine D-

| DISCUSSION
There is significant need for effective therapeutic strategies in the treatment of canine OS as >90% of dogs succumb to metastatic disease that is refractory to chemotherapy. Recent attempts to improve outcomes in OS patients with single-agent targeted molecular therapeutics and immunotherapy has proved largely unrewarding. 6,13 As systemic chemotherapy remains the standard treatment for metastatic disease, novel therapeutic strategies such as the addition of small molecule inhibitors to cytotoxic chemotherapy may enhance their overall efficacy, ultimately improving survival outcomes. One family of novel compounds, termed SINE, inhibit the major nuclear export protein Exportin 1 (XPO1), which is frequently overexpressed in cancer. 14,16-22 SINE compounds work synergistically with standard cytotoxic agents to enhance cancer cell death. 37,39 Verdinexor (KPT-335) is a SINE compound developed for companion animal use that is safe and tolerable in canines at a biologically active dose of 1.5 mg/kg F I G U R E 4 Effects of verdinexor on XPO1 mRNA and protein expression in canine osteosarcoma (OS) cell lines. Canine OS cell lines were treated with 0.1% DMSO, 0.1 μM, or 1 μM for 24 hours and RNA and protein was collected. (A) XPO1 expression measured via quantitative realtime PCR in canine OS cells treated with verdinexor for 24 hours. Data were analysed using a one-way ANOVA with Dunnett's multiple comparisons test. *P < .05, **P < .01, ***P < .001, ****P < .0001. (B) Western blotting for XPO1 was performed on canine OS cells treated with verdinexor for 24 hours. Error bars = SD three times a week. 23 Preliminary evidence of the in vitro activity of verdinexor has been documented in several canine cancer cell lines and verdinexor exhibits biologic activity in canine lymphoma patients; however, XPO1 expression and therapeutic targeting of XPO1 in canine OS has not been evaluated. 23,28,29 We therefore sought to To determine the effects of verdinexor on XPO1 inhibition in canine OS cells, we evaluated XPO1 transcript and protein expression 24 hours following treatment with verdinexor. We observed decreased XPO1 expression in all four cell lines at 0.1 and 1 μM F I G U R E 6 Verdinexor inhibits repair of DNA damage induced by doxorubicin in canine osteosarcoma (OS) cell lines. γH2A.X expression, a marker for DNA damage, was detected via immunofluorescence in OSCA-40 and D-17 cells. (A) Cells were either untreated (control) or treated with 1 μM doxorubicin for 2 hours. Cells treated with doxorubicin were washed and allowed to recover in normal media (1 μM Doxo + Recovery) or 1 μM verdinexor (Combo) for 24 hours. Data were analysed using a two-tailed Student's t-test. ****P < .0001 and presented as mean ± SD. (B) Representative semi-quantitative images of γH2A.X expression. DNA damage because of doxorubicin persisted longer when recovered in verdinexor. Verdinexor alone did not induce DNA damage verdinexor when compared with vehicle treated cells with the exception of the Abrams cell line. At 0.1 μM, Abrams cells expressed more XPO1 than vehicle treated cells; however, at 1.0 μM, XPO1 expression was reduced. Prior in vitro studies have documented that cellular exposure to SINE compounds leads to a time-dependent accumulation of nuclear proteins; however, the nuclear protein abundance progressively decreases as a consequence of enhanced ubiquitination and proteasome-dependent XPO1 degradation. 15 As such, it is possible that in Abrams cells treated with verdinexor at 0.1 μM, XPO1 function is inhibited as evidenced by a compensatory increased in XPO1 mRNA, but not ubiquinated for proteasome degradation. XPO1 transcript was elevated in all cell lines at both concentrations of verdinexor consistent with a compensatory response to XPO1 inhibition, a finding similarly seen in canine melanoma cells. 28 To determine if a combinational approach would enhance the chemotherapeutic efficacy of doxorubicin, a topoisomerase II inhibitor routinely used in the treatment of canine and human OS, OS cell lines were treated with verdinexor and doxorubicin and potential synergism was assessed. Studies performed in human multiple myeloma and acute myeloid leukaemia cell lines demonstrate therapeutic synergistic effects using a combination of selinexor and topoisomerase II inhibitors. [37][38][39] Concordantly, we found that combination treatment with verdinexor and doxorubicin in three canine OS cell lines worked synergistically, causing a significant decrease in cell proliferation compared with either drug alone. Reported mechanisms for this synergism include enhanced DNA damage through nuclear retention of topisomerase IIα (TopoIIα) and reduction of DNA damage repair proteins as a result of XPO1 inhibition. [37][38][39]41 Doxorubicin is a TopoIIα poison which prevents topoisomerase re-ligation of cleaved DNA resulting in double-stranded DNA breaks. 42 Via the forced nuclear retention of TopoIIα by XPO1 inhibition, the efficacy of doxorubicin is enhanced as a result of increased DNA damage leading to cellular apoptosis. 38,41 Alternatively, pre-treatment with DNA damaging agents like doxorubicin, followed by recovery in selinexor prolonged the DNA damage repair process resulting in increased cytotoxicity and apoptosis. 39 To evaluate the mechanism by which doxorubicin and verdinexor work synergistically in canine OS cells, cells were pre-treated with doxorubicin to induce DNA damage and recovered in media with or without verdinexor for 24 hours. To visualize DNA damage, we used the γH2A.X marker, a histone that is phosphorylated in response to double-stranded DNA breaks. 43 In health, γH2A.X forms within minutes of an insult to DNA and can recede to basal levels by 24 hours. Persistent γH2A.X may represent unrepairable doublestranded breaks. 43 When OS cells were treated with doxorubicin and recovered in normal media for 24 hours, γH2A.X returned to near basal levels. In contrast, when OS cells were recovered in verdinexor for 24 hours following treatment with doxorubicin, γH2A.X foci persisted, suggesting a delay or failure to repair damaged DNA. Importantly, treatment with verdinexor alone did not induce γH2A.X foci.
These findings are consistent with data generated in the U2OS human OS cell line. 39 Collectively, these findings suggest that verdinexor alters DNA damage repair processes in vitro and this may, in part, represent a mechanism by which doxorubicin and verdinexor work synergistically in canine OS cells.

| CONCLUSION
In summary, XPO1 is a relevant target for therapeutic intervention in canine OS as XPO1 is overexpressed in canine OS tumours and cell lines. Inhibition of XPO1 using the novel orally bioavailable XPO1