Auranofin inhibition of thioredoxin reductase sensitizes lung neuroendocrine tumor cells (NETs) and small cell lung cancer (SCLC) cells to sorafenib as well as inhibiting SCLC xenograft growth

Abstract

Thioredoxin Reductase (TrxR) functions to recycle thioredoxin (Trx) during hydroperoxide metabolism mediated by peroxiredoxins and is currently being targeted using the FDA-approved anti-rheumatic drug, auranofin (AF), to selectively sensitize cancer cells to therapy. AF treatment decreased TrxR activity and clonogenic survival in small cell lung cancer (SCLC) cell lines (DMS273 and DMS53) as well as the H727 atypical lung carcinoid cell line. AF treatment also significantly sensitized DMS273 and H727 cell lines in vitro to sorafenib, an FDA-approved multi-kinase inhibitor that depleted intracellular glutathione (GSH). The pharmacokinetic, pharmacodynamic, and safety profile of AF was examined in nude mice with DMS273 xenografts administered AF intraperitoneally at 2 mg/kg or 4 mg/kg (IP) once (QD) or twice daily (BID) for 1-5 d. Plasma levels of AF were 10-20 μM (determined by mass spectrometry of gold), and the optimal inhibition of TrxR activity was obtained at 4 mg/kg once daily, with no effect on glutathione peroxidase 1 activity. This AF treatment extended for 14 d, inhibited TrxR (>75%), and resulted in a significant prolongation of median overall survival from 19 to 23 d (p = .04, N = 30 controls, 28 AF). In this experiment, there were no observed changes in animal bodyweight, complete blood counts (CBCs), bone marrow toxicity, blood urea nitrogen, or creatinine. These results support the hypothesis that AF effectively inhibits TrxR both in vitro and in vivo in SCLC, sensitizes NETs and SCLC to sorafenib, and could be repurposed as an adjuvant therapy with targeted agents that induce disruptions in thiol metabolism.

Keywords: auranofin; dms273; glutathione; h727; lung neuroendocrine tumor; small cell lung cancer; sorafenib; thioredoxin reductase.

Figure 1.

AF decreases TrxR activity and clonogenic survival of lung NET and NECs. AF enhances clonogenic cell death with sorafenib treatment. (a) Small cell lung cancer DMS273 and DMS53 and bronchial carcinoid H727 cell line were treated with 1 µM AF for 24 h (gray bars) and then collected for TxrR enzyme activity. All activity was normalized to H727 control cells. (b) Same cell lines and treatment as in (a) with the clonogenic assay. Normalized to each cell line’s respective control. * p < .05 compared to untreated cells two-way ANOVA with Fishers LSD n = 3. (c) Exponentially growing DMS273 and H727 cells were treated for 48 h with sorafenib at 2.5 or 5 µM followed by harvest and glutathione assay. (d,e) DMS273 and H727 cells were treated with 48 h sorafenib and the last 24 h with 1 µM AF followed by clonogenic assay. Clonogenic survival colony counts were normalized to killing with the treatment of AF alone. n ≥ 3 independent experiments two-way ANOVA Fishers LSD * significantly different than control (p < .05), # significantly different than either drug alone (p < .05).

Figure 2.

Dose escalation of AF selectively decreases TrxR activity in SCLC xenografts, kidneys, and liver but does not affect Gpx1 activity. (a, b, c) Female athymic nude mice (6 per group) were xenografted with 5 × 10⁶ DMS273 cells (2 tumors/mouse one on each flank) and given AF IP injections for 24 h or 5 d either QD or BID. Kidney, liver, and tumor samples were collected for TrxR activity. (d) Plasma samples were collected 9 h after the final BID dose (5 d of therapy), and gold content was determined. (e) Gpx1 activity was determined in xenograft tumors after 5 d of AF therapy. * = p-value <.05, ** = p-value <.01, *** = p-value <.001. Analyzed by mix-effects ANOVA with Fishers LSD comparisons.

Figure 3.

AF decreases TrxR activity in DMS273 tumors in vivo independent of tumor size. Two mice each with two flank DMS273 xenografts in the first cohort and four mice with one flank DMS273 xenografts in the third cohort were treated with AF 4 mg/kg IP QD for 14 d and then euthanized and evaluated for TrxR enzyme activity assays, and results are plotted in aggregate (a) or as a function of tumor size at the end of treatment (b). * p < .05 compared to untreated cells two-way ANOVA with Fishers LSD comparisons.

Figure 4.

4 mg/kg QD AF IP for 14 d is nontoxic to bone marrow, liver, and kidneys of mice. Female nude mice-bearing DMS273 xenografts were treated with either vehicle or 4 mg/kg AF IP every day for 14 d. Mice were euthanized when tumors were ≥1000 mm³. This experiment was done on three separate cohorts (9–10 animals/cohort) of animals for a total of 28 AF-treated mice and 30 vehicle control-treated mice. (a) Mice were monitored daily, and the % of the initial weight was plotted as a function of time from the beginning of AF treatment. (b, c) Bone marrow from 9 AF-treated mice (5 from the first cohort, 4 from the second cohort) and 13 control mice (7 from the first cohort, 6 from the second cohort) was harvested from the femurs at euthanasia and subjected to flow cytometry for HSCs (lineage (Lin)− sca-1+ c-Kit+, LSK) shown in (b) and HSPCs (lin- cKit+ Sca1+ CD135-) populations shown in (c). (d-I) Blood was drawn from seven control mice and five AF-treated mice from the third cohort) via cardiac puncture at euthanasia and evaluated for clinical chemistry endpoints by Antech Diagnostics.

Figure 5.

4 mg/kg QD AF IP for 14 d prolongs survival in nude mice with DMS273 xenografts. Tumors from all mice shown in Figure 4a were measured via calipers, and tumor volumes were calculated. When tumor volumes measured 1000 mm³, mice were considered to have reached euthanasia criteria and Kaplan−Meier curves (three cohorts combined) were used to estimate the overall survival. Log-rank (Mantel Cox) test was used to determine significance (p < .04).