Topoisomerase inhibitors promote cancer cell motility via ROS-mediated activation of JAK2-STAT1-CXCL1 pathway

The human colorectal cancer cell lines LoVo, SW480, SW620, gastric cancer cell line AGS and small cell lung cancer cell line NCI-H446 were obtained from ATCC. Cell lines were maintained in RPMI-1640 medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal calf serum (Invitrogen). Cells were cultured at 37 °C with 5% CO₂.

Small interfering RNAs (siRNAs) were provided by GenePharma (Shanghai, China) and transfected into cells with siRNA Mate (GenePharma) following the provider’s instructions.

Interference sequences used were:

CXCL1 #1 -sense:

5′- CUCCAGUCAUUAUGUUAAUTT -3′;

CXCL1 #1 -antisense:

5′- AUUAACAUAAUGACUGGAGTT -3′;

CXCL1 #2 -sense:

5′-GCGGAAAGCUUGCCUCAAUTT -3′;

CXCL1 #2 -antisense:

5′-AUUGAGGCAAGCUUUCCGCTT -3′;

STAT1 #1 -sense:

5′-GCUGGAUGAUCAAUAUAGUTT-3′;

STAT1 #1 -antisense:

5′ -ACUAUAUUGAUCAUCCAGCTT -3′;

STAT1 #2 -sense:

5′- GUGGCAAAGAGUGAUCAGATT-3′;

STAT1 #2 -antisense:

5′ - UCUGAUCACUCUUUGCCACTT-3′;

STAT1 #3 -sense:

5′-GACCAUGCCUUUGGAAAGUTT -3′;

STAT1 #3 -antisense:

5′ -ACUUUCCAAAGGCAUGGUCTT-3′;

JAK2 #1 -sense:

5′-GGAUGGCAGUGUUAGAUAUTT -3′;

JAK2 #1 -antisense:

5′ -AUAUCUAACACUGCCAUCCTT -3′;

JAK2 #2 -sense:

5′-CCACCUGAAUGCAUUGAAATT -3′;

JAK2 #2 -antisense:

5′ -UUUCAAUGCAUUCAGGUGGTT -3′;

JAK2 #3 -sense:

5′- CCUGGUGAAAGUCCCAUAUTT-3′;

JAK2 #3 -antisense:

5′ -AUAUGGGACUUUCACCAGGTT − 3′;

Negative control -sense:

5′- UUCUCCGAACGUGUCACGUTT − 3′;

Negative control -antisense:

5′ -ACGUGACACGUUCGGAGAATT − 3′.

Cells were cultured in triplicate wells of 24-well plate (1.5 × 10⁴ cells/well). The confluences were quantified by the CloneSelect Imager (Molecular Devices, Sunnyvale, CA, USA) every 24 h.

Cell migration assay was performed in 24-well CIM plates (BD Biosciences, CA, USA). Briefly, 1–3 × 10⁴ cells per well were seeded in serum-free medium plus indicated drugs in the upper compartment of the CIM plates. Serum-complemented medium was added to the lower compartment of the chamber. After 24 h incubation, cells that passed through the septum were fixed with cold methanol and stained with crystal violet. The average number of migrated cells in four random microscopic fields was counted. Cell invasion assay was performed in 24-well CIM plates coated with matrigel. Other steps were identical to those of cell migration assay. VP-16 was provided by Hengrui Medicine (Jiangsu, China), ADM and CPT-11 were provided by Pfizer (New York, NY, USA). Fludarabine (HY-B0069), Bay11–7082 (HY-13453), C-176 (HY-112906), KU55933 (HY-12016) and AG490 (HY-12000) were purchased from MCE (Middlesex County, NJ, USA). GSH was purchased from Beyotime (Beijing, China).

Cells were homogenized in loading buffer (0.1 M Tris-HCl, pH 6.8, 1% SDS, 10% β-mercaptoethanol, 11% glycerol), separated by SDS-PAGE and electro-blotted to the nitrocellulose membranes, then blocked with 5% non-fat milk in TBS for 1.5 h at room temperature. The membranes were incubated with the indicated primary antibodies at 4 °C overnight. After washing three times with TBST, membranes were probed by horseradish peroxidase-labeled secondary antibodies for 45 min at room temperature. Protein bands were visualized by enhanced chemiluminescence. Antibodies against p-STAT1 (Tyr701) (#9167), p-JAK2 (T1007/1008) (#3776S) and JAK2 (#3230S) were from Cell Signaling (Boston, MA, USA). Anti-STAT1 (ab2415) and HRP-Protein A (ab7456) were from Abcam (Cambridge, UK). Anti-CXCL1 (100288-T36) was from Sino Biological (Beijing, China). Anti-oxPTP (MAB2844) was from R&D systems (Minneapolis, MN, USA). Anti-PTP1B (MABS197) was from EMD Millopore (Temecula, CA, USA). Anti-GAPDH (60004) was from Proteintech (Chicago, IL, USA).

Total RNA was extracted from cell pellets using Trizol reagent (Invitrogen) according to the manufacturer’s instructions. The RNA samples with an OD260/OD280 ratio between 1.9 and 2.0 were used for cDNA synthesis using High Capacity RNA to cDNA kits (Promega, Madison, WI, USA). Quantitative PCR was performed using SYBR Green master mix (Applied Biosystems) with the housekeeping gene GAPDH as the internal control. The relative expression of CXCL1 was calculated using the comparative Ct method. The primers of CXCL1 and GAPDH were as follows:

CXCL1 forward, 5′- AGCTTTGTTTAAACATGGCCCGCGCTGCTCTC-3′,

CXCL1 reverse, 5′- AGCTTTGTTTAAACCCCTTCTGGTCAGTTGGATTTG-3′;

GAPDH forward, 5′- GGAGCGAGATCCCTCCAAAAT-3′.

GAPDH reverse, 5′- GGCTGTTGTCATACTTCTCATGG-3′.

Animal study was approved by the independent ethics committee of Peking University Cancer Hospital. NOD/SCID mice (HFK Bio-Technology, Beijing, China) were maintained in accordance with the ethics standards of the World Medical Association (Declaration of Helsinki). TI-treated cells were injected to the caudal vein of 16–19 g female NOD/SCID mice as a 100 μL suspension (5 × 10⁵ cells). After 56 days, mice were sacrificed and lungs were stripped for analysis.

Lungs were immersed in 4% paraformaldehyde for 24 h and transferred to 60% ethanol. Individual lobes of lung biopsy material were placed in processing cassettes, dehydrated through the serial alcohol gradients, and embedded in paraffin. Before immunostaining, 5-μm thick lung tissue sections were dewaxed in xylene, rehydrated through decreasing concentrations of ethanol, and washed in PBS. Then sections were stained with hematoxylin and eosin. After staining, sections were dehydrated through increasing concentrations of ethanol and xylene.

Cancer cells grown in 100 mm-cell-culture dishes (80–90% confluence) were treated with VP-16 (20 μM) or DMSO in complete medium for 0.5 h. After removing culture medium and washing with PBS twice to clean the residual drug, cells were cultured in serum-free culture medium for 4 h. For each sample, conditioned medium from 12 dishes (totally 100 ml) was subjected to Mass spectrometric detection by PTM BIO (Hangzhou, China). Samples were sonicated three times on ice using a high intensity ultrasonic processor (Scientz) in lysis buffer (8 M urea plus 1% Protease Inhibitor Cocktail), followed by centrifugation at 12,000 g at 4 °C for 10 min. The protein concentration was determined with BCA kit according to the manufacturer’s instructions. For digestion, the samples were reduced with 5 mM DTT for 30 min at 56 °C and alkylated with 11 mM iodoacetamide for 15 min at room temperature in darkness. The samples were then diluted by adding 100 mM Triethylamonium bicarbonat (TEAB) to urea concentration less than 2 M. Finally, trypsin was added at 1:50 trypsin-to-protein mass ratio for the first digestion overnight and 1:100 trypsin-to-protein mass ratio for a second 4-h digestion. Next, peptides were desalted by Strata X C18 SPE column (Phenomenex) and vacuum-dried. Peptides were reconstituted in 0.5 M TEAB and processed with Tandem Mass Tag (TMT) kit. Briefly, one unit of TMT reagent was thawed and reconstituted in acetonitrile. The peptides were then incubated for 2 h at room temperature and pooled, desalted and dried by vacuum centrifugation. The peptides were subjected to nanospray ionization source followed by tandem mass spectrometry (MS/MS) in Q Exactive Plus (Thermo) coupled online to the UPLC. The electrospray voltage applied was 2.0 kV. The m/z scan range was 350 to 1800 for full scan, and intact peptides were detected in the Orbitrap at a resolution of 70,000. Peptides were then selected for MS/MS using normalized collision energy setting as 28 and the fragments were detected in the Orbitrap at a resolution of 17,500. A data-dependent procedure that alternated between one MS scan followed by 20 MS/MS scans with 15.0 s dynamic exclusion. Automatic gain control was set at 5E4. Fixed first mass was set as 100 m/z.

Intracellular ROS levels were evaluated using the Reactive Oxygen Species Assay Kit (Beyotime) as per manufacturer’s instruction. Briefly, cells were pulsed with 2,7-Dichlorodi -hydrofluorescein diacetate (DCFH-DA) probe (10 μM) in serum-free medium for 20 min. After washing three times with serum-free culture medium, the cells were treated with indicated agents or Rosup (positive control, 50 μg/ml) for 0.5 h. A Zeiss LSM780 confocal microscope (Carl-Zeiss, Oberkochen, Germany) was used to acquire images and the fluorescence intensity of DCFH-DA was quantified by ImageJ software.

Statistical analysis was performed with SPSS 20.0 or GraphPad Prism version 6.0. All values were represented as the Mean ± S.D. of three to four independent experiments with triplicate wells. The unpaired two-tailed t-test was used for in vitro study and one-way ANOVA was used for in vivo study. A two-sided p < 0.05 was considered statistically significant. ** p < 0.01; * p < 0.05; N.S. p > 0.05.

We firstly treated five cancer cell lines (LoVo, SW480, SW620, AGS, H446) with topoisomerase II inhibitors VP-16, ADM and topoisomerase I inhibitor CPT-11. Consistent with the anti-proliferative capacity [27‐29], all these agents inhibited cell growth (Additional file 1: Figure S1). Subsequently, we performed migration and invasion assays. Although three TI inhibited the motility of AGS cells (Additional file 1: Figure S2A), the motility of SW620 cells was not affected (Additional file 1: Figure S2B). Interestingly, these TI promoted migration and invasion of LoVo, SW480, and H446 cells in a dose-dependent manner (Fig. 1a, b). To evaluate the influence of TI on cancer cell motility in vivo, we injected VP-16 or CPT-11-treated LoVo cells into caudal vein of NOD/SCID mice. We noticed that TI-treated LoVo cells had stronger ability to form metastatic nodules on the surface of lungs (Fig. 1c). The Hematoxylin-eosin staining of sectioned lung tissues also showed that TI-treated groups had more metastases (Fig. 1d). These results suggest that VP-16, ADM and CPT-11 promote motility of a subset of cancer cells in vitro and in vivo.

VP-16, ADM and CPT-11 are potent genotoxic stress inducers and DNA damage partially contributes to the anti-proliferative effect of TI [32]. Recently, ATM, and NF-κB, and cGAS-cGAMP-STING pathways were implicated in DNA damage-promoted cell motility and metastasis [25, 33, 34]. To evaluate these pathways’ roles in TI-induced cancer motility, we utilized specific chemical inhibitors. We noticed that pretreatment with inhibitors to ATM, NF-κB, or STING did not prevent VP-16 or CPT-11-promoted migration (Additional file 1: Figure S3A-C), suggesting that TI-promoted motility is independent of the activation of ATM, and NF-κB, or cGAS-cGAMP-STING.

Unexpectedly, serum-free CM from TI-treated cells enhanced both migration and invasion of LoVo, SW480 and H446 cells (Fig. 2a, b). Albeit direct treatment with TI failed to increase the motility of AGS and SW620 cells (Additional file 1: Figure S2A-B), CM from TI-treated cells did enhance migration and invasion of these two cell lines (Fig. 2a, b). We then speculated that secreted factors form TI-treated cells may stimulate cell motility. To characterize the potential secreting factors, we carried out mass spectrometric analysis of CM from VP-16-treated responsive (LoVo, SW480) and unresponsive (SW620) cells. These three cell lines exhibited distinct profiles of protein up-regulation and down-regulation (Additional file 2: Tables S1-S3), while a subset of overlapping proteins was found (Fig. 2c). Among them, seven proteins (HMGB2, PRDX3, S100A7, RPS6, RPS25, NT5DC1, EIF3C) were down-regulated in LoVo and SW480 cells, meanwhile only two proteins (CXCL1, LRSAM1) were up-regulated in LoVo and SW480 cells, but not in SW620 cells (Fig. 2d). As CXCL1 is closely associated with metastasis [35], we focused on the CXCL1 in the subsequent studies. After treatment with VP-16 for 0.5 h, levels of CXCL1 were increased in the supernatants of LoVo, SW480 and H446 cells, but not in those of AGS or SW620 cells (Fig. 2e), which validated the results of mass spectrometry. Furthermore, cellular levels of CXCL1 protein were increased in LoVo, SW480, H446 cells after treatment with TI (Fig. 2f). Correlated with increased CXCL1 protein expression, levels of CXCL1 mRNA were also up-regulated in LoVo, SW480 and H446 cells (Fig. 2g). To the contrary, TI’s effects on CXCL1 protein and mRNA expression were marginal in SW620 and AGS cells (Fig. 2f, g). Thus, TI could promote expression and secretion of chemotactic factor CXCL1.

To corroborate the role of CXCL1, we used siRNA to silence CXCL1 protein expression (Fig. 3a). After knockdown of CXCL1, VP-16 or CPT-11-promoted migration and invasion were significantly alleviated (Fig. 3b). By using a neutralizing antibody against CXCL1, we also found that VP-16 or CPT-11-promoted cell mobility was greatly prevented (Fig. 3c). These results support the notion that CXCL1 mediates TI-promoted cancer cell mobility.

Next, we sought to delineate the mechanism underlying TI-induced CXCL1 expression. Now that TI stimulated CXCL1 transcription (Fig. 2g) and JAK2-STAT1 pathway is involved in transactivation of CXCL1 [36], we examined the roles of JAK2 and STAT1. Treatment with TI for 0.5 h strongly increased phosphorylation levels of JAK2 and STAT1 in LoVo, SW480 and H446 cells, but not in SW620 or AGS cells (Fig. 4a). After interfering the expression of JAK2 (Fig. 4b), TI-induced STAT1 phosphorylation and CXCL1 expression were abrogated (Fig. 4c). Meanwhile, TI-promoted migration was inhibited (Fig. 4d). Similarly, by utilizing the JAK2 inhibitor AG490 [37, 38], we found that TI-promoted STAT1 phosphorylation, CXCL1 induction, and cell migration were all blocked (Fig. 4e, f). Additionally, knockdown of STAT1 abolished TI-promoted CXCL1 expression and migration (Fig. 5a-c). In line with these results, pretreatment with Fludarabine (Flu), a specific inhibitor of STAT1 [37, 38], achieved similar effects (Fig. 5d, e). These results suggest that activation of JAK2-STAT1 pathway mediates TI-promoted CXCL1 expression and cell motility.

Next, we examined the mechanism of TI-promoted JAK2 activation. JAK2 phosphorylation is positively regulated by stimulation with cytokines and growth factors, or negatively regulated by such protein tyrosine phosphatases (PTPs) as PTP1B and TC-PTP [39]. Additionally, JAK2 signaling pathway could be modulated by ROS [40‐42]. Interestingly, JAK2 phosphorylation was revealed to be enhanced by ROS-mediated oxidization and inactivation of PTPs [43]. We found that treatment with VP-16, ADM and CPT-11 enhanced the production of ROS in SW480 cells (Fig. 6a), which was consistent with previous findings [44, 45]. However, ROS status in the unresponsive cell line, SW620, was not affected by TI (Fig. 6a). By using an antibody specific for oxidized motif of PTPs (oxPTP) [46], we found that VP-16 markedly increased protein signals at ~ 50 KD only in responsive cell lines (Fig. 6b). Treatment with ADM or CPT-11 also increased oxPTP levels (Fig. 6b). Prompted by the molecular weight of TI-induced oxPTP, we immumoprecipitated PTP1B and performed Western blot with anti-oxPTP. Results confirmed that oxidization of PTP1B was increased by VP-16 (Fig. 6c). With the ROS scavenger GSH, we found that both phosphorylation of JAK2-STAT1 and expression of CXCL1 induced by TI were negated (Fig. 6d). Consistently, TI-promoted cell migration was prevented (Fig. 6e). Collectively, these results indicate that ROS-mediated PTP1B oxidization contributes to TI-stimulated JAK2-STAT1-CXCL1 pathway and motility.