Transglutaminase-2 mediates acquisition of neratinib resistance in metastatic breast cancer

Abstract

Acquisition of resistance to targeted therapies remains a major clinical obstacle for the HER2⁺ subtype of breast cancer. Using an isogeneic progression series of HER2⁺ breast cancer metastasis we demonstrate that metastatic cells have an increased capacity to acquire resistance to the covalent, pan-ErbB inhibitor, neratinib. RNA sequencing analyses comparing parental and metastatic cells identified upregulation of transglutaminase 2 (TG2). Genetic depletion and overexpression approaches established that TG2 is both necessary and sufficient for acquisition of neratinib resistance. Mechanistically, we describe a pathway in which TG2-mediates activation of NF-κB signaling leading to upregulation of IL-6 in metastatic cells. This autocrine expression of IL-6 functions to maintain enhanced levels of TG2 via JAK:STAT3 signaling. This drug persistence feedback loop can be interrupted through the use of the JAK1/2 inhibitor ruxolitinib. In vivo application of ruxolitinib had no effect on tumor growth under non-treated conditions, but effectively prevented acquisition of resistance, leading to tumor regression upon coadministration with neratinib. Overall, our studies reveal a mechanism in metastatic breast cancer cells that predisposes them to acquisition of resistance to ErbB-targeted therapeutics. Clinically, immediate application of ruxolitinib could prevent acquisition of resistance and improve patient responses to HER2-targeted therapies.

Keywords: Breast cancer; Drug resistance; HER2; IL-6; Metastasis.

Fig. 1

Transglutaminase 2 facilitates neratinib resistance. a HME2-partenal cells were treated with neratinib (1 µM) for the indicated amounts of time and phosphorylated (p) and total (t) HER2 were evaluated by immunoblot. β-tubulin was used as a loading control. b HME2-parental cells were treated (96 h) with neratinib at the indicated concentrations and cell viability was assayed. Data are the mean ± SE of three independent experiments resulting in the indicated IC50 value. c Immunoblot analyses for TG2 in HME2 parental (Par) and bone metastases (BM). HME2-GFP, HME2-TG2, HME2-BM-scram and HME2-BM-shTG2 were also assessed for TG2 expression. β-tubulin was used as a loading control. Data are representative of at least three independent experiments. d TG2 manipulated cells were seeded under single-cell 3D culture conditions in the presence of DMSO or neratinib (100 nM). Images of each condition at day 27 are shown. e TG2 manipulated cells were treated in 2D culture every three days with neratinib (100 nM) for a period of 8 weeks. Crystal violet was used to stain representative wells with at the indicated time points to visualize drug resistant cells. f The neratinib resistant (NR) cells were cultured for an additional 4 weeks without neratinib. The drug resistant cells and passage matched HME2-BM-Scram and HME2-Par-TG2 cells were treated with the indicated concentrations of neratinib for 96 h and cell viability was assayed. Data are the mean of three independent experiments ± SE, resulting in the indicated IC50 and P values

Fig. 2

Transglutaminase 2 facilitates neratinib resistance in vivo. a,c HME2 parental cells expressing GFP or TG2, and HME2-BM cells expressing a control (scram) or TG2-targeted shRNA (shTG2) were injected into the mammary fat pad. Tumor bearing mice were treated with neratinib as described in the material and methods tumors were removed and weighed upon necropsy at the indicated time points. Data are the mean of 5 mice per group ± SE, resulting in the indicated p-values. b,d Upon necropsy, tumors were fixed and imaged

Fig. 3

Transglutaminase 2 causes increased expression of IL-6. a Phosphorylation array comparing lysates derived from HME2-parental and HME2-BM cells. Proteins whose phosphorylation was potentially different between the two cell lines are indicated. b Spot intensity for the indicated sites of phosphorylation. Intensity values were normalized to the HME2-parental cells. c HME2-parental and HME2-BM cells were treated with Lapatinib (500 nM) for 2 h and analyzed for phosphorylation of STAT3 and AKT. Analysis of total levels of STAT3 and AKT served as loading controls. Immunoblots are representative of at least three independent experiments. d Transcript (upper graph) and protein (lower graph) levels for IL-6 in HME-parental and HME2-BM manipulated for TG2 expression were quantified using qRT-PCR and AlphaLISA. Transcript levels are expressed relative to GFP expressing HME2-parental cells. All data are the mean ±SE of three independent experiments resulting in the indicated p-values. e HME2-parental and HME2-BM cells manipulated for TG2 expression were stimulated with IL-6 (20 ng/ml) for 30 min or treated with ruxolitinib (1 µM) for 6 h and analyzed for phosphorylation of STAT3. Analysis of total levels of STAT3 served as loading controls. Immunoblots are representative of at least three independent experiments

Fig. 4

Transglutaminase 2-mediated activation of NF-κB is required for acquisition of neratinib resistance. a,b HME2-BM cells and TG2 overexpressing HME2-Parental (Par) cells were further constructed to express the NF-κB super repressor (S.R.) or GFP as a control and IL-6 was quantified using qRT-PCR (a) and AlphaLISA (b). c The constructed cell lines described in panel a in were serum starved overnight and analyzed for Iκbα (IkBa) expression and phosphorylation of STAT3. Analysis of total levels of STAT3 and β-tubulin served as loading controls. Immunoblots are representative of at least three independent experiments. d The constructed cell lines described in panel a were treated with neratinib every three days. Crystal violet was used to visualize viable cells in representative wells

Fig. 5

STAT3-driven expression of transglutaminase-2 faciliates drug resistance. a Transcript levels for TG2 in HME2-BM cells expressing the NF-κB super repressor (S.R.) or GFP as a control were quantified using qRT-PCR. Data are relative to empty (MT) vector control (HME2-BM-MT) cells. b HME2-BM cells expressing a empty shRNA vectors (shMT) one those targeting STAT3 (shSTAT3) were analyzed for TG2 and STAT3. Analysis of β-tubulin served as a loading control. Immunoblots are representative of at least three independent experiments. c HME2-BM cells described in panel b were treated with neratinib, ruxolitinib, or both drugs every three days for a period of 5 weeks. Crystal violet staining was used at the indicated time points to visualize resistant cells. d HME2-BM cells were treated with no drug (ND) or cultured in the presence of IL-6 or ruxolitinib (Rux) for 14 days and expression of TGM2 was evaluated by RT-PCR. e Immunoblot results for TG2 from cells treated as described in (d), β-tubulin served as a loading control. f HME2-BM cells constructed as described in panel (a) were injected into the mammary fat pad and caliper measurements were used to quantify tumor growth. Arrow indicates initiation of neratinib or ruxolitinib or combination treatment. As a control mice were treated with an equal concentration of drug solvent (DMSO). Data are the mean ± SE of 5 mice per group resulting in the indicated p-values where *p < 0.05, **p < 0.01

Fig. 6

Expression of TG2 and IL6 predict disease recurrence in HER2 + breast cancer. a Left, Correlation analysis of the TCGA dataset comparing IL6 and TGM2. Data were analyzed using a Pearson correlation test, yielding the indicated r and p values. Right, comparison of TGM2 expression in patient samples bearing high (above the mean) or low (below the mean) phosphorylation of STAT3. b Left, Correlation analysis of the METABRIC dataset comparing IL6 and TGM2. Data were analyzed as in panel (a). Right, comparison of Nottingham Prognosis Indexes for patients with high (above the mean) or low (below the mean) expression of TGM2 and IL6. Data were analyzed by a student’s T-test where **** is p < 0.0001. c Kaplan Meier analyses for relapse free survival (RFS) of patients with high (above the mean) or low (below the mean) expression of fibronectin, TGM2, and IL6. Data were analyzed by a logrank test. d A schematic representation of a TG2-mediated drug persistence loop where breast cancer cells upregulate TG2 driving NF-κB-mediated expression of IL-6. Autocrine and paracrine IL-6-mediated activation of JAK:STAT3 signaling contributes to maintained expression of TG2 completing the drug persistence signaling loop. Application of the JAK2 inhibitor ruxolitinib can disrupt this persistence