Research ArticleCancer

JAK2 inhibition sensitizes resistant EGFR-mutant lung adenocarcinoma to tyrosine kinase inhibitors

See allHide authors and affiliations

Science Signaling  29 Mar 2016:
Vol. 9, Issue 421, pp. ra33
DOI: 10.1126/scisignal.aac8460
  • Fig. 1 Synergistic antiproliferative effects of combined EGFR blockade with JAK inhibition.

    (A) MTT-based proliferation assay in H1975, PC-9R, and H1650 cells treated with the TKI erlotinib (Ti) in combination with the JAKi AZD1480 (Ji). Data are means ± SEM from five replicates in three independent experiments. *P < 0.05; **P < 0.001, versus AZD1480 alone (two-tailed Student’s t test). Below each graph are the representative combination indices (CIs) of erlotinib (TKI) in combination with AZD1480 (JAKi): CI < 0.9 indicates synergy, CI between 0.9 and 1.1 is additive, and CI > 1.1 indicates antagonism. The drug dosage combinations used in the MTT assay are boxed. (B) Western blotting in lysates from H1975, PC-9R, and H1650 cells treated with JAKi (AZD1480, 1 μM), TKI (erlotinib, 0.2 μM), or the combination for 1 hour. Blots are representative of three experiments and are quantified in fig. S1D. (C) Tumor volume tracking in mice bearing xenografts of H1975, PC-9R, or H1650 cells and treated with vehicle (C), AZD1480, erlotinib, or the combination (Ji + Ti) for 12 to 25 days. Doses are provided in Materials and Methods. Data are means ± SEM (n = 5 to 7 mice per group). *P < 0.05, AZD1480 versus the combination (two-tailed Student’s t test).

  • Fig. 2 JAK inhibition or depletion enhances EGFR-ERK signaling.

    (A) Western blotting as indicated in lysates from H1975, PC-9R, and H1650 cells treated with AZD1480 (1 μM) for 1 hour. Blots are representative of three experiments and are quantified in fig. S2E. (B) Staining for pSTAT3, pEGFR, EGFR, and pERK in representative tumor sections from H1650 xenografts treated with vehicle control (C) or AZD1480 (Ji) (30 mg/kg, twice daily for 3 weeks). Scale bars, 100 μm. (C) Western blotting as indicated in lysates from H1975, PC-9R, and H1650 cells transfected with control (SCRAMBLE) or JAK2 siRNA. Blots are representative of three experiments and are quantified in fig. S3E.

  • Fig. 3 JAK2 inhibition increases surface EGFR expression through SOCS5.

    (A) Detection of cell surface–bound EGFR on PC-9R cells using Alexa Fluor–EGF at 4°C after a 1-hour treatment with AZD1480 or control. Scale bars, 50 μm. (B) Serum-starved H1975, H1650, and PC-9R cells were pretreated with AZD1480 or control for 1 hour. Surface proteins were biotinylated, precipitated with avidin resin beads, and analyzed by Western blot for EGFR and c-MET. Blots are representative of three experiments and are quantified in fig. S4A. (C) Detection of JAK2-EGFR and SOCS5-EGFR interactions by Duolink staining in PC-9R cells treated with AZD1480 or control for 1 hour. Scale bars, 50 μm. (D) Cell lysates from H1650 and PC-9R cells treated with control or AZD1480 were immunoprecipitated with an antibody against EGFR or JAK2 and analyzed by Western blot for EGFR, JAK2, and ubiquitin. Loading controls were the heavy-chain (H-chain) immunoglobulin G (IgG) for the co-IP, and tubulin for the input. Blots are representative of three experiments and are quantified in fig. S4D. (E) PC-9R cells expressing JAK2 shRNA (JAK2sh) or vector control (Csh) constructs analyzed for SOCS5 and EGFR interactions by Duolink staining. Scale bars, 50 μm. Cell lysates were analyzed for JAK2 and tubulin. (F) Western blot for SOCS5 and tubulin in lysates from H1975 cells expressing scrambled control or SOCS5 shRNA (SOCS5sh). Control or SOCS5sh cells were treated with control, AZD1480 (1 μM), or erlotinib (0.2 μM) for 1 hour and analyzed for pEGFR, EGFR, and tubulin by Western blot. Representative blots are shown (n = 3). (G) Tumor volumes in mice bearing H1975-SOCS5Sh xenografts and treated with vehicle or TKI (25 mg/kg per day) for 9 days. Data are means ± SEM (n = 5 to 7 mice per group). **P < 0.01, control versus TKI (two-tailed Student’s t test). (H) Schematic depicting NSCLC cells expressing EGFR proteins, wherein JAK2 bridges SOCS5-dependent EGFR degradation, and inhibition or reduction of JAK2 uncouples SOCS5 from EGFR, effectively increasing EGFR abundance on the cell surface.

  • Fig. 4 Enhancement of EGFR-ERK signaling by JAK inhibition is mediated through heterodimerization between wild-type/mutant EGFR, which is abrogated by TKI.

    (A) Serum-starved H1975 cells were pretreated with AZD1480 or control (C) for 1 hour, and EGF ligand was added for the indicated times. Surface proteins were biotinylated, precipitated with avidin resin beads, and analyzed by Western blot for surface EGFR (sEGFR) and c-MET (sMET). (B) Duolink staining (left) for wild-type (WT) EGFR (Myc) and mutant L858R EGFR (MUT) interaction in H1975 cells expressing Myc-tagged WT EGFR protein and treated with AZD1480 or control for 1 hour. Scale bars, 50 μm. Western blot (right) in lysates from H1975 parental cells (control) and cells expressing Myc-tagged WT EGFR were analyzed for EGFR, Myc-tagged protein, and tubulin. (C) Top, experimental schematic in which H1975 cells expressing both the WT (blue sphere) and EGFR-L858R/T790M gatekeeper mutant (red sphere) proteins are depicted as a single cell expressing variable amounts of each. H1975 cells were treated with either TKI (0.2 μM) or the T790M-specific inhibitor WZ4002 (WZ; 25 nM) for 30 days, and selected populations are depicted by their relative expression of WT and mutant EGFR per cell. Extracts from WZ4002- and TKI-selected cells treated with either control or AZD1480 were then analyzed by Western blot as indicated (note that EGFR detects both WT and mutant). The growth inhibitory effects of AZD1480 (JAKi) in combination with erlotinib (TKI) are shown below as representative CIs. (D) Western blotting as indicated in lysates from serum-starved PC-9R cells treated with AZD1480 for 1 hour and then EGF for 30 min in the presence of erlotinib at the indicated concentrations. Blots in (A), (C), and (D) are representative of three experiments each and are quantified in figs. S5, A and C, and S6, respectively.

  • Fig. 5 Working model: JAK inhibition enhances TKI-sensitive EGFR signaling and growth inhibition in NSCLC.

    Schematic depicting TKI-resistant NSCLC cells expressing both WT and gatekeeper mutant EGFR proteins, which form homodimers or heterodimers. At steady state, JAK2 promotes SOCS5-dependent EGFR degradation. TKI-resistant, homodimeric mutant EGFR is the principle driver of the RAS–MEK (mitogen-activated protein kinase kinase)–ERK signaling cascade (in bold). Inhibition or reduction of JAK2 uncouples SOCS5 from EGFR, effectively increasing TKI-sensitive, wild-type EGFR homodimers/heterodimers and signaling (in bold). Although the role of EGFR WT:MUT heterodimer signaling has not been clearly defined, we hypothesize that it may regulate TKI sensitivity. This working model may explain the synergistic actions observed between the EGFR-targeted TKI and JAKi on NSCLC tumor growth.

Supplementary Materials

  • www.sciencesignaling.org/cgi/content/full/9/421/ra33/DC1

    Fig. S1. STAT3 activation and inhibition in TKI-resistant NSCLC.

    Fig. S2. JAK2 inhibition increases the abundance of pEGFR and pERK in NSCLC.

    Fig. S3. JAK2 inhibition leads to rapid activation of EGFR/RAS/ERK signaling.

    Fig. S4. JAK2 links SOCS5 to EGFR, regulating sensitivity to TKI.

    Fig. S5. Quantification of Western blots in Fig. 4, A and C.

    Fig. S6. Quantification of Western blots in Fig. 4D.

  • Supplementary Materials for:

    JAK2 inhibition sensitizes resistant EGFR-mutant lung adenocarcinoma to tyrosine kinase inhibitors

    Sizhi P. Gao, Qing Chang, Ninghui Mao, Laura A. Daly, Robert Vogel, Tyler Chan, Shu Hui Liu, Eirini Bournazou, Erez Schori, Haiying Zhang, Monica Red Brewer, William Pao, Luc Morris, Marc Ladanyi, Maria Arcila, Katia Manova-Todorova, Elisa de Stanchina, Larry Norton, Ross L. Levine, Gregoire Altan-Bonnet, David Solit, Michael Zinda, Dennis Huszar, David Lyden,* Jacqueline F. Bromberg*

    *Corresponding author. E-mail: dcl2001{at}med.cornell.edu (D.L.); bromberj{at}mskcc.org (J.F.B.)

    This PDF file includes:

    • Fig. S1. STAT3 activation and inhibition in TKI-resistant NSCLC.
    • Fig. S2. JAK2 inhibition increases the abundance of pEGFR and pERK in NSCLC.
    • Fig. S3. JAK2 inhibition leads to rapid activation of EGFR/RAS/ERK signaling.
    • Fig. S4. JAK2 links SOCS5 to EGFR, regulating sensitivity to TKI.
    • Fig. S5. Quantification of Western blots in Fig. 4, A and C.
    • Fig. S6. Quantification of Western blots in Fig. 4D.

    [Download PDF]

    Technical Details

    Format: Adobe Acrobat PDF

    Size: 1.72 MB


    Citation: S. P. Gao, Q. Chang, N. Mao, L. A. Daly, R. Vogel, T. Chan, S. H. Liu, E. Bournazou, E. Schori, H. Zhang, M. Red Brewer, W. Pao, L. Morris, M. Ladanyi, M. Arcila, K. Manova-Todorova, E. de Stanchina, L. Norton, R. L. Levine, G. Altan-Bonnet, D. Solit, M. Zinda, D. Huszar, D. Lyden, J. F. Bromberg, JAK2 inhibition sensitizes resistant EGFR-mutant lung adenocarcinoma to tyrosine kinase inhibitors. Sci. Signal. 9, ra33 (2016).

    © 2016 American Association for the Advancement of Science

Navigate This Article