Click here to load reader
View
213
Download
0
Embed Size (px)
R E S E A R C H A R T I C L E
C A N C E R
JAK2 inhibition sensitizes resistant EGFR-mutantlung adenocarcinoma to tyrosine kinase inhibitorsSizhi P. Gao,1 Qing Chang,1 Ninghui Mao,1 Laura A. Daly,1 Robert Vogel,2
Tyler Chan,1 Shu Hui Liu,1 Eirini Bournazou,1 Erez Schori,1 Haiying Zhang,3
Monica Red Brewer,4,5 William Pao,4,5 Luc Morris,6 Marc Ladanyi,7,8 Maria Arcila,7
Katia Manova-Todorova,9 Elisa de Stanchina,10 Larry Norton,1,11 Ross L. Levine,1,8,11
Gregoire Altan-Bonnet,2 David Solit,1,8,11,12 Michael Zinda,13 Dennis Huszar,13*David Lyden,3,14,15 Jacqueline F. Bromberg1,11
hD
ownloaded from
Lung adenocarcinomas with mutant epidermal growth factor receptor (EGFR) respond to EGFR-targetedtyrosine kinase inhibitors (TKIs), but resistance invariably occurs. We found that the Janus kinase (JAK)/signal transduction and activator of transcription 3 (STAT3) signaling pathway was aberrantly increased inTKI-resistant EGFR-mutant nonsmall cell lung cancer (NSCLC) cells. JAK2 inhibition restored sensitivityto the EGFR inhibitor erlotinib in TKI-resistant cell lines and xenograft models of EGFR-mutant TKI-resistantlung cancer. JAK2 inhibition uncoupled EGFR from its negative regulator, suppressor of cytokine signaling5 (SOCS5), consequently increasing EGFR abundance and restoring the tumor cells dependence on EGFRsignaling. Furthermore, JAK2 inhibition led to heterodimerization of mutant and wild-type EGFR subunits,the activity of which was then blocked by TKIs. Our results reveal a mechanism whereby JAK2 inhibitionovercomes acquired resistance to EGFR inhibitors and support the use of combination therapy with JAKand EGFR inhibitors for the treatment of EGFR-dependent NSCLC.
ttp://s
on June 14, 2018
tke.sciencemag.org/
INTRODUCTION
Lung cancer is the most frequent cause of cancer death (1), and nonsmallcell lung cancer (NSCLC) is the most common subtype. Somatic activat-ing mutations of the tyrosine kinase domain of the epidermal growthfactor receptor (EGFR) are found in about 26% of all patients with lungadenocarcinoma and confer sensitivity to first-generation EGFR tyrosinekinase inhibitors (TKIs), gefitinib and erlotinib (2, 3). Clinical responsesare variable, although most patients exhibit good response rates to theseinhibitors. However, acquired resistance to TKIs unfailingly occurs, inmost cases (>60%) due to the acquisition of gatekeeper mutations(T790M) in the EGFR, which is thought to alter kinase ATP (adenosine5-triphosphate) affinity above that of gefitinib or erlotinib (4, 5). Progression-free survival with TKI treatment is only 9 to 12 months, and overall sur-vival is less than 20 months (2, 3). Notably, the acquisition of secondarymutations in EGFR emphasizes a continued dependence on EGFR
1Department of Medicine, Memorial Sloan Kettering Cancer Center (MSKCC),New York, NY 10065, USA. 2Computational Biology Program, MSKCC, NewYork, NY 10065, USA. 3Childrens Cancer and Blood Foundation Laboratories,Departments of Pediatrics, Cell and Developmental Biology, Weill CornellMedical College (WCMC), New York, NY 10021, USA. 4Division of Hematology/Oncology, Vanderbilt-Ingram Cancer Center (VICC), Nashville, TN 37232,USA. 5Personalized Cancer Medicine, VICC, Nashville, TN 37232,USA. 6Department of Surgery, MSKCC, New York, NY 10065, USA. 7Depart-ment of Pathology, MSKCC, New York, NY 10065, USA. 8Human Oncologyand Pathogenesis Program, MSKCC, New York, NY 10065, USA. 9MolecularCytology, MSKCC, New York, NY 10065, USA. 10Antitumor Assessment, MSKCC,New York, NY 10065, USA. 11WCMC, New York, NY 10021, USA. 12MetastasisResearch Center, MSKCC, New York, NY 10065, USA. 13Oncology iMED,AstraZeneca, Waltham, MA 02451, USA. 14Department of Pediatrics, MSKCC,New York, NY 10065, USA. 15Drukier Institute for Childrens Health, MeyerCancer Center, WCMC, New York, NY 10021, USA.*Present address: Takeda Pharmaceutical Company Ltd., Cambridge, MA02139, USA.Corresponding author. E-mail: [email protected] (D.L.); [email protected] (J.F.B.)
ww
signaling in these cancers. The need to overcome both innate and acquiredresistance has been a major therapeutic challenge.
EGFR is a member of the ERBB/human epidermal growth factor re-ceptor (HER) family of membrane-bound receptor tyrosine kinases (RTKs)(6). Aberrant regulation of EGFR, including gain-of-functionmutations andoverexpression, is a common feature of many epithelial malignancies,which has led to the development of EGFR TKIs (7). We previously de-scribed that signal transduction and activator of transcription 3 (STAT3)is persistently tyrosine-phosphorylated or activated (pSTAT3) in NSCLC(cell lines and primary tumors) due to EGFR-driven up-regulation ofinterleukin-6 (IL-6) expression, leading to a feed-forward IL-6/Janus kinase(JAK)/STAT3 loop. Furthermore, JAK inhibition abrogates proliferationin NSCLC cell lines, including those that are TKI-resistant (8). JAK1/2 in-hibitors have shown promise in preclinical models of NSCLC (915). In-hibitors of JAKs were developed for immunologic suppression for organtransplantation and for the treatment of myeloproliferative neoplasms inpatients with activating mutations in the JAK2 pathway (16, 17) and arein early-phase clinical trials for lymphomas and solid tumors on the basisof promising preclinical studies (1115, 1820).
Our present study investigated the mechanisms by which JAK inhibi-tion represses cell growth in NSCLC cells, alone or in combination withTKIs. Here, we found that JAK2 inhibition overcame acquired resistanceto TKIs in EGFR-mutant lung adenocarcinoma in vitro and in vivo.
RESULTS
JAK2 inhibition resensitizes TKI-resistant cells andxenograft models to erlotinibWe previously demonstrated by immunohistochemistry that pSTAT3 ispresent in 42% of NSCLC cells that have wild-type EGFR and in 88%of NSCLC cells that have mutant EGFR, mediated through increased
w.SCIENCESIGNALING.org 29 March 2016 Vol 9 Issue 421 ra33 1
http://stke.sciencemag.org/
R E S E A R C H A R T I C L E
on June 14, 2018http://stke.sciencem
ag.org/D
ownloaded from
IL-6/JAK signaling (8). Further examination of this cohort of samplesrevealed that 31% of EGFR-mutant NSCLC tumors had high expres-sion (immunohistochemistry) of pSTAT3. Here, we sought to determinethe relevance of JAK/STAT3 activation in tumors that had developed re-sistance to TKIs. Patients with EGFR-mutant NSCLC had their tumorsrebiopsied upon development of acquired resistance to erlotinib or ge-fitinib (hereafter referred to collectively as TKI) (5). We examined the
ww
abundance of pSTAT3 in 10 TKI-resistant tumors, 4 of which werematched against the untreated primary tumor. We determined that theabundance of pSTAT3 was high (score 2 to 3+) in 68% (4 of 6) of un-matched samples and either similar or increased in all four matched spe-cimens compared to the respective pre-TKI samples (fig. S1A) (21, 22).These results led us to hypothesize that pSTAT3 may be a relevant targetin TKI-resistant disease.
A
B
C
H1975 PC-9R
H1975 PC-9R H1650
60
50
40
30
20
10
0C Ji Ti Ji+Ti
% G
row
th in
hibi
tion
70
60
50
40
30
20
10
0
JAKi (nM)
TKI (nM)
CI
62.5
125
0.2
125
125
0.4
250
125
0.55
**
C Ji Ti Ji+Ti
H1650
908070605040302010
0C Ji Ti Ji+Ti
JAKiTKI
+ +
+ +
JAKiTKI
+ +
+ +
JAKiTKI
+ +
+ +
**
*
3
*
JAKi (nM)
TKI (nM)
CI
187.5
500
0.67
375
250
0.45
750
500
0.79
JAKi (nM)
TKI (nM)
CI
312.5
125
0.23
625
125
0.24
1250
125
0.18
pEGFR
pSTAT3
pERK
Tubulin
EGFR
STAT3
pEGFR
pSTAT3
pERK
Tubulin
EGFR
STAT3
pEGFR
pSTAT3
pERK
Tubulin
EGFR
STAT3
ERK ERK ERK
250-
95 -
55-
250-
95 -
55-
250-
95 -
55-
% G
row
th in
hibi
tion
% G
row
th in
hibi
tion
10 13 16 19 22 25 28 31Days
Tum
or v
olum
e (m
m3 )
H19751400
800
600
400
200
0
Ti
Ji+Ti
C
Ji
Days
Tum
or v
olum
e (m
m)
3T u
mor
vol
ume
(mm
)
400
300
200
100
011 15 19 23 27 31 34
C
Ji+Ti*
Ti
Ji
PC-9R700
600
500
400
300
200
100
020 25 30 35 40 45
Days
Ji
Ji+Ti *
Ti
C
H1650
Fig. 1. Synergistic antiproliferative effects of combined EGFR blockadewith JAK inhibition. (A) MTT-based proliferation assay in H1975, PC-9R,
in the MTT assay are boxed. (B) Western blotting in lysates from H1975,PC-9R, and H1650 cells treated with JAKi (AZD1480, 1 mM), TKI (erlotinib,
and H1650 cells treated with the TKI erlotinib (Ti) in combination with theJAKi AZD1480 (Ji). Data are means SEM from five replicates in threeindependent experiments. *P < 0.05; **P < 0.001, versus AZD1480 alone(two-tailed Students t test). Below each graph are the representativecombination 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 ant