NF-jB signaling is activated and confers resistance to apoptosis in three-dimensionally cultured EGFR-mutant lung adenocarcinoma cells

Yuji Sakuma ⇑, Yukiko Yamazaki, Yoshiyasu Nakamura, Mitsuyo Yoshihara, Shoichi Matsukuma, Shiro Koizume, Yohei Miyagi

A B s t R A C t

Epidermal growth factor receptor (EGFR)-mutant lung adenocarcinoma cells in suspension undergo apop- tosis to a greater extent than adherent cells in a monolayer when EGFR autophosphorylation is inhibited by EGFR tyrosine kinase inhibitors (TKIs). This suggests that cell adhesion to a culture dish may activate an anti-apoptotic signaling pathway other than the EGFR pathway. Since the microenvironment of cells cultured in a monolayer are substantially different to that of cells existing in three-dimension (3D) in vivo, we assessed whether two EGFR-mutant lung adenocarcinoma cell lines, HCC827 and H1975, were more resistant to EGFR TKI-induced apoptosis when cultured in a 3D extracellular matrix (ECM) as com- pared with in suspension. The ECM-adherent EGFR-mutant cells in 3D were significantly less sensitive to treatment with WZ4002, an EGFR TKI, than the suspended cells. Further, a marked degradation of IjBa, the inhibitor of nuclear factor (NF)-jB, was observed only in the 3D-cultured cells, leading to an increase in the activation of NF-jB. Moreover, the inhibition of NF-jB with pharmacological inhibitors enhanced EGFR TKI-induced apoptosis in 3D-cultured EGFR-mutant cells. These results suggest that inhibition of NF-jB signaling would render ECM-adherent EGFR-mutant lung adenocarcinoma cells in vivo more sus- ceptible to EGFR TKI-induced cell death.

Keywords:
Lung adenocarcinoma
EGFR mutation
Three-dimensional culture NF-jB
Apoptosis

1. Introduction

Gain-of-function mutations in the epidermal growth factor receptor (EGFR) gene are detected in ~30–50% of non-small cell lung cancers (NSCLCs) in East Asia and in ~10% of cases in North America and Western Europe. Most EGFR-mutant NSCLCs display adenocarcinoma histology. These EGFR-mutant lung adenocarcino- mas depend on constitutively active EGFR signaling for survival [1,2]. When compared with cells in a monolayer on culture plastic, cancer cells in suspension show a markedly increased resistance to conventional chemotherapeutic drugs, which predominantly affect cells that are rapidly dividing, possibly due to cell cycle arrest in the suspended cells [3,4]. Our previous in vitro experiments, how- ever, have shown that suspended EGFR-mutant lung adenocarci- noma cells undergo apoptosis to a greater extent than do adherent cells in a monolayer when treated with EGFR tyrosine ki- nase inhibitors (TKIs) [5]. This suggests that EGFR-mutant lung adenocarcinoma cells critically depend on constitutive EGFR acti- vation to thrive in anchorage-independent conditions, and EGFR- mutant cells floating in circulation or pleural effusion in vivo may be vulnerable to EGFR TKIs. Alternatively, some anti-apoptotic sig- naling, which is probably derived from the adhesion of cells to cul- ture dishes, may be activated in adherent tumor cells cultured in a monolayer in addition to signaling via the EGFR pathway.
Since the microenvironment of cells in a monolayer culture con- dition is considerably different to that of cells in vivo existing in a three-dimensional (3D) structure, it remains unclear whether extracellular matrix (ECM)-adherent EGFR-mutated carcinoma cells in vivo are less sensitive to EGFR TKIs than cells floating in cir- culation or pleural effusion. The aim of this study was 1) to clarify whether ECM-adherence in a 3D culture model renders EGFR-mu- tant lung adenocarcinoma cells more resistant to EGFR TKI-in- duced apoptosis than those cells treated in suspension; and, if this is the case, 2) to determine what molecule(s) contribute to this increased resistance to apoptosis in ECM-adherent cells.

2. Materials and methods

The experimental procedures were approved by the Institu- tional Review Board at the Kanagawa Cancer Center, Japan.

2.1. 3D on-top culture

We cultured EGFR-mutant lung adenocarcinoma cells above a thin layer of 100% ECM in medium containing 2% ECM, termed ‘‘3D on-top culture’’, as previously described [6,7]. Briefly, culture dishes of 94-mm in diameter (Greiner Bio-One, Tokyo, Japan) were coated with 2350 lL/dish of Matrigel (BD Biosciences, Franklin Lakes, NJ, USA), containing a laminin-rich mixture of ECM proteins produced by Engelbreth-Holm-Swarm (EHS) mouse sarcoma cells, and incubated at 37 °C for 30 min to allow the Matrigel to solidify. Subsequently, spheroids, which were formed from ~60 × 105 tu- mor cells in a suspension culture system using 90-mm low cell binding dishes (Nalge Nunc International KK, Tokyo, Japan), were resuspended in 10 mL of 3D culture medium (RPMI 1640 medium containing 2% Matrigel, and supplemented with 10% heat-inacti- vated fetal bovine serum, 50 units/ml penicillin and 50 lg/ml streptomycin) and added on top of the Matrigel (Fig. 1A). For 3D on-top culture with 24-well culture plates, Matrigel matrix 24-well plates (BD Biosciences) were used, and spheroids were seeded onto the gel at an initial density of ~2 × 105 cells per well in 500 lL of 3D culture medium.

2.2. Cell culture and drugs

Two EGFR-mutant lung adenocarcinoma cell lines, HCC827 (del E746-A750) and NCI-H1975 (L858R/T790 M), were obtained from the American Type Culture Collection (Manassas, VA, USA) and maintained at 37 °C in a humidified incubator with 5% CO2 in air. HCC827 cells or H1975 cells were seeded on low cell binding dishes and then cultured in 10 mL medium without 2% Matrigel for 24 h. In these conditions, the suspended cells aggregated to form tight, densely packed spheroids (Fig. 1A) [5,8]. The spheroids were then cultured in 3D on-top or suspension culture conditions for another 48 h. Subsequently, the spheroids were untreated or treated with the small-molecule inhibitor(s) for another 24 h. The pharmacological agents used in this study were the third-gen- eration EGFR TKI, WZ4002 (Selleck Chemicals, Houston, TX, USA) [9] and a potent, specific IKKb inhibitor, TPCA-1 (Tocris Bioscience, Ellisville, MO, USA) [10]. As a control, cells were treated with the same concentration of dimethyl sulfoxide (Sigma–Aldrich, St. Louis, MO, USA). Stock solutions of the drugs were prepared at 10 mM in dimethyl sulfoxide.

2.3. Western blotting

3D-cultured cells were harvested by treatment with Dispase (BD Biosciences) to resolve Matrigel, according to the manufac- turer’s instructions. The cells, untreated or treated with the indi- cated agent(s) and cultured in 3D or suspension conditions, were then lysed in NuPAGE LDS Sample Buffer (Invitrogen, Carlsbad, CA, USA). The whole cell lysates were subjected to SDS–PAGE (Nu- PAGE 4–12% Bis-Tris Gel; Invitrogen) followed by blotting with specific antibodies, and detection using the Supersignal West Pico Chemiluminescent Substrate (Thermo Scientific, Rockford, IL, USA). The types and dilutions of primary antibodies used were: anti- PARP-1 (p116/p85) (E78; 1:1000 dilution; Epitomics, Burlingame, CA, USA), anti-EGFR (1:1000 dilution), anti-phospho-EGFR (Y1173) (53A5; 1:2000 dilution), anti-IKKb (2C8; 1:1000 dilution), anti-phospho-IKKa (S176)/IKKb (S177) (C84E11; 1:500 dilution), anti-IjBa (44D4; 1:1000 dilution), anti-NF-jB p65 (1:2000 dilu- tion), or anti-phospho-NF-jB p65 (S536) (93H1; 1:2000 dilution), and anti-b-actin (AC–15; 1:10,000 dilution; Sigma–Aldrich). All antibodies used, except anti-PARP-1 and anti-b-actin, were pur- chased from Cell Signaling Technology Japan (Tokyo, Japan).

2.4. Apoptosis assays

Apoptosis was examined by western blot analysis of the cleaved poly(ADP-ribose) polymerase-1 (PARP-1) and Caspase-Glo 3/7 As- say (Promega, Madison, WI, USA). PARP-1 is an intracellular ‘‘death substrate’’ and its cleaved-form is used as a representative marker of apoptosis [11]. For the PARP-1 cleavage assay, the band intensi- ties of the pro- and cleaved-forms of PARP-1 were measured on X- ray films using Image J software. For Caspase 3/7 activation, tumor spheroids of the two cell lines were seeded on Matrigel matrix 24- well plates for 3D culture at an initial density of ~2.0 × 105 cells per well in 500 lL of 3D culture medium, and allowed to adhere to the gel for 24 h. The cells were then untreated or treated with the indicated drug(s) for another 24 h. After 48 h, cell apoptosis was assessed using the Caspase-Glo 3/7 Assay, according to the manufacturer’s instructions.

2.5. Statistical analysis

Differences in rates of apoptosis between treated cells in 3D or suspension culture conditions, or between 3D-cultured cells trea- ted with WZ4002 alone or in combination with TPCA-1, were eval- uated by paired t-tests. P-values of less than 0.05 were considered significant. All statistical calculations were performed with the JMP software system (JMP for Windows version 7; SAS Institute Japan; Tokyo, Japan).

3. Results

3.1. Three dimensionally cultured EGFR-mutant lung adenocarcinoma cells are more resistant to WZ4002-induced apoptosis than suspended cells

Our 3D culture model seems to recapitulate well carcinoma tis- sue architecture in vivo with evidence of tumor cell spheroids, where cells are interconnected with one another and adhering to adjacent ECM. During the period of observation, spheroids of HCC827 or H1975 cells in 3D culture settings maintained almost the same round shape as suspended spheroids (Fig. 1A, 1B). The microenvironment of 3D-cultured spheroids was thus similar to that of spheroids in suspension, with the exception that 3D-cul- tured spheroids interacted with the ECM. However, the PARP-1 cleavage assay demonstrated that 3D-cultured EGFR-mutant cells were significantly less sensitive to WZ4002-induced apoptosis than suspended cells (Fig. 1C, 1D). Phosphorylation of EGFR ex- pressed in EGFR-mutant cells in the 3D culture was completely suppressed by WZ4002 treatment (Fig. 1C), indicating that the adhesion of spheroids to the ECM did not elevate drug efflux activ- ity, and that the 2% Matrigel within the media did not affect drug diffusion. These observations suggest that some anti-apoptotic sig- naling, probably derived from the interaction between the spher- oids and the ECM, was activated only in 3D-cultured spheroids.

3.2. NF-jB is activated in 3D-cultured EGFR-mutant lung adenocarcinoma cells

We found that the expression of IjBa, the inhibitor of nuclear factor-jB (NF-jB), was markedly reduced in 3D-cultured cells, as compared with suspended cells (Fig. 2A). As a result, the activation of NF-jB was enhanced in 3D-cultured cells, as confirmed by the increased expression of phosphorylated NF-jB component p65 (also termed RelA). The marked reduction in the levels of IjBa in 3D-cultured cells seemed to be due to activation of IjB kinase (IKK) b (Fig. 2A), which phosphorylates IjB. This phosphorylation, in turn, causes IjB to become ubiquitylated and rapidly degraded in proteasomes [12]. In the absence of IjB, NF-jB is free to move into the nucleus, where it activates the expression of anti-apoptotic genes. We surmised that these anti-apoptotic genes activated by the NF-jB transcription factor play an important role in the in- creased resistance of 3D-cultured EGFR-mutant cells to WZ4002- induced apoptosis. The obvious decrease in IjBa expression levels, leading to increased NF-jB activation, was observed as early as 6 h after tumor cell spheroids were cultured in 3D conditions (Fig. 2B).

3.3. TPCA-1(a specific inhibitor of IKKb) treatment enhances the levels of apoptosis induced by EGFR TKI in 3D-cultured EGFR-mutant lung adenocarcinoma cells

The findings mentioned above suggested that NF-jB activation played a role, at least in part, in the increased resistance of 3D-cultured cells to WZ4002-induced apoptosis. NF-jB pathway inhibi- tion can be achieved through the inhibition of IKKb, the primary ki- nase that promotes IjB instability and leads to NF-jB activation. We thus tested whether a potent, selective IKKb inhibitor, TPCA- 1 [10], could enhance the susceptibility of EGFR-mutant lung ade- nocarcinoma cells to WZ4002 in our 3D-culture. The action of TPCA-1 treatment was confirmed by the increased levels of IjBa. Fig. 3A shows that caspase 3/7 activation induced by the combined treatment of WZ4002 plus TPCA-1 was significantly greater than that induced by WZ4002 alone in EGFR-mutant cells in 3D culture. Further, the combination treatment induced apoptosis to a greater degree than WZ4002 alone in terms of PARP-1 cleavage (Fig. 3B). The efficacy of the combined treatment of WZ4002 plus TPCA-1 was also supported by a significant increase in the number of frag- mented cells observed in the treated spheroids (Fig. 3C). Another IKKb inhibitor, BMS-345541 [10], also increased the level of apoptosis in 3D-cultured cells treated with WZ4002 (data not shown). These findings indicate that enhanced NF-jB activation in 3D settings plays a certain role in the decreased vulnerability of 3D-cul- tured EGFR-mutant spheroids to WZ4002, as compared with suspended spheroids.

4. Discussion

Although the 3D on-top culture model exploited in this study does not fully recapitulate all aspects of the in vivo microenviron- ment, it appears to offer a more natural culture setting than a monolayer culture [6,7]. In particular, our 3D culture model seems to be similar, at least partly, to carcinoma tissue architecture in vivo where developed tumor cell nests, in which cells intercon- nect with each other, adhere to the ECM through hemidesmo- somes. This is because tumor cell spheroids, where cells are interconnected with one another, adhere to adjacent ECM in our 3D model. Here, we have demonstrated that 3D-cultured spheroids of two EGFR-mutant lung adenocarcinoma cell lines (HCC827 and H1975) were significantly less sensitive to EGFR TKI-induced apoptosis than suspended spheroids. Interestingly, IjBa was markedly degraded only in the 3D culture, probably mediated by IKKb, lead- ing to aberrant NF-jB activation. IKKb phosphorylates IjB, and the phosphorylated IjB then becomes ubiquitylated and rapidly de- graded in proteasomes [12]. It seems that such activation of the NF-jB pathway was mainly caused by interactions between tumor spheroids and the ECM (100% Matrigel), because only spheroids in the 3D culture, and not those in suspension, adhered to and interacted with the ECM. More importantly, we confirmed that the pharmacological inhibition of the enzymatic activity of IKKb en- hanced the efficacy of EGFR TKI in 3D-cultured EGFR-mutant cells. The findings on the NF-jB activation induced in 3D culture ob- tained in the present study are essentially in accord with those established by the seminal study of Weaver and colleagues in breast epithelial cells [13]: Weaver et al. previously clarified that 3D cell organization and polarity were responsible for the activa- tion of NF-jB and regulation of cellular sensitivity to drug-induced apoptosis. However, our study is the first report to demonstrate that NF-jB is activated and confers resistance to EGFR TKI-induced apoptosis in 3D-cultured, polarized spheroids of EGFR-mutant lung adenocarcinoma cells.
Although only a minority of cells constituting spheroids at- tached themselves to the ECM in our 3D on-top culture system, we observed a marked degradation of IjBa in these 3D-cultured spheroids. This suggests that IjBa was degraded in most, if not all, cells constituting spheroids. Adhesion of a spheroid to the ECM probably induces spheroid polarization, and we surmise that the ECM may also induce polarization of the cells constituting the spheroid, with IjB being degraded in the polarized cells (Fig. 4). It remains to be elucidated precisely whether and to what degree IjB is degraded; in other words, the degree to which NF-jB is activated in individual cells constituting a spheroid.
NF-jB activation has been shown to be essential for the devel- opment of K-ras driven lung adenocarcinomas [14]. More recently, the transcription factor has been reported to play a certain role in non-uniform response to EGFR TKI treatment in EGFR-mutant lung cancers [15]. The results presented here clearly reveal that NF-jB is activated and significantly correlated with increased resistance to EGFR TKI-induced apoptosis in EGFR-mutant lung adenocarci- noma cells cultured in 3D conditions that better reflect the in vivo microenvironment. Moreover, clinical data exist showing a positive association between low IjB expression (=high NF-jB activation state) and worse progression-free survival and de- creased overall survival in EGFR-mutant lung cancer patients trea- ted with EGFR TKI [15]. These data collectively TPCA-1 suggest that pharmacological inhibition of NF-jB using an IKKb inhibitor may improve the clinical outcome for EGFR-mutant lung adenocarci- noma patients treated with EGFR TKIs.

References

[1] S.V. Sharma, D.W. Bell, J. Settleman, et al., Epidermal growth factor receptor mutations in lung cancer, Nat. Rev. Cancer 7 (2007) 169–181.
[2] W. Pao, J. Chmielecki, Rational, biologically based treatment of EGFR-mutant non-small-cell lung cancer, Nat. Rev. Cancer 10 (2010) 760–774.
[3] R.E. Durand, P.L. Olive, Resistance of tumor cells to chemo- and radiotherapy modulated by the three-dimensional architecture of solid tumors and spheroids, Methods Cell Biol. 64 (2001) 211–233.
[4] Y. Hua, K. Gorshkov, Y. Yang, et al., Slow down to stay alive: HER4 protects against cellular stress and confers chemoresistance in neuroblastoma, Cancer (in press) doi: 10.1002/cncr.27496.
[5] Y. Sakuma, Y. Yamazaki, Y. Nakamura, et al., WZ4002, a third-generation EGFR inhibitor, can overcome anoikis resistance in EGFR-mutant lung adenocarcinomas more efficiently than Src inhibitors, Lab. Invest. 92 (2012) 371–383.
[6] G.Y. Lee, P.A. Kenny, E.H. Lee, et al., Three-dimensional culture models of normal and malignant breast epithelial cells, Nat. Methods 4 (2007) 359–365.
[7] T. Shibue, R.A. Weinberg, Integrin beta1-focal adhesion kinase signaling directs the proliferation of metastatic cancer cells disseminated in the lungs, Proc. Natl. Acad. Sci. U S A 106 (2009) 10290–10295.
[8] Y. Sakuma, T. Takeuchi, Y. Nakamura, et al., Lung adenocarcinoma cells floating in lymphatic vessels resist anoikis by expressing phosphorylated Src, J. Pathol. 220 (2010) 574–585.
[9] W. Zhou, D. Ercan, L. Chen, et al., Novel mutant-selective EGFR kinase inhibitors against EGFR T790M, Nature 462 (2009) 1070–1074.
[10] D. De Silva, M.D. Mitchell, J.A. Keelan, Inhibition of choriodecidual cytokine production and inflammatory gene expression by selective I-kappaB kinase (IKK) inhibitors, Br. J. Pharmacol. 160 (2010) 1808–1822.
[11] V.A. Soldatenkov, M. Smulson, Poly(ADP-ribose) polymerase in DNA damage- response pathway: implications for radiation oncology, Int. J. Cancer 90 (2000) 59–67.
[12] C. Nakanishi, M. Toi, Nuclear factor-kappaB inhibitors as sensitizers to anticancer drugs, Nat. Rev. Cancer 5 (2005) 297–309.
[13] V.M. Weaver, S. Lelièvre, J.N. Lakins, et al., Beta4 integrin-dependent formation of polarized three-dimensional architecture confers resistance to apoptosis in normal and malignant mammary epithelium, Cancer Cell 2 (2002) 205–216.
[14] S.E. Meylan, A.L. Dooley, D.M. Feldser, et al., Requirement for NF-kappaB signalling in a mouse model of lung adenocarcinoma, Nature 462 (2009) 104– 107.
[15] T.G. Bivona, H. Hieronymus, J. Parker, et al., FAS and NF-jB signalling modulate dependence of lung cancers on mutant EGFR, Nature 471 (2011) 523–526.