The role of HGF-MET pathway and CCDC66 cirRNA expression in EGFR resistance and epithelial-to-mesenchymal transition of lung adenocarcinoma cells

The role of HGF-MET pathway and CCDC66 cirRNA expression in EGFR resistance and... Background: Epithelial-to-mesenchymal transition (EMT) has, in recent years, emerged as an important tumor cell behavior associated with high metastatic potential and drug resistance. Interestingly, protein SUMOylation and hepatocyte growth factor could respectively reduce the effect of small molecule inhibitors on tyrosine kinase activity of mutated epidermal growth factor receptor of lung adenocarcinomas (LADC). The actual mechanism is yet to be resolved. Methods: Immunohistochemistry was used to stain proteins in LADC specimens. Protein expression was confirmed by Western blotting. In vitro, expression of proteins was determined by Western blotting and immunocytochemistry. Levels of circular RNA were determined by reverse transcription-polymerase chain reaction. Results: SAE2 and cirRNA CCDC66 were highly expressed in LADC. Expression of SAE2 was mainly regulated by EGFR; however, expression of cirRNA CCDC66 was positively regulated by FAK and c-Met but negatively modulated by nAchR7α. EGFR-resistant H1975 also highly expressed cirRNA CCDC66. Immediate response of hypoxia increased phosphorylated c-Met, SAE2, and epithelial-to-mesenchymal transition. Either activation of FAK or silencing of nAchR7α increased cirRNA CCDC66. Conclusions: HGF/c-Met regulates expression of SAE2 and cirRNA CCDC66toincreaseEMT anddrugresistanceofLADC cells. Multimodality drugs concurrently aiming at these targets would probably provide more benefits for cancer patients. Keywords: SUMOylation, EGFR, SAE2, EMT, HGF, c-MET, cirRNA, CCDC66 Background improved living environments, the annual incidence and Elevated tumor cell proliferation and metastatic potential death of lung cancer increase incessantly, including the are major characteristics of lung adenocarcinoma (LADC), USA and Taiwan [3](https://www.mohw.gov.tw/lp-3327- associating with increased resistance to radio- and chemo- 2.html). The appalling escalation was particularly notable in therapy, as well as early tumor recurrence, fast disease female patients. Unfortunately, many new patients were progression, and high mortality rate [1, 2]. Although pub- mostly young non-smokers who died of lung cancer due to lic policy of restricting cigarette retails and smoking has the early recurrence and multi-organ involvement [1–3] (http://www.mohw.gov.tw/cht/DOS/Statistic.aspx). Recent development of chemotherapy targeting mutated * Correspondence: shchiou@dragon.nchu.edu.tw; hssun@mail.ncku.edu.tw; kcchow668@gmail.com human epidermal growth factor receptor (EGFR) by tyro- Graduate Institute of Microbiology and Public Health, National Chung Hsing sine kinase inhibitors (TKIs) has become a major strategic University, Taichung, Taiwan 5 treatment for patients in the Far East [4]. However, not all Institute of Molecular Medicine, National Chung Kung University, Tainan, Taiwan Graduate Institute of Biomedical Sciences, National Chung Hsing University, patients with the same clinical features had the same treat- Taichung, Taiwan ment achievements [5]. Moreover, TKI improved only Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 2 of 14 progression-free survival but not overall survival, suggesting In this report, we showed their correlations via coiled-coil that TKIs might not be effective for full-blown and large- domain containing 66 (CCDC66) circular RNA (cirRNA) scale metastasis and that sensitivity of tumor cells might be to induce EGFR resistance as well as to increase EMT and affected by other factors. To determine whether the disease metastatic potential of LADC cells. progression correlates with specific gene expression pro- files, we combined differential displays and microarrays to Results analyze gene expression patterns in LADC specimens at Expressions of intracellular markers, ATAD3A and SAE2, variousstagesofdisease progression[6]. We identified sev- influence the sensitivity of EGFR-mutated LADC cells to eral potential genes, including aldo-keto reductases (also tyrosine kinase inhibitors named dihydrodiol dehydrogenase, DDH), hepatocyte LADC cells with mutant EGFR (H1975 cells) overexpressed growth factor (HGF), HGF receptor (or c-Met), dynamin- ATAD3A (Fig. 1a, b), an anti-apoptotic factor [9], but rarely related protein 1 (DRP1) and the ATPase family, and expressed SAE2. Molecular weight of EGFR in H1975 AAA domain-containing 3A (ATAD3A) [6–9]. Interest- (~ 185 kDa) was higher than in A549 (170 kDa), suggest- ingly, the protein level of ATAD3A increased dramatically ing that the protein was phosphorylated. Interestingly, when cancer cells were serum-starved. Furthermore, some of H1975 cells (dark arrows) resembled phenotypes through nicotinic acetylcholine receptor alpha 7 of mesenchymal cells [21](Fig. 1b). Silencing of SAE2 in- (nAChRα7), nicotine could increase expression of HGF creased ATAD3A expression (Fig. 1c) and sensitivity of and eukaryotic elongation factor 2 (eEF2), which in turn A549 cells to gefitinib (Fig. 1d) and erlotinib (Fig. 1e). Sen- KD upregulate ATAD3A and DRP1 [6, 10]. Nicotine also in- sitivity of SAE2 A549 cells to gefitinib was equivalent to KD creases eEF2 SUMOylation, which is vital for maintaining that of H1975 cells (ID ~25 μM). However, SAE2 anti-apoptotic capability of LADC. A549 cells (ID ~50 μM) were more resistant than SUMOylation is a post-translational modification of H1975 cells (ID ~5.6 μM) to erlotinib. Silencing of KD proteins by a SUMO (small ubiquitin-related modifier) ATAD3A (ATAD3A ), on the other hand, reduced ex- protein (conjugating to an epsilon lysine residue of the pression of EGFR and AIF (Fig. 1f), however, increased substrate protein) and induces protein conformation sensitivity to both (Fig. 1g)gefitinib and(Fig. 1h) erlotinib. change to enhance transcription and cell cycle progression [11]. Before reaction, the C-terminus of SUMO precursor Expression pattern of membrane receptors and is removed by SUMO-specific proteases (SUPs) to expose intracellular transport markers in LADC cell lines the glycine-glycine motif, which is activated by a heterodi- In vitro, expression of EGFR was low in H23 and H2009, meric enzyme complex, composed of SUMO-activating moderate in H838 and H2087 (5- to 10-fold higher than in enzyme subunit 1 (SAE1) and SAE2, with the expense of H23 and H2009 cells), and high in A549, H226, and H1437 an ATP [12]. The activated SUMO is transferred from cells (10- to 30-fold higher than in H23 and H2009 cells) SAE1/2 to an E2 conjugating protein, ubiquitin-conjugating (Fig. 2a, the upper panel). H1437 cells that had a molecular (UBC) 9, and subsequently conjugated to the final sub- weight upshift was also positive for c-Met and HGF. Ex- strates by an E3 ligase, e.g., SP-RING-type (Siz/PIAS-family) pression of SAE2 was low in H838 and H1437, moderate in E3 ligase or IR-type (internal repeats) E3 ligase [11]. Four H125 (5- to 10-fold higher than in H838 and H1437 cells), types of SUMO proteins have been identified; however, they and high in A549, H23, H226, and H2087 cells (20- to 40- are all activated by the same SAE1/2 and transferred to the fold higher than in H838 and H1437 cells) (Fig. 2a,the substrate proteins via UBC9 [11–14]. lower panel). Interestingly, molecular weights of EGFR in Using a genome-wide RNA interference method to H838, H2087 (~ 175 kDa), and H1437 (~ 185 kDa) were screen cancer cell growth-related genes, Kessler et al. found higher than those in A549, H23, H226, and H2009 that expression of SAE1/2 was correlated with higher meta- (~ 170 kDa) cells. Moreover, sensitivity of H1437 to gefi- static potential of cancer cells [15]. SUMOylation of BrCa1 tinib was equivalent to that of H1975 [22]. has been shown to increase DNA damage repair and drug Although addition of EGF increased levels of phosphory- resistance [16]. Moreover, SUMOylation of SIRT1, a class lated EGFR (pEGFR), no obvious molecular weight upshift III deacetylase sirtuin-1, was found to drive epithelial-to- of EGFR was observed (Fig. 2b). Addition of hepatocyte mesenchymal transition (EMT) [17]. Interestingly, growth factor (HGF), on the other hand, increased molecu- SUMOylation of GTPase Rac1, which was activated by lar weight upshift of EGFR (from 170 to 185 kDa) (Fig. 2c). HGF and PI3K/Akt, not only increased cell migration Treatment with 0.1 U/μgofcalfintestinalphosphatase [18], but also reduced effectiveness of gefitinib on EGFR (CIP) downshifted EGFR molecular weight in H1437 cells mutants [19]. A recent study by Du et al. showed that (Fig. 2d). Silencing of SAE2 did not evidently affect protein SAE2 increased Oct-1 stability to maintain stemness of level of EGFR but reduced HER2 levels (Fig. 2e). Silencing cancer cells [20]. However, the relationships among SAE2, of EGFR reduced protein level of EGFR and that of SAE2 ATAD3A, EGFR, HGF, and EMT are yet to be determined. (Fig. 2f). Short-term silencing of EGFR increased levels of Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 3 of 14 Fig. 1 Expression of intracellular markers, SAE2 and ATAD3A, affect sensitivity of EGFR-mutated lung adenocarcinoma (LADC) cells to tyrosine kinase inhibitors, gefitinib or erlotinib. a LADC cells, carrying mutated EGFR (H1975), expressed abundant ATAD3A, an anti-apoptotic factor [9], as determined by a Western blotting analysis. b ATAD3A protein was highly expressed in EGFR-mutated H1975 (the left panel) and, intermediately, in the wild-type A549 (the right panel) cells. Some of the EGFR-mutated H1975 cells were elongated, scattered, and crescent shape (dark arrows), resembling phenotypes of mesenchymal cells [21], implicating that mutated EGFR could be associated with epithelial-to-mesenchymal transition KD (EMT) of LADC cells. c Knockdown of SAE2 expression (SAE2 ) increased ATAD3A expression. d In SAE2 cells, gefitinib sensitivity was also increased, KD suggesting that SAE2 could be an intracellular sensor of gefitinib. It is therefore worth noting that gefitinib sensitivity of SAE2 A549 cells was KD equivalent to that of H1975 cells (ID50 ~ 25 μM). e Silencing of SAE2 expression increased erlotinib sensitivity of A549 cells as well. However, SAE2 A549 cells (ID ~ 50 μM) were more resistant than H1975 cells (ID ~ 5.6 μM) to erlotinib. f Silencing of ATAD3A (ATAD3A ) reduced expression of EGFR KD KD and AIF. g Increase of gefitinib sensitivity was thus anticipated in ATAD3A cells. h Increment of erlotinib sensitivity was also noted in ATAD3A cells c-Met and pro-c-Met. Prolonged silencing of EGFR for 831, 836, 837, 840, and 844 above 170 kDa; tumor over 96 h, on the other hand, diminished both c-Met and specimens 839, 842, and 843 about 170 kDa) (Fig. 3a). ATAD3A (Fig. 2g), suggesting a physiological feedback cir- Protein level of mutant EGFR was in general lower cuit for regulating homeostasis of intracellular material than that of the wild type. Samples having EGFR transport system, which could then sustain cell survival molecular weight upshifting were mostly positive for under conditions of growth factor deficiency or with c-Met, supporting our previous results that HGF/c- defected receptors [6–10]. Met system was critical for the function of EGFR. While SAE2 expression correlated with EGFR, expres- Biological correlations of EGFR with drug resistance, sion of ATAD3A and AIF was markedly reduced in epithelial-to-mesenchymal transition, and metastatic LADC samples with mutant EGFR. potential of LADC cells Previously, we had shown that c-Met was highly In patient’s biopsies, molecular weights of EGFR var- expressed in LADC biopsies and cell lines [6]. Patients ied from patient to patient (tumor specimens 830, with smoking habits usually had higher HGF level in Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 4 of 14 Fig. 2 Expression pattern of membrane receptors and intracellular transport markers in lung cancer cell lines. a Expression of EGFR detected by immunoblotting was high in A549, H226, and H1437; moderate in H838 and H2087; and low in H23 and H2009 cells. EGFR was not detected in H125 cells. Molecular weight of EGFR in H838, H2087 (about 175 kDa), and H1437 (about 185 kDa) was higher than that in A549, H23, H226, and H2009 (about 170 kDa) cells. Interestingly, only H1437 cells expressed both HGF and c-Met. Expression of SAE2, on the other hand, was detected as a 100-kDa protein in lung cancer cells. Expression of β-actin was a monitoring standard. SAE2 expression was high in A549, H23, H226, and H2087; moderate in H125; and low in H838 and H1437 cells. SAE2 was not detected in H2009 cells. b Addition of EGF increased expression level and the phosphorylated forms (pEGFR) of EGFR in H2009 cells, which did not express c-Met. However, no evident molecular weight upshifting of EGFR was detected. Addition of EGF increased SAE2 protein expression as well. c Addition of HGF increased molecular weight upshifting of EGFR from 170 to 185 kDa in A549 cells. d Following treatment of H1437 cells with calf intestinal phosphatase (CIP), protein levels and molecular weights of 185 kDa EGFR reduced, suggesting that phosphorylation KD was critical for EGFR stability. e Using siRNA to knockdown EGFR (EGFR ) expression for 48 h reduced protein levels of SAE2. f Silencing of EGFR gene KD KD expression for 48 h (EGFR -48hr) reduced SAE2 expression. g However, expression of c-Met and pc-Met was increased in EGFR -48hr cells. Such short-term effect dwindled when EGFR gene silencing was continued to 96 h; and at this time, protein level of c-Met and ATAD3A reduced markedly the diseased lungs. In vitro, hypoxia increased HGF (as measured by Western blotting analysis), hypoxia level and c-Met phosphorylation. Using CoCl to deplete altered intracellular protein distribution (Fig. 3d,as intracellular oxygen also increased phosphorylated c-Met shown by confocal immunofluorescence microscopy). and levels of EGFR and SAE2, as well as markers of Vimentins were gathered in cytoplasmic vacuole-like epithelial-to-mesenchymal transition (EMT), e.g., vimen- structures. In this way, vimentins on plasma membranes tin and paxillin, in LADC cells. The increase was dose- were markedly reduced. Hypoxia affected intracellular dis- and time-dependent (Fig. 3b, c). Silencing of SAE2 or tribution of paxillin and nuclear ataxia-telangiectasia- c-MET alleviated the effect of hypoxia on EMT mutated (ATM) kinase as well. Expression of ATM was markers’ expression (Fig. 3c). Besides protein levels clearly reduced, and most of the proteins were retained Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 5 of 14 Fig. 3 Biological correlations of EGFR with drug resistance, epithelial-to-mesenchymal transition, and metastatic potential of LADC cells. a In vitro, we had shown that the different molecular weights of EGFR in various NSCLC cell lines were closely associated with HGF and c-Met pathways. Molecular weights of EGFR in surgical biopsies of LADC also varied from patient to patient (tumor samples 830, 831, 836, 840, and 844 above 170 kDa; tumor specimens 839, 842, and 843 around 170 kDa). Interestingly, protein levels of mutant EGFR were generally lower than those of wild type. Samples that had higher molecular weights were mostly positive for c-Met and higher serum HGF, supporting our in vitro results that EGFR could be modified through HGF/c-Met pathway. b Using CoCl to induce hypoxia-activated c-Met phosphorylation (p-c-Met) increased protein level of EGFR and SAE2, as well as epithelial-to-mesenchymal transition (EMT) markers, e.g., vimentin and paxillin, in a dose-dependent manner in H23 cells. c c-Met phosphorylation and increase of the related proteins were initiated about 30–60 min following CoCl treatment of H23 cells. Appearance of HIF-1α became visible about 3–6 h following CoCl challenge, indicating that the increase of EMT-related protein expression occurred before traditional hypoxic activation. The hypoxia-mediated protein increases were abolished when SAE2 gene was silenced. Moreover, addition of actinomycin D did not affect CoCl -induced cell responses. d In addition to the increases of proteins as shown by Western blotting analysis, confocal immunofluorescence micrographs revealed that hypoxia altered protein distribution inside the cells as well. For instance, vimentin was accumulated in the enlarged cytoplasmic vacuole-like structures, but not on the plasma membrane. Intracellular distribution of paxillin was generally similar to that of vimentin, except that some cells started to have nuclear signals of paxillin. Nuclear signals of ATM, however, were markedly reduced, and most of the proteins were retained in the cytoplasm in elongated spherical forms. e Exposure of H23 cells (□, control) to hypoxic condition increased resistance to cisplatin (○,hypoxia for12 h; △, hypoxia for 24 h). Using shRNA-lentivirus to silence SAE2 gene expression (by shSAE2-1 [▲]orbyshSAE2-2[●]) increased sensitivity of LADC cells to cisplatin. f Silencing of SAE2 expression in A549 cells reduced cell mobility across extracellular matrix protein-coated carbonate membrane (the left panel). When H23 cells were exposed to hypoxia prior to the cell mobility examination, the cell migration ability increased significantly (the right panel). The statistical differences were significant (WT A549 vs. shSAE2-1 and shSAE2-2, P < 0.0004 [***], ANOVA; control H23 vs. H23 cells exposed to hypoxia, P < 0.0358 [*], ANOVA). [*P < 0.05; ***P < 0.001] mostly in the cytoplasm as irregularly orbicular vesicles, Pathologically, EGFR and SAE2 were concurrently suggesting that hypoxia could interfere intracellular expressed in LADC patients with poor prognosis material transport, which fortuitously activated cell sur- Expression of EGFR and SAE2 was determined by West- vival response [6–10, 23]. ern blotting in 66 pairs of NTLT and LADC specimens. Exposure of LADC cells to CoCl for 12–24 h increased Both EGFR (84.8%, 56/66) and SAE2 (91%, 60/66) were SAE2 expression and cisplatin resistance (Fig. 3e). Silencing detected in most of the tumor fractions (Fig. 4a, b). Ex- of SAE2 expression reduced not only cisplatin resistance pression of SAE2 correlated with EGFR (Pearson r = (Fig. 3e), but also cell mobility across Matrigel-coated 0.454, 95% confidential interval was between 0.238 and membrane (Fig. 3f). Hypoxia increased cell mobility. The 0.627, and R = 0.206). Pathologically, SAE2 was identi- difference was statistically significant (WT vs. shSAE2-1- or fied in the tumor nests (302/372, 81.2%) (Fig. 4c1–c4). shSAE2-2-treated A549 cells, P < 0.0004, ANOVA; control In NTLT, SAE2 signal was detected in type II H23 vs. hypoxic H23 cells, P < 0.0358, ANOVA). pneumocytes (Fig. 4c5). SAE2 was also detected in Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 6 of 14 Fig. 4 Pathologically, EGFR and SAE2 were concurrently expressed in LADC patients with poor survival. a Using a Western blotting method, expression of EGFR was detected with various molecular weights in 18 tumor and eight non-tumor fractions of 20 pair of LADC biopsies. β-actin was used as a monitoring standard. b Using the same method, expression of SAE2 was detected as a 100-kDa protein in 19 tumor and two non-tumor fractions of LADC specimens. N non-tumor lung tissue (NTLT), T tumor fraction of surgical resections. c Representative examples of SAE2 expression in pathological specimens of LADC was determined by immunohistochemical staining (crimson precipitates). Expression of SAE2 was detected in the LADC tumor nests of (c1) papillary, (c2) acinar, (c3) solid, and (c4) mixed subtypes. (c5) In the NTLT, signal of SAE2 was detected in type II pneumocytes (the enlarged area), but not in type I pneumocytes of the air sac. d Comparison of Kaplan-Meier product limit estimates of survival analysis in LADC patients. (d1) Patients were divided into two groups depending on the expression of SAE2. The difference in survival between the two groups was compared by the log-rank test. Survival of patients with low SAE2 level was significantly better than those with high SAE2 level. The hazard ratio between these two groups was 1.708, and the difference in cumulative survival was significant (P = 0.002). (d2) Survival of stage I patients with low SAE2 level was significantly better as well. The hazard ratio was 4.08, and the difference in cumulative survival was significant (P = 0.0053) 87.77% (244/278) of metastatic lymph nodes (data not SAE2, 132 (43.7%) had tumor recurrence. Among the 70 shown). Statistical analysis showed that SAE2 expression patients who had low SAE2, 15 had tumor recurrence correlated with cigarette smoking, tumor stage, and lym- (22.8%). All recurrences were detected within 24 months phovascular involvement (Table 1), suggesting that SAE2 after operation. The recurrence risk of high SAE2 was expression was closely associated with metastatic poten- 2.85-fold higher than that of low SAE2 (P < 0.001). Sur- tial. Interestingly, among the 302 patients who had high vival rate of low SAE2 patients was better than that of Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 7 of 14 Table 1 Correlation of SAE2 expression with clinicopathological (Fig. 5g), suggesting that nAchRα7 could be a feedback parameters in patients with LADC mechanism of HGF-c-MET circuit. Our results showed Expression P value that expression of pEGFR, c-MET, and FAK were higher of SAE2 in EGFR-mutated H1975 than that in A549 cells (Fig. 5h). Parameter High Low Univariate Multivariate Expression of cirRNA CCDC66 was also higher in H1975 (n = 302) (n = 70) cells (Fig. 5i). Levels of nAchRα7, in contrast, were higher Gender in A549 cells. Addition of FAK inhibitor Y15 (1,2,4,5-ben- Male (n = 292) 243 49 0.0549 0.107 zenetetramine tetrahydrochloride, Merck KGaA, Darm- Female (n = 80) 59 21 stadt, Germany) inhibited expression of FAK, nAchR7α, and CCDC66α (Fig. 5j), however, increased cirRNA Cigarette smoking CCDC66 expression (Fig. 5k, l). Smoker (n = 266) 227 39 0.0012 0.0025 Non-smoker (n = 106) 75 31 Relationship between expression of cirRNA CCDC66 and Stage EMT markers in LADC cells I(n = 83) 65 18 0.042 0.091 Expression of cirRNA CCDC66, FAK, and paxillin was II (n = 109) 81 28 determined by RT-PCR in 11 pairs of NTLT and LADC specimens. All three makers were detected in most of the IIIa (n = 180) 156 24 tumor fractions [cirRNA CCDC66 (82%, 9/11), FAK (82%, Cell differentiation 9/11), and paxillin (91%, 10/11)] (Fig. 6a). Pathologically, Well (n = 56) 42 14 0.205 0.41 paxillin was identified in tumor nests and the neighboring Moderate (n = 211) 172 39 area (Fig. 6b1–b3), supporting that paxillin expression Poor (n = 105) 88 17 could correlate with metastatic potential. Knockdown of KD Lymphovascular invasion cirRNA CCDC66 (cirRNA CCDC66 ) expression (Fig. 6c) did not affect resistance to gefitinib or erlotinib in H1975 Positive (n = 278) 244 34 < 0.001 0.0021 cells (Fig. 6d1, d2) but increased cisplatin resistance in H23 Negative (n = 94) 58 36 KD cells (Fig. 6d3). In cirRNA CCDC66 cells, expression of † 2 Two-sided P value determined by χ test ‡ 2 EMT markers (Fig. 6e) and invasion ability (Fig. 6f, g1)were Two-sided P value determined by χ test for trend significantly abrogated. In pCIRC2-cirRNA CCDC66- TF high SAE2 patients (Fig. 4d1). The hazard ratio was transfected A549 cells (cirR ), although invasion ability TF 1.708, and the difference in cumulative survival was increased slightly (compared to cirR H2009 cells), the significant (P = 0.002). Multivariate analysis showed that increment was significant (Fig. 6g2). the difference in SAE2 expression was significant between the two groups (P = 0.0047). Survival rate of Discussion stage I patients with low SAE2 level was also better EGFRs, especially those with tyrosine kinase mutations, (Fig. 4d2, P = 0.0053). The hazard ratio was 4.08. have recently become major targets of TKIs, e.g., gefitinib, erlotinib, and afatinib, in LADC chemotherapy [4, 5, 25]. Expression of CCDC66 cirRNA in LADC cells However, patient’s responses are equivocal [5]. The fact As shown in Fig. 3b, c, the increase of protein synthesis that TKI therapy improves progression-free but not over- before elevation of HIF-1α following exposure to hypoxia all survival [25–27] has raised sincere queries into drug’s suggested that translation efficacy might be increased by effectiveness, which may be affected by other factors. reduction of functional microRNA (miRNA), which Kasahara et al. showed that LADC patients with higher repressed mRNA translation (a list of potential miRNA is HGF level were more resistant to TKIs [26]. Yano et al. shown in Additional file 1). Recently, Hsiao et al. showed demonstrated that gefitinib resistance was mediated via that cirRNA CCDC66 could promote cancer cell growth HGF/c-Met circuit to activate PI3K/AKT [25]. Moreover, and metastasis by adsorbing miRNA [24]. Interestingly, continuous exposure of tumor cells to TKIs transpires cirRNA CCDC66 was also detected in LADC cells extra EGFR mutations and drug resistance [27], suggesting (Fig. 5a), and its expression correlated with the levels of nuclear events of DNA sequence alteration and gene acti- FAK mRNA, a vital marker closely associated with cancer vation. Nonetheless, induction mechanisms of gene muta- metastasis and EMT (Fig. 5b). Silencing of FAK expression tions and phenotype changes, e.g., increase of EMT and (Fig. 5c) reduced cirRNA CCDC66 (Fig. 5d). Addition of metastatic potential [28, 29], are yet to be resolved. HGF, on the contrary, increased expression of cirRNA Kessler et al. had found that higher metastatic potential CCDC66 (Fig. 5e). It is worth noting that silencing of of cancer cells was correlated with expression of SAE1/2 nicotinic acetylcholine receptor alpha 7 (nAchRα7) in- [15]. It was further demonstrated that the elevated meta- creased levels of CCDC66β (Fig. 5f) and cirRNA CCDC66 static potential was activated by SUMOylation of Rac1, Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 8 of 14 Fig. 5 Detection of cirRNA CCDC66 expression in LADC cells. a Expression of cirRNA CCDC66 was detected by RT-PCR in LADC cells. b Its expression correlated with levels of FAK mRNA, a vital marker closely associated with cancer metastasis and EMT. c Silencing of FAK expression reduced d cirRNA CCDC66. e Addition of HGF (50 ng/ml) increased expression of cirRNA CCDC66. f Silencing of nicotinic acetylcholine receptor alpha 7 (nAchRα7), in contrast, increased levels of CCDC66β and g cirRNA CCDC66. h Comparing gene expression patterns of pEGFR, c-MET, and FAK in EGFR-mutated H1975 and wild-type A549 cells. i Expression of pEGFR, c-MET, and FAK was higher in EGFR-mutated H1975 cells. Level of nAchRα7, in contrast, was higher in A549 cells. i Expression of cirRNA CCDC66 was also higher in H1975 cells. j Addition of FAK inhibitor Y15 (1,2,4,5-benzenetetramine tetrahydrochloride) inhibited expression of FAK, nAchR7α, and CCDC66α. Y15 inhibited expression of nAchR7α at a concentration of 7.5 μM, CCDC66α at a concentration of 10 μM, and FAK at a concentration of 15 μM. k Y15, however, increased cirRNA CCDC66 expression. l Addition of nicotine also increased cirRNA CCDC66 expression KD cAMP response element binding (CREB) protein, silencing of SAE2 (SAE2 ) inhibited hypoxic effects, our eEF2, and SIRT1 to increase EMT and cell migration data suggest that SAE2 is able to maintain protein stability [10, 16–18]. By showing that both EGFR and SAE2 are through protein SUMOylation. Moreover, because highly expressed in LADC, and that SAE2 expression cor- addition of actinomycin D did not inhibit hypoxic effects, relates with patient’s poor outcomes, our data support our data also suggest that hypoxia may increase protein their observations that SUMOylation is crucial for tumor synthesis from preexisted mRNA. The rapid rise (0.5 to progression. In particular, via HGF, hypoxia could enhance 1 h, such response was therefore named as immediate SUMOylation of these factors [10, 17, 18]. Among these, reaction) of protein levels before evident accumulation of SUMOylated Rac1 could further induce expression of a HIF-1α supported our findings especially that the panel of genes, including SNAI (snail family zinc finger) 1, increased translational efficacy might be initiated before VEGF, and N-cadherin, to facilitate EMT and metastasis upregulation of transcription (compared to the immediate of cancer cells [30, 31]. reaction, such response was named as delayed reaction). Hypoxia increased expression of EGFR, SAE2, and EMT In fact, the different increase rate of respective pro- markers, such as paxillin and vimentin as well [10]. Since teins suggested that the differential augmentation of Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 9 of 14 Fig. 6 (See legend on next page.) Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 10 of 14 (See figure on previous page.) Fig. 6 Expression of cirRNA CCDC66 correlated with that of FAK and paxillin, and knockdown of cirRNA CCDC66 reduced expression of EMT markers and invasion ability of LADC cells. a Expression of cirRNA CCDC66, FAK, and paxillin was determined by RT-PCR in 11 pairs of LADC specimens. All three tumor makers were detected in most of the tumor fractions [cirRNA CCDC66 (82%, 9/11), FAK (82%, 9/11), and paxillin (91%, 10/11)]. Paxillin was detected in tumor nests (b1–b3; b3, #836) and the neighboring areas (b2, #838) by an immunohistochemical staining. Paxillin-positive cells were in dark brown KD color (indicated by an arrow in b1, #859), showing EMT-like fibroblasts characteristics. c Knockdown of cirRNA CCDC66 (cirRNA CCDC66 )expressionin KD H1975 and H23 cells. Resistance to gefitinib (d1) or erlotinib (d2) was not markedly affected in cirRNA CCDC66 H1975 cells (d1, d2), but clearly increased KD cisplatin resistance in cirRNA CCDC66 H23 cells (d3). e Knockdown of cirRNA CCDC66 expression, however, clearly diminished expression of EMT markers (left panel, H1975 cells were stained for vimentin, red fluorescence; right panel, H23 cells were stained for paxillin, green fluorescence. Nuclei were stained with DAPI, blue fluorescence). f Silencing of cirRNA CCDC66 expression markedly reduced cell mobility across extracellular matrix protein-coated carbonate membrane. H23 cells were labeled with green fluorescence protein prior to Matrigel transmembrane assay. Cells in bottom wells of the Boyden chamber were directly detected by green fluorescence. Nuclei were stained with DAPI (blue fluorescence). g Effect of cirRNA CCDC66 on cell invasion ability. KD g1 Invasion ability of cirRNA CCDC66 , H1975, and H23 cells decreased markedly. Statistical differences were significant (H1975 vs. Mock vs. cirRNA KD KD CCDC66 H1975, P < 0.0001 [***], ANOVA; WT H23 vs. Mock vs. cirRNA CCDC66 H23 cells, P = 0.0048 [**], ANOVA). [*P <0.05; **P < 0.01; ***P < 0.001]; g2 When cirRNA CCDC66 was overexpressed in A549 cells, the invasion ability increased a little (wild-type vs. pCIRC2-mCherry-transfected [pCIRC2m] TF TF vs. pCIRC2-cirRNA CCDC66-transfected A549 cells [cirR ], P = 0.037 [*], ANOVA; wild-type vs. cirR A549 cells, P =0.1054, t test). Comparing wild-type, TF TF pCIRC2m, and cirR H2009 cells, increment of invasion ability was significant (P = 0.0025 [**], ANOVA; H2009 vs. cirR H2009, P = 0.0137 [*], t test). H23 cells exposed to 24-h hypoxia were used as internal comparison parameter (P = 0.0031, t test) translation efficacy could be a result of reducing levels of EGFR, indicating that post-translational modification of suppressive microRNA (miRNA) following hypoxia (a EGFR was taken place on intracellular vesicles, and these list of potential miRNA is shown in Additional file 1). events could occur without the presence of EGF when cells However, we did not find an expression surge of specific were under hypoxic conditions [32]. Our results showed miRNA under hypoxia. Recently, circular RNA (cirRNA) that HGF/c-Met circuit was accountable for such modi- CCDC66 was shown to promote colon cancer cell fication, and this circuit fortuitously activated EGFR. growth and metastasis possibly by adsorbing miRNA, Because HGF-induced phosphorylation did not occur which had inhibited expression of several nuclear gene uniquely on tyrosine residues, phosphorylated serine regulators, including DNA methyltransferase 3β and threonine residues were less sensitive to TKIs. (DNMT3B), enhancer of zeste homolog 2 (EZH2), c-Myc, Moreover, joined effects of c-Met and EGFR on inhibit- and Yes-associated protein 1 (YAP1) [24]. Showing that ing cancer cell metastatic ability could be carried out cirRNA CCDC66 was expressed in LADC cells, and its ex- by repressing miR-200 family and methylcytosine diox- pression correlated with FAK, a marker that was associated ygenases (also called ten-eleven translocases, TET) [33]. with cancer metastasis, our results supported their findings. Two proteins, TET1 (235 kDa) and TET2 (224 kDa), have KD Moreover, silencing of FAK (FAK ) reduced cirRNA been identified in TET family. Both TET1 and TET2 pro- KD CCDC66. Knockdown of nAChRα7(nAChRα7 ), on the teins possess transmembrane and coiled coil domains (as other hand, increased cirRNA CCDC66 expression, even in shown in Additional file 5), which could respectively inter- thepresenceofY15,which inhibitednAChRα7expression act with DRP1 and ATAD3A, the essential components of at low concentrations. The results suggested that the intracellular transport system, to regulate nuclear entry of nAChRα7 effect on cirRNA CCDC66 expression was prior these two proteins and other DNA repair-related enzymes to that of FAK. [9, 23]. Although the nuclear events of these proteins have In addition to cirRNA, miRNA levels are also regulated yet to be elucidated, it is plausible that TKI inhibits EGFR by the miRNA maturation process. Argonaute 2 (AGO2), a activity as well as reduces ATAD3A and SAE2 expression member of RNA-induced silencing complex (RISC), was to limit nuclear entry of DNA repair-related proteins, e.g., inhibited by activated EGFR to decelerate miRNA matur- human RAD23 homolog A (hHR23A) and ataxia- ation [32]. Moreover, like AGO2, SAE2 and cirRNA telangiectasia-mutated (ATM) kinase, to increase DNA CCDC66 were also located on the endoplasmic reticulum damage and mutation frequency [23, 34]. In an ongoing (ER) (intracellular localization of SAE2 and cirRNA study, we are focusing on these issues. CCDC66 is shown in Additional files 2, 3,and 4), suggest- ing that the miRNA maturation and miRNA adsorption Conclusions process could be carried out at the same time on the ER. From this study, we identify novel downstream substrates of Hypoxia-increased buildup of AGO2 was also located on EGFR and c-Met, SAE2, and cirRNA CCDC66, which are the membrane of intracellular vesicles, which removed also expressed in LADC. Through SUMOylation, SAE2 miRNA-associated gene suppression. Hypoxia also aug- could maintain stability of other proteins, including mented levels of SAE2 and EMT markers [8, 9, 19]. Inhib- ATAD3A and EMT markers, such as vimentin and paxillin, ition of mitochondrial oxidative phosphorylation increased which are crucial for metastatic potential and drug resist- molecular weight of AGO2 too, which co-precipitated with ance, and these bioparameters correlate with LADC patient’s Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 11 of 14 prognosis. Therefore, multimodality drugs concurrently The primer sequences were selected by Primer3 aiming at several targets in the circuit of putative feedback (http://bioinfo.ut.ee/primer3/). mechanism of EGFR and drug resistance-related enzymes For FAK, the primers are FAKs: 5′-TGAATTTCTTC would probably provide more benefits for cancer patients. TATCAACAG-3′ [sense primer, nts 485-504, NM_00 1352746] and FAKa: 5′-CCTGGCTTCATCTATTCCAT Methods -3′ [antisense primer, nts 1142-1161]. The anticipated Tissue specimens and NSCLC lines DNA fragment is 677 base-pair (bp). For SAE2, the Some of the patients in this study were from the cohorts primers are SAE2s: 5′-TGAAAGTGGAACAGCTGG used in the previous studies [8, 10]. Briefly, from January GT-3′ [nts 481-500, NM_005499] and SAE2a: 5′-GCA 2007 to December 2012, tissue specimens were collected GGAATAATGTTCCCTGCC-3′ [nts 1207-1227]. The from 628 patients with newly diagnosed non-small cell anticipated cDNA fragment is 847 bp. For nAchRα7, the lung cancer cell (NSCLC). Samples from all patients, for primers are nAchRα7s: 5′-GCCGCAGGACGCTCTACT whom at least one follow-up or death was documented, AT-3′ [nts 912-931, NM_148911] and nAchRα7a: 5′-A were pathologically confirmed NSCLC. Of the 628 CGGCACTCATCTCCACACT-3′ [nts 1334-1353]. The patients, 372 were diagnosed as having LADC. The stage anticipated DNA fragment is 442 base-pair (bp). For cir- of the disease was classified according to the new inter- RNA CCDC66 (circular form_10-8, exons 10 to 8), the national staging system for lung cancer [35]. The primers are CCDC66s: 5′-TCTCTTGGACCCAGCTCAG- Medical Ethics Committee had approved the protocol, 3′ [nts 1428-1446, NM_001141947] and CCDC66a: 5′- and written informed consent was obtained from every TGAATCAAAGTGCATTGCCC-3′ [nts 1175-1194]. The patient before surgery (DMR99-IRB-203). All patients anticipated DNA fragment is 272 bp. For CCDC66 mRNA had undergone surgical resection and radical N2 lymph (linear form_16-17, exons 16 to 17), the primers are node dissection. Tumor size, numbers of lymph node in- CCDC66s: 5′-ACCTTGATCCCGATGCACC-3′ [nts 2657- volvement, differentiation, vascular invasion, and mitotic 2675, NM_001141947] and CCDC66a: 5′-ACGGCACTCA number were evaluated. Patients with lymph node TCTCCACACT-3′ [nts 2801-2820]. The anticipated DNA involvement or loco-regional recurrence received irradi- fragment is 164 bp. For β-actin, the primers are β-actins: 5′- ation at the afflicted areas. Those with distant metastasis AGA GCT ACG AGC TGC CTG AC-3′ [nts 797-816, were treated with chemotherapy. After treatment, patients NM_001101] and β-actins: 5′-CAC CTT CAC CGT TCC were routinely followed every 3 to 6 months in the AGT TT-3′ [nts 1337-1356]. The anticipated DNA frag- outpatient department. Results from blood examination, ment is 560 bp. biochemical studies, chest radiography, abdominal sonog- raphy, whole body bone scan, and computerized tomog- Immunoblotting analysis and immunocytochemistry raphy scans of chest that indicated any evidence of disease Immunoblotting and immunohistochemistry were per- were interpreted as tumor recurrence or metastasis. The formed as described previously [6]. Antibodies for β-actin average age of the male patients (n = 205) was 62.8 ± were obtained from Chemicon International (Temecula, 12.49 years and that of the female patients (n = 167) was CA). Antibodies to EGFR, ATAD3A, DRP1, and SAE2 58.9 ± 9.41 years (t test, P < 0.001). Immunohistochemical were home-made and characterized in the lab (for SAE2 staining was carried out using a single-blind procedure. antibody characterization, please refer to Additional files 6 Eight NSCLC cell lines (H125, H23, H226, H838, H1437, and 7). For immunocytochemistry, the cells were grown H2009, H2087, and A549) from American Type Culture overnight on slides and fixed with cold methanol/acetone Collection (ATCC, Manassas, VA) were used for evaluation for 10 min before staining. Immunological staining was of gene expression. H125, H23, H838, H1437, H2009, performed by an immunoperoxidase method [6–10]. SAE2 H2087, and A549 are LADC cells, and H226 is an epithe- antibodies were not added in the negative control group. lial cell type. Cells were grown at 37 °C in a monolayer in RPMI 1640 supplemented with 10% fetal calf serum (FCS), 100 IU/mL penicillin, and 100 μg/mL streptomycin. Silencing of gene expression using lentivirus-carrying shRNA RT-PCR Lentivirus carrying shRNA sequence to the target gene Reverse transcription-polymerase chain reaction (RT-PCR) was prepared using a three-plasmid transfection method was performed as described before [6–10]. Following ex- [36] and used to infect cells. Cells with gene knockdown KD KD KD KD KD traction of total RNA and synthesis of the first-strand (SAE2 , EGFR ,FAK , c-Met or nAchRα7 ) cDNA, an aliquot of cDNA was subjected to 35 cycles of were selected using 1 μg/mL puromycin. The siRNA PCR to determine the integrity of β-actin mRNA [7]. The used for silencing gene expression was obtained from cDNA used in RT-PCR was adjusted according to the the National RNAi Core Facility (Institute of Molecular quality and quantity of β-actin mRNA. Biology/Genomic Research Centre, Academia Sinica, Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 12 of 14 Taipei, Taiwan). Knockdown of cirRNA CCDC66 was adapted for this study [38]. A specimen was considered KD (cirRNA CCDC66 ) was carried out by direct delivery having strong signals when more than 50% of cancer cells of siRNAs (siJCT1 and siJCT2) [24] into H23 and were positively stained; intermediate, if 25–50% of the cells H1975 cells using Ambion Silencer® siRNA transfection stained positive; weak, if less than 25% or more than 10% of kit [10]. Reduction of cirRNA CCDC66 was determined the cells were positively stained; and negative, if less than by RT-PCR and gel electrophoresis. 10% of the cancer cells were stained. Cases with strong and intermediate signals were classified as SAE2 .Those with Drug sensitivity assay weak or negative signals were classified as SAE2 [6–10]. Drug sensitivity was measured by a WST-1 assay [37]. Cells were seeded at different cell numbers in a 96-well plate Statistical analysis 18 h prior to drug challenge. Cells were pulse-treated with Correlation between SAE2 expression and clinicopatho- various concentrations of cisplatin for 4 h. The control cells logical factors was analyzed by either a chi-square test or were treated with the solvent of the drug. Total survival of a chi-square test for trend. Survival curves were plotted the cells was determined 72 h following drug challenge, using the Kaplan-Meier estimator [39] and analyzed by and percent survival was estimated by dividing optical the log-rank test [40]. Statistical analysis was performed absorbance resulted from each experiment group with that using GraphPad Prism6 (La Jolla, CA). Statistical signifi- of the control group. Each experiment was done in trip- cance was set at P < 0.05. licates, and the optical absorbance was measured by coloration of reacted substrate, WST-1 (BioVision, Additional files Mountain View, CA), which was catalyzed by mito- chondrial dehydrogenases. Additional file 1: The potential microRNA (miRNA) which could suppress protein synthesis of paxillin, vimentin, HIF-1α, SAE2, and EGFR (the data was obtained from the published data or a web search by using online software Matrigel invasion assay programs, miRanda [http://34.236.212.39/microrna/home.do]). (DOCX 26 kb) Matrigel invasion assay was performed following the Additional file 2: Prediction of transmembrane domain in amino acid protocol suggested by BD Biosciences (Bedford, MA, sequences of SAE2. (DOCX 25 kb) USA). Briefly, a vial of BD Matrigel™ basement mem- Additional file 3: The intracellular location of SAE2 and eEF2 in A549 LADC cells as determined by fluorescence immunocytochemical staining. brane matrix (BD-MBM, 356234) was thawed on ice (DOCX 294 kb) overnight and diluted to ½ and ¼ with ice-cold serum- Additional file 4: The intracellular location of cirRNA CCDC66 in A549 free Dulbecco’s modified Eagle’s medium (DMEM). Five LADC cells as determined by fluorescence immunocytochemical staining. milliliters of the diluted BD-MBM was spread to a petri (DOCX 523 kb) dish on ice, before a piece of polycarbonate membrane Additional file 5: Prediction of transmembrane and coiled-coil domains in amino acid sequences of TET1 and TET2. (DOCX 335 kb) (with 5.0-μm pore size) was submerged into the suspen- Additional file 6: Characterization of monoclonal antibodies to SAE2 sion mixture. Membrane coating was carried out at and the immunoprecipitated products. (DOCX 807 kb) room temperature for 1 h. The membrane was rinsed Additional file 7: Supplemental methods. (DOCX 20 kb) with serum-free DMEM once and mounted onto the Boyden chamber. The lower chamber contained the full Abbreviations medium. 1 × 10 cells were pipetted into the well of the AIF: Apoptosis-inducing factor; ATAD3A: ATPase family, AAA domain-containing upper chamber at a 1-h interval for 8 h and incubated at 3A; ATM: Ataxia-telangiectasia-mutated kinase; CCDC66: Coiled-coil domain containing 66; CIP: Calf intestinal phosphatase; cirRNA: Circular RNA; 37 °C for 24 h in a humidified incubator with 5% CO . EGFR: Human epidermal growth factor receptor (HER-1, or ErbB-1); Following complete removal of non-invading cells, the EMT: Epithelial-to-mesenchymal transition; HGF: Hepatocyte growth factor; membrane was lifted from the chamber and fixed in LADC: Lung adenocarcinoma; MALDI-TOF: Matrix-assisted laser desorption/ ionization time of flight mass spectrometry; NSCLC: Non-small cell lung 100% methanol for 2 min. Cells on the membrane were carcinoma; pCIRC2-cirRNA CCDC66: Plasmid containing cDNA insert for stained with 1% toluidine blue for 2 min and washed expressing cirRNA CCDC66; PI3K: Phosphatidylinositol 3-kinase; RT-PCR: Reverse twice with distilled water. After counting cells, percent transcription-polymerase chain reaction; SAE1: SUMO-activating enzyme subunit 1; SAE2: SUMO-activating enzyme subunit 2; SNAI 1: Snail family zinc invasion on membrane were calculated by comparing finger 1; SUMO: Small ubiquitin-related modifier; TET: Ten-eleven translocases; the experimental to the control group. Some assays were TKIs: Tyrosine kinase inhibitors done by green fluorescence-labeled cells. Acknowledgements We thank Drs. Kuei-Yang Hsiao and Shaw-Jeng Tsai (Department of Physiology, Slide evaluation of SAE2 expression by National Chung Kung University, Tainan, Taiwan) for providing the primers for immunohistochemical staining detecting cirRNA CCDC66 expression. In each pathological section, non-tumor lung tissue (NTLT) Funding served as the internal negative control. Slides were evaluated This work was supported, in part, by the Ministry of Science and Technology by two independent pathologists blinded to the clinicopath- (MOST 104-2320-B-005-012-, K.C. Chow), Taiwan, for the design of this study ological knowledge. The Immunoreactive Scoring System as well as the collection, analysis, and interpretation of data. This work was Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 13 of 14 also supported, in part, by the Cancer Prevention Research Institute of Texas 9. Fang HY, Chang CL, Hsu SH, Huang CY, Chiang SF, Chiou SH, et al. ATPase training grant (RP160283; S.K. Gupta), and by the Cancer Prevention Research family AAA domain-containing 3A is a novel anti-apoptotic factor in lung Institute of Texas HIHRRA (RP160795; L. Yen) for constructing the expression adenocarcinoma cells. J Cell Sci. 2010;123:1171–80. vector pCIRC2-cirRNA CCDC66 and interpretation of the transfection data. 10. Chen CY, Fang HY, Chiou SH, Yi SE, Huang CY, Chiang SF, et al. Sumoylation of eukaryotic elongation factor 2 is vital for protein stability and anti-apoptotic Availability of data and materials activity in lung adenocarcinoma cells. Cancer Sci. 2011;102:1582–9. Data sharing is not applicable to this article as no datasets were generated 11. Wang Y, Dasso M. SUMOylation and deSUMOylation at a glance. J Cell Sci. or analyzed during the current study. 2009;122:4249–52. 12. Flotho A, Melchior F. Sumoylation: a regulatory protein modification in Authors’ contributions health and disease. Annu Rev Biochem. 2013;82:357–85. KCC conceived and designed the experiments; NAJ, SHC, ZL, CLY, TYL, HWC, 13. Tatham MH, Jaffray E, Vaughan OA, Desterro JM, Botting CH, Naismith JH, SKG, LY, SDW, and KCC performed the experiments; SHC, TYL, HSS, LY, and et al. Polymeric chains of SUMO-2 and SUMO-3 are conjugated to protein KCC contributed to the statistical analysis and interpretation of the data; NAJ, substrates by SAE1/SAE2 and Ubc9. J Biol Chem. 2001;276:35368–74. SHC, and KCC wrote the paper; SHC, TYL, HSS, SKG, LY, and KCC contributed 14. Guo D, Li M, Zhang Y, Yang P, Eckenrode S, Hopkins D, et al. A functional reagents/materials/analysis tools; HSS, SHC, and KCC contributed to the variant of SUMO4, a new I kappa B alpha modifier, is associated with type 1 study supervision and coordination. All authors read and approved the diabetes. Nat Genet. 2004;36:837–41. final manuscript. 15. Kessler JD, Kahle KT, Sun T, Meerbrey KL, Schlabach MR, Schmitt EM, et al. A sumoylation-dependent transcriptional subprogram is required Ethics approval and consent to participate for Myc-driven tumorigenesis. Science. 2012;335:348–53. The Medical Ethics Committee of China Medical University Hospital had 16. Morris JR, Boutell C, Keppler M, Densham R, Weekes D, Alamshah A, et al. approved the protocol, and written informed consent was obtained from The SUMO modification pathway is involved in the BRCA1 response to every patient before surgery (DMR99-IRB-203). genotoxic stress. Nature. 2009;462:886–90. 17. Sun L,LiH,ChenJ, Dehennaut V, ZhaoY,YangY,et al. A Consent for publication SUMOylation-dependent pathway regulates SIRT1 transcription and lung Not applicable cancer metastasis. J Natl Cancer Inst. 2013;105:887–98. 18. Castillo-Lluva S, Tatham MH, Jones RC, Jaffray EG, Edmondson RD, Hay RT, Competing interests et al. Sumoylation of the GTPase Rac1 is required for optimal cell migration. The authors declare that they have no competing interests. Nat Cell Biol. 2010;12:1078–85. 19. Yano S, Wang W, Li Q, Matsumoto K, Sakurama H, Nakamura T, et al. Hepatocyte growth factor induces gefitinib resistance of lung Publisher’sNote adenocarcinoma with epidermal growth factor receptor-activating Springer Nature remains neutral with regard to jurisdictional claims in mutations. Cancer Res. 2008;68:9479–87. published maps and institutional affiliations. 20. Du L, Li YJ, Fakih M, Wiatrek RL, Duldulao M, Chen Z, et al. Role of SUMO activating enzyme in cancer stem cell maintenance and self-renewal. Nat Author details 1 Commun. 2016;7:12326. https://doi.org/10.1038/ncomms12326. Graduate Institute of Biomedical Sciences, National Chung Hsing University, 2 21. Dasari V, Gallup M, Lemjabbar H, Maltseva I, McNamara N. Induction of Taichung, Taiwan. Graduate Institute of Microbiology and Public Health, 3 epithelial-mesenchymal transition in lung cancer: is tobacco the “smoking National Chung Hsing University, Taichung, Taiwan. Department of gun”? Am J Respir Cell Mol Biol. 2006;35:3–9. Molecular and Cell Biology, University of California, Berkeley, CA, USA. 22. Tominaga T, Tsuchiya T, Mochinaga K, Arai J, Yamasaki N, Matsumoto K, Department of Pathology, China Medical University Hospital, Taichung, 5 et al. Epidermal growth factor signals regulate dihydropyrimidine Taiwan. Institute of Molecular Medicine, National Chung Kung University, Tainan, dehydrogenase expression in EGFR-mutated non-small-cell lung cancer. Taiwan. Department of Pathology and Immunology, Baylor College of Medicine, BMC Cancer. 2016;16:354. https://doi.org/10.1186/s12885-016-2392-0. Houston, TX, USA. School of Post-Baccalaureate Chinese Medicine, College of 23. You WC, Chiou SH, Huang CY, Chiang SF, Yang CL, Sudhakar JN, et al. Chinese Medicine, China Medical University, Taichung, Taiwan. Mitochondrial protein ATPase family, AAA domain containing 3A correlates with radioresistance in glioblastoma. Neuro-Oncology. 2013;15:1342–52. Received: 27 October 2017 Accepted: 22 January 2018 24. Hsiao KY, Lin YC, Gupta SK, Chang N, Yen L, Sun HS, et al. Noncoding effects of circular RNA CCDC66 promote colon cancer growth and metastasis. Cancer Res. 2017;77:2339–50. References 25. Wu JY, Wu SG, Yang CH, Chang YL, Chang YC, Hsu YC, et al. Comparison of 1. Annual reports of the Department of Health, the Executive Yuan, Republic gefitinib and erlotinib in advanced NSCLC and the effect of EGFR of China, 2010. mutations. Lung Cancer. 2011;72:205–12. 2. Schiller JH, Harrington D, Belani CP, Langer C, Sandler A, Krook J, et al. 26. Kasahara K, Arao T, Sakai K, Matsumoto K, Sakai A, Kimura H, et al. Impact of Comparison of four chemotherapy regimens for advanced non-small-cell serum hepatocyte growth factor on treatment response to epidermal lung cancer. N Engl J Med. 2002;346:92–8. growth factor receptor tyrosine kinase inhibitors in patients with non-small 3. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin. cell lung adenocarcinoma. Clin Cancer Res. 2010;16:4616–24. 2014;64:9–29. 27. Pao W, Miller VA, Politi KA, Riely GJ, Somwar R, Zakowski MF, et al. Acquired 4. Huang SF, Liu HP, Li LH, Ku YC, Fu YN, Tsai HY, et al. High frequency of resistance of lung adenocarcinomas to gefitinib or erlotinib is associated epidermal growth factor receptor mutations with complex patterns in non-small with a second mutation in the EGFR kinase domain. PLoS Med. 2005;2:e73. cell lung cancers related to gefitinib responsiveness in Taiwan. Clin Cancer Res. 28. Maseki S, Ijichi K, Tanaka H, Fujii M, Hasegawa Y, Ogawa T, et al. Acquisition 2004;10:8195–203. of EMT phenotype in the gefitinib-resistant cells of a head and neck 5. Lee CK, Brown C, Gralla RJ, Hirsh V, Thongprasert S, Tsai CM, et al. Impact of squamous cell carcinoma cell line through Akt/GSK-3β/snail signaling EGFR inhibitor in non-small cell lung cancer on progression-free and overall pathway. Br J Cancer. 2012;106:1196–204. survival: a meta-analysis. J Natl Cancer Inst. 2013;105:595–605. 29. Levy BP, Rao P, Becker DJ, Becker K. Attacking a moving target: understanding 6. Chen JT, Lin TS, Chow KC, Huang HH, Chiou SH, Chiang SF, et al. Cigarette resistance and managing progression in EGFR-positive lung cancer patients smoking induces overexpression of HGF in type II pneumocytes and lung treated with tyrosine kinase inhibitors. Oncology (Williston Park). 2016; cancer cells. Am J Respir Cell Mol Biol. 2006;34:264–73. 7. Hsu NY, Ho HC, Chow KC, Lin TY, Shih CS, Wang LS, et al: Overexpression of 30. Yin J, Lee JH, Zhang J, Gao Z, Polotsky VY, Ye J. Regulation of hepatocyte dihydrodiol dehydrogenase as a prognostic marker of non-small cell lung growth factor expression by NF-κB and PPARγ in adipose tissue. cancer. Cancer Res 2001;61:2727-2731. Am J Physiol Endocrinol Metab. 2014;306:E929–36. 8. Chiang YY, Chen SL, Hsiao YT, Huang CH, Lin TY, Chiang IP, et al. Nuclear 31. Nagai T, Arao T, Furuta K, Sakai K, Kudo K, Kaneda H, et al. Sorafenib inhibits expression of dynamin-related protein 1 in lung adenocarcinomas. Mod the hepatocyte growth factor-mediated epithelial mesenchymal transition Pathol. 2009;22:1139–50. in hepatocellular carcinoma. Mol Cancer Ther. 2011;10:169–77. Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 14 of 14 32. Shen J, Xia W, Khotskaya YB, Huo L, Nakanishi K, Lim SO, et al. EGFR modulates microRNA maturation in response to hypoxia through phosphorylation of AGO2. Nature. 2013;497:383–7. 33. Song SJ, Poliseno L, Song MS, Ala U, Webster K, Ng C, et al. MicroRNA- antagonism regulates breast cancer stemness and metastasis via TET-family- dependent chromatin remodeling. Cell. 2013;154:311–24. 34. Sudhakar JN, Chow KC. Human RAD23 homolog A is required for the nuclear translocation of apoptosis-inducing factor during induction of cell death. Biol Cell. 2014;106:359–76. 35. Mountain CF. Revisions in the international system for staging lung cancer. Chest. 1997;111:1710–7. 36. Smith JS, Alderete B, Minn Y, Borell TJ, Perry A, Mohapatra G, et al. Localization of common deletion regions on 1p and 19q in human glioma and their association with histological subtype. Oncogene. 1999;18:4144–52. 37. Ishiyama M, Tominaga H, Shiga M, Sasamoto K, Ohkura Y, Ueno K. A combined assay of cell viability and in vitro cytotoxicity with a highly water-soluble tetrazolium salt, neutral red and crystal violet. Biol Pharm Bull. 1996;19:1518–20. 38. Remmele W, Schicketanz KH. Immunohistochemical determination of estrogen and progesterone receptor content in human breast cancer. Computer-assisted image analysis (QIC score) vs. subjective grading (IRS). Pathol Res Pract. 1993;189:862–6. 39. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457–81. 40. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep. 1966;50:163–70. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries � Our selector tool helps you to find the most relevant journal � We provide round the clock customer support � Convenient online submission � Thorough peer review � Inclusion in PubMed and all major indexing services � Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Hematology & Oncology Springer Journals

The role of HGF-MET pathway and CCDC66 cirRNA expression in EGFR resistance and epithelial-to-mesenchymal transition of lung adenocarcinoma cells

Free
14 pages
Loading next page...
 
/lp/springer_journal/the-role-of-hgf-met-pathway-and-ccdc66-cirrna-expression-in-egfr-z5PmheUx2O
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Oncology; Hematology; Cancer Research
eISSN
1756-8722
D.O.I.
10.1186/s13045-018-0557-9
Publisher site
See Article on Publisher Site

Abstract

Background: Epithelial-to-mesenchymal transition (EMT) has, in recent years, emerged as an important tumor cell behavior associated with high metastatic potential and drug resistance. Interestingly, protein SUMOylation and hepatocyte growth factor could respectively reduce the effect of small molecule inhibitors on tyrosine kinase activity of mutated epidermal growth factor receptor of lung adenocarcinomas (LADC). The actual mechanism is yet to be resolved. Methods: Immunohistochemistry was used to stain proteins in LADC specimens. Protein expression was confirmed by Western blotting. In vitro, expression of proteins was determined by Western blotting and immunocytochemistry. Levels of circular RNA were determined by reverse transcription-polymerase chain reaction. Results: SAE2 and cirRNA CCDC66 were highly expressed in LADC. Expression of SAE2 was mainly regulated by EGFR; however, expression of cirRNA CCDC66 was positively regulated by FAK and c-Met but negatively modulated by nAchR7α. EGFR-resistant H1975 also highly expressed cirRNA CCDC66. Immediate response of hypoxia increased phosphorylated c-Met, SAE2, and epithelial-to-mesenchymal transition. Either activation of FAK or silencing of nAchR7α increased cirRNA CCDC66. Conclusions: HGF/c-Met regulates expression of SAE2 and cirRNA CCDC66toincreaseEMT anddrugresistanceofLADC cells. Multimodality drugs concurrently aiming at these targets would probably provide more benefits for cancer patients. Keywords: SUMOylation, EGFR, SAE2, EMT, HGF, c-MET, cirRNA, CCDC66 Background improved living environments, the annual incidence and Elevated tumor cell proliferation and metastatic potential death of lung cancer increase incessantly, including the are major characteristics of lung adenocarcinoma (LADC), USA and Taiwan [3](https://www.mohw.gov.tw/lp-3327- associating with increased resistance to radio- and chemo- 2.html). The appalling escalation was particularly notable in therapy, as well as early tumor recurrence, fast disease female patients. Unfortunately, many new patients were progression, and high mortality rate [1, 2]. Although pub- mostly young non-smokers who died of lung cancer due to lic policy of restricting cigarette retails and smoking has the early recurrence and multi-organ involvement [1–3] (http://www.mohw.gov.tw/cht/DOS/Statistic.aspx). Recent development of chemotherapy targeting mutated * Correspondence: shchiou@dragon.nchu.edu.tw; hssun@mail.ncku.edu.tw; kcchow668@gmail.com human epidermal growth factor receptor (EGFR) by tyro- Graduate Institute of Microbiology and Public Health, National Chung Hsing sine kinase inhibitors (TKIs) has become a major strategic University, Taichung, Taiwan 5 treatment for patients in the Far East [4]. However, not all Institute of Molecular Medicine, National Chung Kung University, Tainan, Taiwan Graduate Institute of Biomedical Sciences, National Chung Hsing University, patients with the same clinical features had the same treat- Taichung, Taiwan ment achievements [5]. Moreover, TKI improved only Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 2 of 14 progression-free survival but not overall survival, suggesting In this report, we showed their correlations via coiled-coil that TKIs might not be effective for full-blown and large- domain containing 66 (CCDC66) circular RNA (cirRNA) scale metastasis and that sensitivity of tumor cells might be to induce EGFR resistance as well as to increase EMT and affected by other factors. To determine whether the disease metastatic potential of LADC cells. progression correlates with specific gene expression pro- files, we combined differential displays and microarrays to Results analyze gene expression patterns in LADC specimens at Expressions of intracellular markers, ATAD3A and SAE2, variousstagesofdisease progression[6]. We identified sev- influence the sensitivity of EGFR-mutated LADC cells to eral potential genes, including aldo-keto reductases (also tyrosine kinase inhibitors named dihydrodiol dehydrogenase, DDH), hepatocyte LADC cells with mutant EGFR (H1975 cells) overexpressed growth factor (HGF), HGF receptor (or c-Met), dynamin- ATAD3A (Fig. 1a, b), an anti-apoptotic factor [9], but rarely related protein 1 (DRP1) and the ATPase family, and expressed SAE2. Molecular weight of EGFR in H1975 AAA domain-containing 3A (ATAD3A) [6–9]. Interest- (~ 185 kDa) was higher than in A549 (170 kDa), suggest- ingly, the protein level of ATAD3A increased dramatically ing that the protein was phosphorylated. Interestingly, when cancer cells were serum-starved. Furthermore, some of H1975 cells (dark arrows) resembled phenotypes through nicotinic acetylcholine receptor alpha 7 of mesenchymal cells [21](Fig. 1b). Silencing of SAE2 in- (nAChRα7), nicotine could increase expression of HGF creased ATAD3A expression (Fig. 1c) and sensitivity of and eukaryotic elongation factor 2 (eEF2), which in turn A549 cells to gefitinib (Fig. 1d) and erlotinib (Fig. 1e). Sen- KD upregulate ATAD3A and DRP1 [6, 10]. Nicotine also in- sitivity of SAE2 A549 cells to gefitinib was equivalent to KD creases eEF2 SUMOylation, which is vital for maintaining that of H1975 cells (ID ~25 μM). However, SAE2 anti-apoptotic capability of LADC. A549 cells (ID ~50 μM) were more resistant than SUMOylation is a post-translational modification of H1975 cells (ID ~5.6 μM) to erlotinib. Silencing of KD proteins by a SUMO (small ubiquitin-related modifier) ATAD3A (ATAD3A ), on the other hand, reduced ex- protein (conjugating to an epsilon lysine residue of the pression of EGFR and AIF (Fig. 1f), however, increased substrate protein) and induces protein conformation sensitivity to both (Fig. 1g)gefitinib and(Fig. 1h) erlotinib. change to enhance transcription and cell cycle progression [11]. Before reaction, the C-terminus of SUMO precursor Expression pattern of membrane receptors and is removed by SUMO-specific proteases (SUPs) to expose intracellular transport markers in LADC cell lines the glycine-glycine motif, which is activated by a heterodi- In vitro, expression of EGFR was low in H23 and H2009, meric enzyme complex, composed of SUMO-activating moderate in H838 and H2087 (5- to 10-fold higher than in enzyme subunit 1 (SAE1) and SAE2, with the expense of H23 and H2009 cells), and high in A549, H226, and H1437 an ATP [12]. The activated SUMO is transferred from cells (10- to 30-fold higher than in H23 and H2009 cells) SAE1/2 to an E2 conjugating protein, ubiquitin-conjugating (Fig. 2a, the upper panel). H1437 cells that had a molecular (UBC) 9, and subsequently conjugated to the final sub- weight upshift was also positive for c-Met and HGF. Ex- strates by an E3 ligase, e.g., SP-RING-type (Siz/PIAS-family) pression of SAE2 was low in H838 and H1437, moderate in E3 ligase or IR-type (internal repeats) E3 ligase [11]. Four H125 (5- to 10-fold higher than in H838 and H1437 cells), types of SUMO proteins have been identified; however, they and high in A549, H23, H226, and H2087 cells (20- to 40- are all activated by the same SAE1/2 and transferred to the fold higher than in H838 and H1437 cells) (Fig. 2a,the substrate proteins via UBC9 [11–14]. lower panel). Interestingly, molecular weights of EGFR in Using a genome-wide RNA interference method to H838, H2087 (~ 175 kDa), and H1437 (~ 185 kDa) were screen cancer cell growth-related genes, Kessler et al. found higher than those in A549, H23, H226, and H2009 that expression of SAE1/2 was correlated with higher meta- (~ 170 kDa) cells. Moreover, sensitivity of H1437 to gefi- static potential of cancer cells [15]. SUMOylation of BrCa1 tinib was equivalent to that of H1975 [22]. has been shown to increase DNA damage repair and drug Although addition of EGF increased levels of phosphory- resistance [16]. Moreover, SUMOylation of SIRT1, a class lated EGFR (pEGFR), no obvious molecular weight upshift III deacetylase sirtuin-1, was found to drive epithelial-to- of EGFR was observed (Fig. 2b). Addition of hepatocyte mesenchymal transition (EMT) [17]. Interestingly, growth factor (HGF), on the other hand, increased molecu- SUMOylation of GTPase Rac1, which was activated by lar weight upshift of EGFR (from 170 to 185 kDa) (Fig. 2c). HGF and PI3K/Akt, not only increased cell migration Treatment with 0.1 U/μgofcalfintestinalphosphatase [18], but also reduced effectiveness of gefitinib on EGFR (CIP) downshifted EGFR molecular weight in H1437 cells mutants [19]. A recent study by Du et al. showed that (Fig. 2d). Silencing of SAE2 did not evidently affect protein SAE2 increased Oct-1 stability to maintain stemness of level of EGFR but reduced HER2 levels (Fig. 2e). Silencing cancer cells [20]. However, the relationships among SAE2, of EGFR reduced protein level of EGFR and that of SAE2 ATAD3A, EGFR, HGF, and EMT are yet to be determined. (Fig. 2f). Short-term silencing of EGFR increased levels of Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 3 of 14 Fig. 1 Expression of intracellular markers, SAE2 and ATAD3A, affect sensitivity of EGFR-mutated lung adenocarcinoma (LADC) cells to tyrosine kinase inhibitors, gefitinib or erlotinib. a LADC cells, carrying mutated EGFR (H1975), expressed abundant ATAD3A, an anti-apoptotic factor [9], as determined by a Western blotting analysis. b ATAD3A protein was highly expressed in EGFR-mutated H1975 (the left panel) and, intermediately, in the wild-type A549 (the right panel) cells. Some of the EGFR-mutated H1975 cells were elongated, scattered, and crescent shape (dark arrows), resembling phenotypes of mesenchymal cells [21], implicating that mutated EGFR could be associated with epithelial-to-mesenchymal transition KD (EMT) of LADC cells. c Knockdown of SAE2 expression (SAE2 ) increased ATAD3A expression. d In SAE2 cells, gefitinib sensitivity was also increased, KD suggesting that SAE2 could be an intracellular sensor of gefitinib. It is therefore worth noting that gefitinib sensitivity of SAE2 A549 cells was KD equivalent to that of H1975 cells (ID50 ~ 25 μM). e Silencing of SAE2 expression increased erlotinib sensitivity of A549 cells as well. However, SAE2 A549 cells (ID ~ 50 μM) were more resistant than H1975 cells (ID ~ 5.6 μM) to erlotinib. f Silencing of ATAD3A (ATAD3A ) reduced expression of EGFR KD KD and AIF. g Increase of gefitinib sensitivity was thus anticipated in ATAD3A cells. h Increment of erlotinib sensitivity was also noted in ATAD3A cells c-Met and pro-c-Met. Prolonged silencing of EGFR for 831, 836, 837, 840, and 844 above 170 kDa; tumor over 96 h, on the other hand, diminished both c-Met and specimens 839, 842, and 843 about 170 kDa) (Fig. 3a). ATAD3A (Fig. 2g), suggesting a physiological feedback cir- Protein level of mutant EGFR was in general lower cuit for regulating homeostasis of intracellular material than that of the wild type. Samples having EGFR transport system, which could then sustain cell survival molecular weight upshifting were mostly positive for under conditions of growth factor deficiency or with c-Met, supporting our previous results that HGF/c- defected receptors [6–10]. Met system was critical for the function of EGFR. While SAE2 expression correlated with EGFR, expres- Biological correlations of EGFR with drug resistance, sion of ATAD3A and AIF was markedly reduced in epithelial-to-mesenchymal transition, and metastatic LADC samples with mutant EGFR. potential of LADC cells Previously, we had shown that c-Met was highly In patient’s biopsies, molecular weights of EGFR var- expressed in LADC biopsies and cell lines [6]. Patients ied from patient to patient (tumor specimens 830, with smoking habits usually had higher HGF level in Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 4 of 14 Fig. 2 Expression pattern of membrane receptors and intracellular transport markers in lung cancer cell lines. a Expression of EGFR detected by immunoblotting was high in A549, H226, and H1437; moderate in H838 and H2087; and low in H23 and H2009 cells. EGFR was not detected in H125 cells. Molecular weight of EGFR in H838, H2087 (about 175 kDa), and H1437 (about 185 kDa) was higher than that in A549, H23, H226, and H2009 (about 170 kDa) cells. Interestingly, only H1437 cells expressed both HGF and c-Met. Expression of SAE2, on the other hand, was detected as a 100-kDa protein in lung cancer cells. Expression of β-actin was a monitoring standard. SAE2 expression was high in A549, H23, H226, and H2087; moderate in H125; and low in H838 and H1437 cells. SAE2 was not detected in H2009 cells. b Addition of EGF increased expression level and the phosphorylated forms (pEGFR) of EGFR in H2009 cells, which did not express c-Met. However, no evident molecular weight upshifting of EGFR was detected. Addition of EGF increased SAE2 protein expression as well. c Addition of HGF increased molecular weight upshifting of EGFR from 170 to 185 kDa in A549 cells. d Following treatment of H1437 cells with calf intestinal phosphatase (CIP), protein levels and molecular weights of 185 kDa EGFR reduced, suggesting that phosphorylation KD was critical for EGFR stability. e Using siRNA to knockdown EGFR (EGFR ) expression for 48 h reduced protein levels of SAE2. f Silencing of EGFR gene KD KD expression for 48 h (EGFR -48hr) reduced SAE2 expression. g However, expression of c-Met and pc-Met was increased in EGFR -48hr cells. Such short-term effect dwindled when EGFR gene silencing was continued to 96 h; and at this time, protein level of c-Met and ATAD3A reduced markedly the diseased lungs. In vitro, hypoxia increased HGF (as measured by Western blotting analysis), hypoxia level and c-Met phosphorylation. Using CoCl to deplete altered intracellular protein distribution (Fig. 3d,as intracellular oxygen also increased phosphorylated c-Met shown by confocal immunofluorescence microscopy). and levels of EGFR and SAE2, as well as markers of Vimentins were gathered in cytoplasmic vacuole-like epithelial-to-mesenchymal transition (EMT), e.g., vimen- structures. In this way, vimentins on plasma membranes tin and paxillin, in LADC cells. The increase was dose- were markedly reduced. Hypoxia affected intracellular dis- and time-dependent (Fig. 3b, c). Silencing of SAE2 or tribution of paxillin and nuclear ataxia-telangiectasia- c-MET alleviated the effect of hypoxia on EMT mutated (ATM) kinase as well. Expression of ATM was markers’ expression (Fig. 3c). Besides protein levels clearly reduced, and most of the proteins were retained Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 5 of 14 Fig. 3 Biological correlations of EGFR with drug resistance, epithelial-to-mesenchymal transition, and metastatic potential of LADC cells. a In vitro, we had shown that the different molecular weights of EGFR in various NSCLC cell lines were closely associated with HGF and c-Met pathways. Molecular weights of EGFR in surgical biopsies of LADC also varied from patient to patient (tumor samples 830, 831, 836, 840, and 844 above 170 kDa; tumor specimens 839, 842, and 843 around 170 kDa). Interestingly, protein levels of mutant EGFR were generally lower than those of wild type. Samples that had higher molecular weights were mostly positive for c-Met and higher serum HGF, supporting our in vitro results that EGFR could be modified through HGF/c-Met pathway. b Using CoCl to induce hypoxia-activated c-Met phosphorylation (p-c-Met) increased protein level of EGFR and SAE2, as well as epithelial-to-mesenchymal transition (EMT) markers, e.g., vimentin and paxillin, in a dose-dependent manner in H23 cells. c c-Met phosphorylation and increase of the related proteins were initiated about 30–60 min following CoCl treatment of H23 cells. Appearance of HIF-1α became visible about 3–6 h following CoCl challenge, indicating that the increase of EMT-related protein expression occurred before traditional hypoxic activation. The hypoxia-mediated protein increases were abolished when SAE2 gene was silenced. Moreover, addition of actinomycin D did not affect CoCl -induced cell responses. d In addition to the increases of proteins as shown by Western blotting analysis, confocal immunofluorescence micrographs revealed that hypoxia altered protein distribution inside the cells as well. For instance, vimentin was accumulated in the enlarged cytoplasmic vacuole-like structures, but not on the plasma membrane. Intracellular distribution of paxillin was generally similar to that of vimentin, except that some cells started to have nuclear signals of paxillin. Nuclear signals of ATM, however, were markedly reduced, and most of the proteins were retained in the cytoplasm in elongated spherical forms. e Exposure of H23 cells (□, control) to hypoxic condition increased resistance to cisplatin (○,hypoxia for12 h; △, hypoxia for 24 h). Using shRNA-lentivirus to silence SAE2 gene expression (by shSAE2-1 [▲]orbyshSAE2-2[●]) increased sensitivity of LADC cells to cisplatin. f Silencing of SAE2 expression in A549 cells reduced cell mobility across extracellular matrix protein-coated carbonate membrane (the left panel). When H23 cells were exposed to hypoxia prior to the cell mobility examination, the cell migration ability increased significantly (the right panel). The statistical differences were significant (WT A549 vs. shSAE2-1 and shSAE2-2, P < 0.0004 [***], ANOVA; control H23 vs. H23 cells exposed to hypoxia, P < 0.0358 [*], ANOVA). [*P < 0.05; ***P < 0.001] mostly in the cytoplasm as irregularly orbicular vesicles, Pathologically, EGFR and SAE2 were concurrently suggesting that hypoxia could interfere intracellular expressed in LADC patients with poor prognosis material transport, which fortuitously activated cell sur- Expression of EGFR and SAE2 was determined by West- vival response [6–10, 23]. ern blotting in 66 pairs of NTLT and LADC specimens. Exposure of LADC cells to CoCl for 12–24 h increased Both EGFR (84.8%, 56/66) and SAE2 (91%, 60/66) were SAE2 expression and cisplatin resistance (Fig. 3e). Silencing detected in most of the tumor fractions (Fig. 4a, b). Ex- of SAE2 expression reduced not only cisplatin resistance pression of SAE2 correlated with EGFR (Pearson r = (Fig. 3e), but also cell mobility across Matrigel-coated 0.454, 95% confidential interval was between 0.238 and membrane (Fig. 3f). Hypoxia increased cell mobility. The 0.627, and R = 0.206). Pathologically, SAE2 was identi- difference was statistically significant (WT vs. shSAE2-1- or fied in the tumor nests (302/372, 81.2%) (Fig. 4c1–c4). shSAE2-2-treated A549 cells, P < 0.0004, ANOVA; control In NTLT, SAE2 signal was detected in type II H23 vs. hypoxic H23 cells, P < 0.0358, ANOVA). pneumocytes (Fig. 4c5). SAE2 was also detected in Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 6 of 14 Fig. 4 Pathologically, EGFR and SAE2 were concurrently expressed in LADC patients with poor survival. a Using a Western blotting method, expression of EGFR was detected with various molecular weights in 18 tumor and eight non-tumor fractions of 20 pair of LADC biopsies. β-actin was used as a monitoring standard. b Using the same method, expression of SAE2 was detected as a 100-kDa protein in 19 tumor and two non-tumor fractions of LADC specimens. N non-tumor lung tissue (NTLT), T tumor fraction of surgical resections. c Representative examples of SAE2 expression in pathological specimens of LADC was determined by immunohistochemical staining (crimson precipitates). Expression of SAE2 was detected in the LADC tumor nests of (c1) papillary, (c2) acinar, (c3) solid, and (c4) mixed subtypes. (c5) In the NTLT, signal of SAE2 was detected in type II pneumocytes (the enlarged area), but not in type I pneumocytes of the air sac. d Comparison of Kaplan-Meier product limit estimates of survival analysis in LADC patients. (d1) Patients were divided into two groups depending on the expression of SAE2. The difference in survival between the two groups was compared by the log-rank test. Survival of patients with low SAE2 level was significantly better than those with high SAE2 level. The hazard ratio between these two groups was 1.708, and the difference in cumulative survival was significant (P = 0.002). (d2) Survival of stage I patients with low SAE2 level was significantly better as well. The hazard ratio was 4.08, and the difference in cumulative survival was significant (P = 0.0053) 87.77% (244/278) of metastatic lymph nodes (data not SAE2, 132 (43.7%) had tumor recurrence. Among the 70 shown). Statistical analysis showed that SAE2 expression patients who had low SAE2, 15 had tumor recurrence correlated with cigarette smoking, tumor stage, and lym- (22.8%). All recurrences were detected within 24 months phovascular involvement (Table 1), suggesting that SAE2 after operation. The recurrence risk of high SAE2 was expression was closely associated with metastatic poten- 2.85-fold higher than that of low SAE2 (P < 0.001). Sur- tial. Interestingly, among the 302 patients who had high vival rate of low SAE2 patients was better than that of Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 7 of 14 Table 1 Correlation of SAE2 expression with clinicopathological (Fig. 5g), suggesting that nAchRα7 could be a feedback parameters in patients with LADC mechanism of HGF-c-MET circuit. Our results showed Expression P value that expression of pEGFR, c-MET, and FAK were higher of SAE2 in EGFR-mutated H1975 than that in A549 cells (Fig. 5h). Parameter High Low Univariate Multivariate Expression of cirRNA CCDC66 was also higher in H1975 (n = 302) (n = 70) cells (Fig. 5i). Levels of nAchRα7, in contrast, were higher Gender in A549 cells. Addition of FAK inhibitor Y15 (1,2,4,5-ben- Male (n = 292) 243 49 0.0549 0.107 zenetetramine tetrahydrochloride, Merck KGaA, Darm- Female (n = 80) 59 21 stadt, Germany) inhibited expression of FAK, nAchR7α, and CCDC66α (Fig. 5j), however, increased cirRNA Cigarette smoking CCDC66 expression (Fig. 5k, l). Smoker (n = 266) 227 39 0.0012 0.0025 Non-smoker (n = 106) 75 31 Relationship between expression of cirRNA CCDC66 and Stage EMT markers in LADC cells I(n = 83) 65 18 0.042 0.091 Expression of cirRNA CCDC66, FAK, and paxillin was II (n = 109) 81 28 determined by RT-PCR in 11 pairs of NTLT and LADC specimens. All three makers were detected in most of the IIIa (n = 180) 156 24 tumor fractions [cirRNA CCDC66 (82%, 9/11), FAK (82%, Cell differentiation 9/11), and paxillin (91%, 10/11)] (Fig. 6a). Pathologically, Well (n = 56) 42 14 0.205 0.41 paxillin was identified in tumor nests and the neighboring Moderate (n = 211) 172 39 area (Fig. 6b1–b3), supporting that paxillin expression Poor (n = 105) 88 17 could correlate with metastatic potential. Knockdown of KD Lymphovascular invasion cirRNA CCDC66 (cirRNA CCDC66 ) expression (Fig. 6c) did not affect resistance to gefitinib or erlotinib in H1975 Positive (n = 278) 244 34 < 0.001 0.0021 cells (Fig. 6d1, d2) but increased cisplatin resistance in H23 Negative (n = 94) 58 36 KD cells (Fig. 6d3). In cirRNA CCDC66 cells, expression of † 2 Two-sided P value determined by χ test ‡ 2 EMT markers (Fig. 6e) and invasion ability (Fig. 6f, g1)were Two-sided P value determined by χ test for trend significantly abrogated. In pCIRC2-cirRNA CCDC66- TF high SAE2 patients (Fig. 4d1). The hazard ratio was transfected A549 cells (cirR ), although invasion ability TF 1.708, and the difference in cumulative survival was increased slightly (compared to cirR H2009 cells), the significant (P = 0.002). Multivariate analysis showed that increment was significant (Fig. 6g2). the difference in SAE2 expression was significant between the two groups (P = 0.0047). Survival rate of Discussion stage I patients with low SAE2 level was also better EGFRs, especially those with tyrosine kinase mutations, (Fig. 4d2, P = 0.0053). The hazard ratio was 4.08. have recently become major targets of TKIs, e.g., gefitinib, erlotinib, and afatinib, in LADC chemotherapy [4, 5, 25]. Expression of CCDC66 cirRNA in LADC cells However, patient’s responses are equivocal [5]. The fact As shown in Fig. 3b, c, the increase of protein synthesis that TKI therapy improves progression-free but not over- before elevation of HIF-1α following exposure to hypoxia all survival [25–27] has raised sincere queries into drug’s suggested that translation efficacy might be increased by effectiveness, which may be affected by other factors. reduction of functional microRNA (miRNA), which Kasahara et al. showed that LADC patients with higher repressed mRNA translation (a list of potential miRNA is HGF level were more resistant to TKIs [26]. Yano et al. shown in Additional file 1). Recently, Hsiao et al. showed demonstrated that gefitinib resistance was mediated via that cirRNA CCDC66 could promote cancer cell growth HGF/c-Met circuit to activate PI3K/AKT [25]. Moreover, and metastasis by adsorbing miRNA [24]. Interestingly, continuous exposure of tumor cells to TKIs transpires cirRNA CCDC66 was also detected in LADC cells extra EGFR mutations and drug resistance [27], suggesting (Fig. 5a), and its expression correlated with the levels of nuclear events of DNA sequence alteration and gene acti- FAK mRNA, a vital marker closely associated with cancer vation. Nonetheless, induction mechanisms of gene muta- metastasis and EMT (Fig. 5b). Silencing of FAK expression tions and phenotype changes, e.g., increase of EMT and (Fig. 5c) reduced cirRNA CCDC66 (Fig. 5d). Addition of metastatic potential [28, 29], are yet to be resolved. HGF, on the contrary, increased expression of cirRNA Kessler et al. had found that higher metastatic potential CCDC66 (Fig. 5e). It is worth noting that silencing of of cancer cells was correlated with expression of SAE1/2 nicotinic acetylcholine receptor alpha 7 (nAchRα7) in- [15]. It was further demonstrated that the elevated meta- creased levels of CCDC66β (Fig. 5f) and cirRNA CCDC66 static potential was activated by SUMOylation of Rac1, Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 8 of 14 Fig. 5 Detection of cirRNA CCDC66 expression in LADC cells. a Expression of cirRNA CCDC66 was detected by RT-PCR in LADC cells. b Its expression correlated with levels of FAK mRNA, a vital marker closely associated with cancer metastasis and EMT. c Silencing of FAK expression reduced d cirRNA CCDC66. e Addition of HGF (50 ng/ml) increased expression of cirRNA CCDC66. f Silencing of nicotinic acetylcholine receptor alpha 7 (nAchRα7), in contrast, increased levels of CCDC66β and g cirRNA CCDC66. h Comparing gene expression patterns of pEGFR, c-MET, and FAK in EGFR-mutated H1975 and wild-type A549 cells. i Expression of pEGFR, c-MET, and FAK was higher in EGFR-mutated H1975 cells. Level of nAchRα7, in contrast, was higher in A549 cells. i Expression of cirRNA CCDC66 was also higher in H1975 cells. j Addition of FAK inhibitor Y15 (1,2,4,5-benzenetetramine tetrahydrochloride) inhibited expression of FAK, nAchR7α, and CCDC66α. Y15 inhibited expression of nAchR7α at a concentration of 7.5 μM, CCDC66α at a concentration of 10 μM, and FAK at a concentration of 15 μM. k Y15, however, increased cirRNA CCDC66 expression. l Addition of nicotine also increased cirRNA CCDC66 expression KD cAMP response element binding (CREB) protein, silencing of SAE2 (SAE2 ) inhibited hypoxic effects, our eEF2, and SIRT1 to increase EMT and cell migration data suggest that SAE2 is able to maintain protein stability [10, 16–18]. By showing that both EGFR and SAE2 are through protein SUMOylation. Moreover, because highly expressed in LADC, and that SAE2 expression cor- addition of actinomycin D did not inhibit hypoxic effects, relates with patient’s poor outcomes, our data support our data also suggest that hypoxia may increase protein their observations that SUMOylation is crucial for tumor synthesis from preexisted mRNA. The rapid rise (0.5 to progression. In particular, via HGF, hypoxia could enhance 1 h, such response was therefore named as immediate SUMOylation of these factors [10, 17, 18]. Among these, reaction) of protein levels before evident accumulation of SUMOylated Rac1 could further induce expression of a HIF-1α supported our findings especially that the panel of genes, including SNAI (snail family zinc finger) 1, increased translational efficacy might be initiated before VEGF, and N-cadherin, to facilitate EMT and metastasis upregulation of transcription (compared to the immediate of cancer cells [30, 31]. reaction, such response was named as delayed reaction). Hypoxia increased expression of EGFR, SAE2, and EMT In fact, the different increase rate of respective pro- markers, such as paxillin and vimentin as well [10]. Since teins suggested that the differential augmentation of Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 9 of 14 Fig. 6 (See legend on next page.) Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 10 of 14 (See figure on previous page.) Fig. 6 Expression of cirRNA CCDC66 correlated with that of FAK and paxillin, and knockdown of cirRNA CCDC66 reduced expression of EMT markers and invasion ability of LADC cells. a Expression of cirRNA CCDC66, FAK, and paxillin was determined by RT-PCR in 11 pairs of LADC specimens. All three tumor makers were detected in most of the tumor fractions [cirRNA CCDC66 (82%, 9/11), FAK (82%, 9/11), and paxillin (91%, 10/11)]. Paxillin was detected in tumor nests (b1–b3; b3, #836) and the neighboring areas (b2, #838) by an immunohistochemical staining. Paxillin-positive cells were in dark brown KD color (indicated by an arrow in b1, #859), showing EMT-like fibroblasts characteristics. c Knockdown of cirRNA CCDC66 (cirRNA CCDC66 )expressionin KD H1975 and H23 cells. Resistance to gefitinib (d1) or erlotinib (d2) was not markedly affected in cirRNA CCDC66 H1975 cells (d1, d2), but clearly increased KD cisplatin resistance in cirRNA CCDC66 H23 cells (d3). e Knockdown of cirRNA CCDC66 expression, however, clearly diminished expression of EMT markers (left panel, H1975 cells were stained for vimentin, red fluorescence; right panel, H23 cells were stained for paxillin, green fluorescence. Nuclei were stained with DAPI, blue fluorescence). f Silencing of cirRNA CCDC66 expression markedly reduced cell mobility across extracellular matrix protein-coated carbonate membrane. H23 cells were labeled with green fluorescence protein prior to Matrigel transmembrane assay. Cells in bottom wells of the Boyden chamber were directly detected by green fluorescence. Nuclei were stained with DAPI (blue fluorescence). g Effect of cirRNA CCDC66 on cell invasion ability. KD g1 Invasion ability of cirRNA CCDC66 , H1975, and H23 cells decreased markedly. Statistical differences were significant (H1975 vs. Mock vs. cirRNA KD KD CCDC66 H1975, P < 0.0001 [***], ANOVA; WT H23 vs. Mock vs. cirRNA CCDC66 H23 cells, P = 0.0048 [**], ANOVA). [*P <0.05; **P < 0.01; ***P < 0.001]; g2 When cirRNA CCDC66 was overexpressed in A549 cells, the invasion ability increased a little (wild-type vs. pCIRC2-mCherry-transfected [pCIRC2m] TF TF vs. pCIRC2-cirRNA CCDC66-transfected A549 cells [cirR ], P = 0.037 [*], ANOVA; wild-type vs. cirR A549 cells, P =0.1054, t test). Comparing wild-type, TF TF pCIRC2m, and cirR H2009 cells, increment of invasion ability was significant (P = 0.0025 [**], ANOVA; H2009 vs. cirR H2009, P = 0.0137 [*], t test). H23 cells exposed to 24-h hypoxia were used as internal comparison parameter (P = 0.0031, t test) translation efficacy could be a result of reducing levels of EGFR, indicating that post-translational modification of suppressive microRNA (miRNA) following hypoxia (a EGFR was taken place on intracellular vesicles, and these list of potential miRNA is shown in Additional file 1). events could occur without the presence of EGF when cells However, we did not find an expression surge of specific were under hypoxic conditions [32]. Our results showed miRNA under hypoxia. Recently, circular RNA (cirRNA) that HGF/c-Met circuit was accountable for such modi- CCDC66 was shown to promote colon cancer cell fication, and this circuit fortuitously activated EGFR. growth and metastasis possibly by adsorbing miRNA, Because HGF-induced phosphorylation did not occur which had inhibited expression of several nuclear gene uniquely on tyrosine residues, phosphorylated serine regulators, including DNA methyltransferase 3β and threonine residues were less sensitive to TKIs. (DNMT3B), enhancer of zeste homolog 2 (EZH2), c-Myc, Moreover, joined effects of c-Met and EGFR on inhibit- and Yes-associated protein 1 (YAP1) [24]. Showing that ing cancer cell metastatic ability could be carried out cirRNA CCDC66 was expressed in LADC cells, and its ex- by repressing miR-200 family and methylcytosine diox- pression correlated with FAK, a marker that was associated ygenases (also called ten-eleven translocases, TET) [33]. with cancer metastasis, our results supported their findings. Two proteins, TET1 (235 kDa) and TET2 (224 kDa), have KD Moreover, silencing of FAK (FAK ) reduced cirRNA been identified in TET family. Both TET1 and TET2 pro- KD CCDC66. Knockdown of nAChRα7(nAChRα7 ), on the teins possess transmembrane and coiled coil domains (as other hand, increased cirRNA CCDC66 expression, even in shown in Additional file 5), which could respectively inter- thepresenceofY15,which inhibitednAChRα7expression act with DRP1 and ATAD3A, the essential components of at low concentrations. The results suggested that the intracellular transport system, to regulate nuclear entry of nAChRα7 effect on cirRNA CCDC66 expression was prior these two proteins and other DNA repair-related enzymes to that of FAK. [9, 23]. Although the nuclear events of these proteins have In addition to cirRNA, miRNA levels are also regulated yet to be elucidated, it is plausible that TKI inhibits EGFR by the miRNA maturation process. Argonaute 2 (AGO2), a activity as well as reduces ATAD3A and SAE2 expression member of RNA-induced silencing complex (RISC), was to limit nuclear entry of DNA repair-related proteins, e.g., inhibited by activated EGFR to decelerate miRNA matur- human RAD23 homolog A (hHR23A) and ataxia- ation [32]. Moreover, like AGO2, SAE2 and cirRNA telangiectasia-mutated (ATM) kinase, to increase DNA CCDC66 were also located on the endoplasmic reticulum damage and mutation frequency [23, 34]. In an ongoing (ER) (intracellular localization of SAE2 and cirRNA study, we are focusing on these issues. CCDC66 is shown in Additional files 2, 3,and 4), suggest- ing that the miRNA maturation and miRNA adsorption Conclusions process could be carried out at the same time on the ER. From this study, we identify novel downstream substrates of Hypoxia-increased buildup of AGO2 was also located on EGFR and c-Met, SAE2, and cirRNA CCDC66, which are the membrane of intracellular vesicles, which removed also expressed in LADC. Through SUMOylation, SAE2 miRNA-associated gene suppression. Hypoxia also aug- could maintain stability of other proteins, including mented levels of SAE2 and EMT markers [8, 9, 19]. Inhib- ATAD3A and EMT markers, such as vimentin and paxillin, ition of mitochondrial oxidative phosphorylation increased which are crucial for metastatic potential and drug resist- molecular weight of AGO2 too, which co-precipitated with ance, and these bioparameters correlate with LADC patient’s Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 11 of 14 prognosis. Therefore, multimodality drugs concurrently The primer sequences were selected by Primer3 aiming at several targets in the circuit of putative feedback (http://bioinfo.ut.ee/primer3/). mechanism of EGFR and drug resistance-related enzymes For FAK, the primers are FAKs: 5′-TGAATTTCTTC would probably provide more benefits for cancer patients. TATCAACAG-3′ [sense primer, nts 485-504, NM_00 1352746] and FAKa: 5′-CCTGGCTTCATCTATTCCAT Methods -3′ [antisense primer, nts 1142-1161]. The anticipated Tissue specimens and NSCLC lines DNA fragment is 677 base-pair (bp). For SAE2, the Some of the patients in this study were from the cohorts primers are SAE2s: 5′-TGAAAGTGGAACAGCTGG used in the previous studies [8, 10]. Briefly, from January GT-3′ [nts 481-500, NM_005499] and SAE2a: 5′-GCA 2007 to December 2012, tissue specimens were collected GGAATAATGTTCCCTGCC-3′ [nts 1207-1227]. The from 628 patients with newly diagnosed non-small cell anticipated cDNA fragment is 847 bp. For nAchRα7, the lung cancer cell (NSCLC). Samples from all patients, for primers are nAchRα7s: 5′-GCCGCAGGACGCTCTACT whom at least one follow-up or death was documented, AT-3′ [nts 912-931, NM_148911] and nAchRα7a: 5′-A were pathologically confirmed NSCLC. Of the 628 CGGCACTCATCTCCACACT-3′ [nts 1334-1353]. The patients, 372 were diagnosed as having LADC. The stage anticipated DNA fragment is 442 base-pair (bp). For cir- of the disease was classified according to the new inter- RNA CCDC66 (circular form_10-8, exons 10 to 8), the national staging system for lung cancer [35]. The primers are CCDC66s: 5′-TCTCTTGGACCCAGCTCAG- Medical Ethics Committee had approved the protocol, 3′ [nts 1428-1446, NM_001141947] and CCDC66a: 5′- and written informed consent was obtained from every TGAATCAAAGTGCATTGCCC-3′ [nts 1175-1194]. The patient before surgery (DMR99-IRB-203). All patients anticipated DNA fragment is 272 bp. For CCDC66 mRNA had undergone surgical resection and radical N2 lymph (linear form_16-17, exons 16 to 17), the primers are node dissection. Tumor size, numbers of lymph node in- CCDC66s: 5′-ACCTTGATCCCGATGCACC-3′ [nts 2657- volvement, differentiation, vascular invasion, and mitotic 2675, NM_001141947] and CCDC66a: 5′-ACGGCACTCA number were evaluated. Patients with lymph node TCTCCACACT-3′ [nts 2801-2820]. The anticipated DNA involvement or loco-regional recurrence received irradi- fragment is 164 bp. For β-actin, the primers are β-actins: 5′- ation at the afflicted areas. Those with distant metastasis AGA GCT ACG AGC TGC CTG AC-3′ [nts 797-816, were treated with chemotherapy. After treatment, patients NM_001101] and β-actins: 5′-CAC CTT CAC CGT TCC were routinely followed every 3 to 6 months in the AGT TT-3′ [nts 1337-1356]. The anticipated DNA frag- outpatient department. Results from blood examination, ment is 560 bp. biochemical studies, chest radiography, abdominal sonog- raphy, whole body bone scan, and computerized tomog- Immunoblotting analysis and immunocytochemistry raphy scans of chest that indicated any evidence of disease Immunoblotting and immunohistochemistry were per- were interpreted as tumor recurrence or metastasis. The formed as described previously [6]. Antibodies for β-actin average age of the male patients (n = 205) was 62.8 ± were obtained from Chemicon International (Temecula, 12.49 years and that of the female patients (n = 167) was CA). Antibodies to EGFR, ATAD3A, DRP1, and SAE2 58.9 ± 9.41 years (t test, P < 0.001). Immunohistochemical were home-made and characterized in the lab (for SAE2 staining was carried out using a single-blind procedure. antibody characterization, please refer to Additional files 6 Eight NSCLC cell lines (H125, H23, H226, H838, H1437, and 7). For immunocytochemistry, the cells were grown H2009, H2087, and A549) from American Type Culture overnight on slides and fixed with cold methanol/acetone Collection (ATCC, Manassas, VA) were used for evaluation for 10 min before staining. Immunological staining was of gene expression. H125, H23, H838, H1437, H2009, performed by an immunoperoxidase method [6–10]. SAE2 H2087, and A549 are LADC cells, and H226 is an epithe- antibodies were not added in the negative control group. lial cell type. Cells were grown at 37 °C in a monolayer in RPMI 1640 supplemented with 10% fetal calf serum (FCS), 100 IU/mL penicillin, and 100 μg/mL streptomycin. Silencing of gene expression using lentivirus-carrying shRNA RT-PCR Lentivirus carrying shRNA sequence to the target gene Reverse transcription-polymerase chain reaction (RT-PCR) was prepared using a three-plasmid transfection method was performed as described before [6–10]. Following ex- [36] and used to infect cells. Cells with gene knockdown KD KD KD KD KD traction of total RNA and synthesis of the first-strand (SAE2 , EGFR ,FAK , c-Met or nAchRα7 ) cDNA, an aliquot of cDNA was subjected to 35 cycles of were selected using 1 μg/mL puromycin. The siRNA PCR to determine the integrity of β-actin mRNA [7]. The used for silencing gene expression was obtained from cDNA used in RT-PCR was adjusted according to the the National RNAi Core Facility (Institute of Molecular quality and quantity of β-actin mRNA. Biology/Genomic Research Centre, Academia Sinica, Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 12 of 14 Taipei, Taiwan). Knockdown of cirRNA CCDC66 was adapted for this study [38]. A specimen was considered KD (cirRNA CCDC66 ) was carried out by direct delivery having strong signals when more than 50% of cancer cells of siRNAs (siJCT1 and siJCT2) [24] into H23 and were positively stained; intermediate, if 25–50% of the cells H1975 cells using Ambion Silencer® siRNA transfection stained positive; weak, if less than 25% or more than 10% of kit [10]. Reduction of cirRNA CCDC66 was determined the cells were positively stained; and negative, if less than by RT-PCR and gel electrophoresis. 10% of the cancer cells were stained. Cases with strong and intermediate signals were classified as SAE2 .Those with Drug sensitivity assay weak or negative signals were classified as SAE2 [6–10]. Drug sensitivity was measured by a WST-1 assay [37]. Cells were seeded at different cell numbers in a 96-well plate Statistical analysis 18 h prior to drug challenge. Cells were pulse-treated with Correlation between SAE2 expression and clinicopatho- various concentrations of cisplatin for 4 h. The control cells logical factors was analyzed by either a chi-square test or were treated with the solvent of the drug. Total survival of a chi-square test for trend. Survival curves were plotted the cells was determined 72 h following drug challenge, using the Kaplan-Meier estimator [39] and analyzed by and percent survival was estimated by dividing optical the log-rank test [40]. Statistical analysis was performed absorbance resulted from each experiment group with that using GraphPad Prism6 (La Jolla, CA). Statistical signifi- of the control group. Each experiment was done in trip- cance was set at P < 0.05. licates, and the optical absorbance was measured by coloration of reacted substrate, WST-1 (BioVision, Additional files Mountain View, CA), which was catalyzed by mito- chondrial dehydrogenases. Additional file 1: The potential microRNA (miRNA) which could suppress protein synthesis of paxillin, vimentin, HIF-1α, SAE2, and EGFR (the data was obtained from the published data or a web search by using online software Matrigel invasion assay programs, miRanda [http://34.236.212.39/microrna/home.do]). (DOCX 26 kb) Matrigel invasion assay was performed following the Additional file 2: Prediction of transmembrane domain in amino acid protocol suggested by BD Biosciences (Bedford, MA, sequences of SAE2. (DOCX 25 kb) USA). Briefly, a vial of BD Matrigel™ basement mem- Additional file 3: The intracellular location of SAE2 and eEF2 in A549 LADC cells as determined by fluorescence immunocytochemical staining. brane matrix (BD-MBM, 356234) was thawed on ice (DOCX 294 kb) overnight and diluted to ½ and ¼ with ice-cold serum- Additional file 4: The intracellular location of cirRNA CCDC66 in A549 free Dulbecco’s modified Eagle’s medium (DMEM). Five LADC cells as determined by fluorescence immunocytochemical staining. milliliters of the diluted BD-MBM was spread to a petri (DOCX 523 kb) dish on ice, before a piece of polycarbonate membrane Additional file 5: Prediction of transmembrane and coiled-coil domains in amino acid sequences of TET1 and TET2. (DOCX 335 kb) (with 5.0-μm pore size) was submerged into the suspen- Additional file 6: Characterization of monoclonal antibodies to SAE2 sion mixture. Membrane coating was carried out at and the immunoprecipitated products. (DOCX 807 kb) room temperature for 1 h. The membrane was rinsed Additional file 7: Supplemental methods. (DOCX 20 kb) with serum-free DMEM once and mounted onto the Boyden chamber. The lower chamber contained the full Abbreviations medium. 1 × 10 cells were pipetted into the well of the AIF: Apoptosis-inducing factor; ATAD3A: ATPase family, AAA domain-containing upper chamber at a 1-h interval for 8 h and incubated at 3A; ATM: Ataxia-telangiectasia-mutated kinase; CCDC66: Coiled-coil domain containing 66; CIP: Calf intestinal phosphatase; cirRNA: Circular RNA; 37 °C for 24 h in a humidified incubator with 5% CO . EGFR: Human epidermal growth factor receptor (HER-1, or ErbB-1); Following complete removal of non-invading cells, the EMT: Epithelial-to-mesenchymal transition; HGF: Hepatocyte growth factor; membrane was lifted from the chamber and fixed in LADC: Lung adenocarcinoma; MALDI-TOF: Matrix-assisted laser desorption/ ionization time of flight mass spectrometry; NSCLC: Non-small cell lung 100% methanol for 2 min. Cells on the membrane were carcinoma; pCIRC2-cirRNA CCDC66: Plasmid containing cDNA insert for stained with 1% toluidine blue for 2 min and washed expressing cirRNA CCDC66; PI3K: Phosphatidylinositol 3-kinase; RT-PCR: Reverse twice with distilled water. After counting cells, percent transcription-polymerase chain reaction; SAE1: SUMO-activating enzyme subunit 1; SAE2: SUMO-activating enzyme subunit 2; SNAI 1: Snail family zinc invasion on membrane were calculated by comparing finger 1; SUMO: Small ubiquitin-related modifier; TET: Ten-eleven translocases; the experimental to the control group. Some assays were TKIs: Tyrosine kinase inhibitors done by green fluorescence-labeled cells. Acknowledgements We thank Drs. Kuei-Yang Hsiao and Shaw-Jeng Tsai (Department of Physiology, Slide evaluation of SAE2 expression by National Chung Kung University, Tainan, Taiwan) for providing the primers for immunohistochemical staining detecting cirRNA CCDC66 expression. In each pathological section, non-tumor lung tissue (NTLT) Funding served as the internal negative control. Slides were evaluated This work was supported, in part, by the Ministry of Science and Technology by two independent pathologists blinded to the clinicopath- (MOST 104-2320-B-005-012-, K.C. Chow), Taiwan, for the design of this study ological knowledge. The Immunoreactive Scoring System as well as the collection, analysis, and interpretation of data. This work was Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 13 of 14 also supported, in part, by the Cancer Prevention Research Institute of Texas 9. Fang HY, Chang CL, Hsu SH, Huang CY, Chiang SF, Chiou SH, et al. ATPase training grant (RP160283; S.K. Gupta), and by the Cancer Prevention Research family AAA domain-containing 3A is a novel anti-apoptotic factor in lung Institute of Texas HIHRRA (RP160795; L. Yen) for constructing the expression adenocarcinoma cells. J Cell Sci. 2010;123:1171–80. vector pCIRC2-cirRNA CCDC66 and interpretation of the transfection data. 10. Chen CY, Fang HY, Chiou SH, Yi SE, Huang CY, Chiang SF, et al. Sumoylation of eukaryotic elongation factor 2 is vital for protein stability and anti-apoptotic Availability of data and materials activity in lung adenocarcinoma cells. Cancer Sci. 2011;102:1582–9. Data sharing is not applicable to this article as no datasets were generated 11. Wang Y, Dasso M. SUMOylation and deSUMOylation at a glance. J Cell Sci. or analyzed during the current study. 2009;122:4249–52. 12. Flotho A, Melchior F. Sumoylation: a regulatory protein modification in Authors’ contributions health and disease. Annu Rev Biochem. 2013;82:357–85. KCC conceived and designed the experiments; NAJ, SHC, ZL, CLY, TYL, HWC, 13. Tatham MH, Jaffray E, Vaughan OA, Desterro JM, Botting CH, Naismith JH, SKG, LY, SDW, and KCC performed the experiments; SHC, TYL, HSS, LY, and et al. Polymeric chains of SUMO-2 and SUMO-3 are conjugated to protein KCC contributed to the statistical analysis and interpretation of the data; NAJ, substrates by SAE1/SAE2 and Ubc9. J Biol Chem. 2001;276:35368–74. SHC, and KCC wrote the paper; SHC, TYL, HSS, SKG, LY, and KCC contributed 14. Guo D, Li M, Zhang Y, Yang P, Eckenrode S, Hopkins D, et al. A functional reagents/materials/analysis tools; HSS, SHC, and KCC contributed to the variant of SUMO4, a new I kappa B alpha modifier, is associated with type 1 study supervision and coordination. All authors read and approved the diabetes. Nat Genet. 2004;36:837–41. final manuscript. 15. Kessler JD, Kahle KT, Sun T, Meerbrey KL, Schlabach MR, Schmitt EM, et al. A sumoylation-dependent transcriptional subprogram is required Ethics approval and consent to participate for Myc-driven tumorigenesis. Science. 2012;335:348–53. The Medical Ethics Committee of China Medical University Hospital had 16. Morris JR, Boutell C, Keppler M, Densham R, Weekes D, Alamshah A, et al. approved the protocol, and written informed consent was obtained from The SUMO modification pathway is involved in the BRCA1 response to every patient before surgery (DMR99-IRB-203). genotoxic stress. Nature. 2009;462:886–90. 17. Sun L,LiH,ChenJ, Dehennaut V, ZhaoY,YangY,et al. A Consent for publication SUMOylation-dependent pathway regulates SIRT1 transcription and lung Not applicable cancer metastasis. J Natl Cancer Inst. 2013;105:887–98. 18. Castillo-Lluva S, Tatham MH, Jones RC, Jaffray EG, Edmondson RD, Hay RT, Competing interests et al. Sumoylation of the GTPase Rac1 is required for optimal cell migration. The authors declare that they have no competing interests. Nat Cell Biol. 2010;12:1078–85. 19. Yano S, Wang W, Li Q, Matsumoto K, Sakurama H, Nakamura T, et al. Hepatocyte growth factor induces gefitinib resistance of lung Publisher’sNote adenocarcinoma with epidermal growth factor receptor-activating Springer Nature remains neutral with regard to jurisdictional claims in mutations. Cancer Res. 2008;68:9479–87. published maps and institutional affiliations. 20. Du L, Li YJ, Fakih M, Wiatrek RL, Duldulao M, Chen Z, et al. Role of SUMO activating enzyme in cancer stem cell maintenance and self-renewal. Nat Author details 1 Commun. 2016;7:12326. https://doi.org/10.1038/ncomms12326. Graduate Institute of Biomedical Sciences, National Chung Hsing University, 2 21. Dasari V, Gallup M, Lemjabbar H, Maltseva I, McNamara N. Induction of Taichung, Taiwan. Graduate Institute of Microbiology and Public Health, 3 epithelial-mesenchymal transition in lung cancer: is tobacco the “smoking National Chung Hsing University, Taichung, Taiwan. Department of gun”? Am J Respir Cell Mol Biol. 2006;35:3–9. Molecular and Cell Biology, University of California, Berkeley, CA, USA. 22. Tominaga T, Tsuchiya T, Mochinaga K, Arai J, Yamasaki N, Matsumoto K, Department of Pathology, China Medical University Hospital, Taichung, 5 et al. Epidermal growth factor signals regulate dihydropyrimidine Taiwan. Institute of Molecular Medicine, National Chung Kung University, Tainan, dehydrogenase expression in EGFR-mutated non-small-cell lung cancer. Taiwan. Department of Pathology and Immunology, Baylor College of Medicine, BMC Cancer. 2016;16:354. https://doi.org/10.1186/s12885-016-2392-0. Houston, TX, USA. School of Post-Baccalaureate Chinese Medicine, College of 23. You WC, Chiou SH, Huang CY, Chiang SF, Yang CL, Sudhakar JN, et al. Chinese Medicine, China Medical University, Taichung, Taiwan. Mitochondrial protein ATPase family, AAA domain containing 3A correlates with radioresistance in glioblastoma. Neuro-Oncology. 2013;15:1342–52. Received: 27 October 2017 Accepted: 22 January 2018 24. Hsiao KY, Lin YC, Gupta SK, Chang N, Yen L, Sun HS, et al. Noncoding effects of circular RNA CCDC66 promote colon cancer growth and metastasis. Cancer Res. 2017;77:2339–50. References 25. Wu JY, Wu SG, Yang CH, Chang YL, Chang YC, Hsu YC, et al. Comparison of 1. Annual reports of the Department of Health, the Executive Yuan, Republic gefitinib and erlotinib in advanced NSCLC and the effect of EGFR of China, 2010. mutations. Lung Cancer. 2011;72:205–12. 2. Schiller JH, Harrington D, Belani CP, Langer C, Sandler A, Krook J, et al. 26. Kasahara K, Arao T, Sakai K, Matsumoto K, Sakai A, Kimura H, et al. Impact of Comparison of four chemotherapy regimens for advanced non-small-cell serum hepatocyte growth factor on treatment response to epidermal lung cancer. N Engl J Med. 2002;346:92–8. growth factor receptor tyrosine kinase inhibitors in patients with non-small 3. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin. cell lung adenocarcinoma. Clin Cancer Res. 2010;16:4616–24. 2014;64:9–29. 27. Pao W, Miller VA, Politi KA, Riely GJ, Somwar R, Zakowski MF, et al. Acquired 4. Huang SF, Liu HP, Li LH, Ku YC, Fu YN, Tsai HY, et al. High frequency of resistance of lung adenocarcinomas to gefitinib or erlotinib is associated epidermal growth factor receptor mutations with complex patterns in non-small with a second mutation in the EGFR kinase domain. PLoS Med. 2005;2:e73. cell lung cancers related to gefitinib responsiveness in Taiwan. Clin Cancer Res. 28. Maseki S, Ijichi K, Tanaka H, Fujii M, Hasegawa Y, Ogawa T, et al. Acquisition 2004;10:8195–203. of EMT phenotype in the gefitinib-resistant cells of a head and neck 5. Lee CK, Brown C, Gralla RJ, Hirsh V, Thongprasert S, Tsai CM, et al. Impact of squamous cell carcinoma cell line through Akt/GSK-3β/snail signaling EGFR inhibitor in non-small cell lung cancer on progression-free and overall pathway. Br J Cancer. 2012;106:1196–204. survival: a meta-analysis. J Natl Cancer Inst. 2013;105:595–605. 29. Levy BP, Rao P, Becker DJ, Becker K. Attacking a moving target: understanding 6. Chen JT, Lin TS, Chow KC, Huang HH, Chiou SH, Chiang SF, et al. Cigarette resistance and managing progression in EGFR-positive lung cancer patients smoking induces overexpression of HGF in type II pneumocytes and lung treated with tyrosine kinase inhibitors. Oncology (Williston Park). 2016; cancer cells. Am J Respir Cell Mol Biol. 2006;34:264–73. 7. Hsu NY, Ho HC, Chow KC, Lin TY, Shih CS, Wang LS, et al: Overexpression of 30. Yin J, Lee JH, Zhang J, Gao Z, Polotsky VY, Ye J. Regulation of hepatocyte dihydrodiol dehydrogenase as a prognostic marker of non-small cell lung growth factor expression by NF-κB and PPARγ in adipose tissue. cancer. Cancer Res 2001;61:2727-2731. Am J Physiol Endocrinol Metab. 2014;306:E929–36. 8. Chiang YY, Chen SL, Hsiao YT, Huang CH, Lin TY, Chiang IP, et al. Nuclear 31. Nagai T, Arao T, Furuta K, Sakai K, Kudo K, Kaneda H, et al. Sorafenib inhibits expression of dynamin-related protein 1 in lung adenocarcinomas. Mod the hepatocyte growth factor-mediated epithelial mesenchymal transition Pathol. 2009;22:1139–50. in hepatocellular carcinoma. Mol Cancer Ther. 2011;10:169–77. Joseph et al. Journal of Hematology & Oncology (2018) 11:74 Page 14 of 14 32. Shen J, Xia W, Khotskaya YB, Huo L, Nakanishi K, Lim SO, et al. EGFR modulates microRNA maturation in response to hypoxia through phosphorylation of AGO2. Nature. 2013;497:383–7. 33. Song SJ, Poliseno L, Song MS, Ala U, Webster K, Ng C, et al. MicroRNA- antagonism regulates breast cancer stemness and metastasis via TET-family- dependent chromatin remodeling. Cell. 2013;154:311–24. 34. Sudhakar JN, Chow KC. Human RAD23 homolog A is required for the nuclear translocation of apoptosis-inducing factor during induction of cell death. Biol Cell. 2014;106:359–76. 35. Mountain CF. Revisions in the international system for staging lung cancer. Chest. 1997;111:1710–7. 36. Smith JS, Alderete B, Minn Y, Borell TJ, Perry A, Mohapatra G, et al. Localization of common deletion regions on 1p and 19q in human glioma and their association with histological subtype. Oncogene. 1999;18:4144–52. 37. Ishiyama M, Tominaga H, Shiga M, Sasamoto K, Ohkura Y, Ueno K. A combined assay of cell viability and in vitro cytotoxicity with a highly water-soluble tetrazolium salt, neutral red and crystal violet. Biol Pharm Bull. 1996;19:1518–20. 38. Remmele W, Schicketanz KH. Immunohistochemical determination of estrogen and progesterone receptor content in human breast cancer. Computer-assisted image analysis (QIC score) vs. subjective grading (IRS). Pathol Res Pract. 1993;189:862–6. 39. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457–81. 40. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep. 1966;50:163–70. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries � Our selector tool helps you to find the most relevant journal � We provide round the clock customer support � Convenient online submission � Thorough peer review � Inclusion in PubMed and all major indexing services � Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit

Journal

Journal of Hematology & OncologySpringer Journals

Published: May 31, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off