Etk Interaction with PFKFB4 Modulates Chemoresistance of Small-cell Lung Cancer by Regulating Autophagy

Lung cancer is the leading cause of cancer-related death worldwide, with millions of new cases diagnosed annually (1). Small-cell lung cancer (SCLC) is a highly aggressive malignancy that accounts for approximately 15%–18% of all lung cancers (2). SCLC is characterized by a rapid doubling time, the early development of metastases, and a subsequently poor prognosis (3). A platinum–etoposide combination regimen remains the first treatment option for SCLC, and majority of patients exhibit sensitivity to this initial regimen. However, most SCLC patients rapidly develop multidrug resistance to chemotherapy (4, 5). Therefore, it is critical to elucidate the molecular mechanisms of chemoresistance and to identify effective therapeutic targets for SCLC.

Epithelial and endothelial tyrosine kinase (Etk), also known as bone marrow tyrosine kinase (Bmx), is a member of the Btk (Bruton tyrosine kinase) family of kinases. The Btk family is characterized by four conserved structural motifs: a pleckstrin homology (PH) domain, Src homology 3 (SH3) and SH2 domains, and a catalytic tyrosine kinase domain (6). Specifically, Etk is expressed in epithelial and endothelial cells (including cancer cells) rather than just in lympho-hematopoietic cells where Btk is expressed (7). Several studies have suggested that Etk plays significant modulatory roles in tumorigenicity, apoptosis and treatment resistance of many human malignancies, including prostate cancer, breast cancer, renal cell carcinoma, etc. (8–16). For example, Etk interacts with p53 in the cytoplasm to protect tumor cells from apoptosis and confers resistance to doxorubicin in prostate cancer (10). In breast cancer, Etk promotes tumorigenic growth and resistance to antiestrogen therapies by phosphorylating Pak1 (P21-activated kinase 1; refs. 9, 16). Our previous study identified that Etk inhibits apoptosis and functions as a chemoresistance-associated protein by mediating either Bcl-2 or Bcl-X_L in SCLC (17, 18). Another study showed that CTN06, a dual inhibitor of Btk and Etk, induced autophagy and exhibited increased sensitivity to docetaxel in prostate cancer cells, which implied Etk might affect chemoresistance by modulating autophagy (19). Autophagy is a natural physiologic process for cellular homeostasis to disassemble unnecessary or dysfunctional cellular components, which plays a dual role in cancer (20), and accumulating researches suggested that autophagy facilitates resistance to chemotherapy in many human tumors (21–24). Etk was identified as an important chemoresistance-associated protein in SCLC (17, 18), but the molecular mechanisms of Etk leading to the poor therapeutic response remains obscure.

6-Phosphofructo-2-kinase/fructose-2,6-biphosphatase 4 (PFKFB4) belongs to a family of bifunctional PFKFB enzymes, which regulates glycolysis by controlling fructose-2,6-bisphosphate (F2,6-BP) levels (25, 26). Overexpression of PFKFB4 was observed in many human cancers (27–29), with the majority of studies focusing on glycolytic flux modulation by PFKFB4 during hypoxia (30–35). Strohecker and colleagues recently identified PFKFB4 as a novel autophagy regulator by a high-throughput shRNA screen after deleting the autophagy cargo receptor and substrate p62/SQSTM1 (36). Knockdown of PFKFB4 increased autophagic flux (36). This report revealed the function of PFKFB4 on autophagy, which implied us that PFKFB4 might modulate autophagy in SCLC.

As the reports that implied a relationship between Etk, chemoresistance, and autophagy, we hypothesized that Etk may contribute to chemoresistance in SCLC by modulating autophagy, which is validated in the following tests, and inhibition of autophagy sensitized the SCLC cells to chemotherapy. Furthermore, we performed a microarray analysis to identify downstream effectors of Etk that may affect autophagy. After screening, we identified PFKFB4 expression was associated with Etk and its contribution to autophagy in SCLC. Besides, we validated the direct interaction of PFKFB4 and Etk in protein level. Then, we investigated the biological function of PFKFB4 on the chemotherapy response both in vivo and in vitro, and its expression associated with various clinicopathologic characteristics. Taken together, our results identify a novel mechanism of Etk in promoting chemoresistance through regulating autophagy, suggesting the potential application of PFKFB4 expression in predicting drug resistance and in customizing optimal treatments of SCLC.

A total of 68 formalin-fixed, paraffin-embedded (FFPE) tissues and 43 blood samples were collected from patients who had received bronchofiberscopy or biopsy for SCLC between the January 2013 and October 2015 and received care and follow-up in Zhujiang Hospital (Southern Medical University, Guangzhou, China) or the First Affiliated Hospital of Guangzhou Medical University (Guangzhou, China). Informed consent was obtained from all patients before specimen collection. The experiments were approved by the Ethics Committee of the Southern Medical University (Guangzhou, China). All samples were independently reviewed by at least two pathologists and were divided as limited disease or extensive disease according to the Veterans Administration Lung Study Group. Samples were further distinguished as “sensitive” (complete response or partial response) and “refractory” (stable disease or progressive disease) groups according to Response Evaluation Criteria in Solid Tumors (RECIST Edition 1.1).

Human SCLC cell lines NCI-H69, NCI-H446, and the chemoresistant H69AR were purchased from ATCC. Human SCLC cell lines NCI-H82, H209, H345, H146, and H526 were obtained as a generous gift from Dr. Ji Lin of MD Anderson Cancer Center (Houston, TX). The other drug-resistant subline, H446DDP, was established in our laboratory by culturing H446 cells in cisplatin and is described in detail in the Supplementary Materials and Methods section.

Cells were transiently transfected with validated siRNAs for Etk and PFKFB4 and corresponding negative control (GenePharma) and the lentiviral particles of shEtk, Etk-GFP, or shPFKFB4, PFKFB4-GFP (GenePharma). Detailed descriptions of the experiment can be found in the Supplementary Materials and Methods section.

RNA isolation, real-time qRT-PCR, Western blot analysis, flow cytometry, and Cell counting kit-8 (CCK-8) assay were performed as described previously (37), and as detailed in the Supplementary Materials and Methods section.

Ibrutinib (PCI-32765) was purchased from Selleck Biochemicals as a powder. For Western blot analysis, ibrutinib was dissolved in DMSO and diluted to 1, 2, or 5 μmol/L in culture media. For animal experiments, ibrutinib powder was dissolved in a 10% HP-beta-cyclodextrin solution (1.6 mg/mL) and administered to mice in the drinking water.

Tissues of SCLC clinical samples were fixed in 4% paraformaldehyde, and then embedded in paraffin blocks. Four-micron-thick sections were cut and analyzed for PFKFB4 (Abcam) or Etk (Abcam) protein expression. The slides were incubated in PFKFB4 antibody or Etk antibody diluted to 1:100 at 4°C overnight and then with the secondary antibodies for 2 hours. The sections were viewed under EnVision peroxidase system (Dako) and then analyzed with Image-Pro Plus analysis software.

cDNA expression microarray was performed as described previously (38), and as detailed in the Supplementary Materials and Methods section.

Total of 1 mg whole cell was incubated overnight with 2 μg of antibodies against Etk (Abcam). Then, 100 μL preblocked agarose beads was added to the antibody/lysate mixture and incubated for another 2 hours at 4°C. After incubation, immunocomplexes were washed for three times in HEGNDT buffer [10 mmol/L HEPES (pH 8.0), 1 mmol/L EDTA, 10% glycerol, 50 mmol/L NaCl, 2 mmol/L dithiothreitol (DTT)] containing 0.1% Triton X-100 and once in HEGNDT buffer without Triton X-100. The protein complex was finally eluted off the beads for Western blot analysis, and the solubilized proteins were subjected to SDS-PAGE, transferred onto a polyvinylidene difluoride membrane, and analyzed by Western blotting using the antibody of PFKFB4.

Etk cDNA was isolated by RT-PCR and cloned into the XhoI and EcoRI sites of GST-tagged pGEX-4T-1 vector by T4 DNA Ligase (Thermo Scientific), whereas PFKFB4 cDNA was cloned into the XhoI and EcoRI sites of His-tagged pET-28a(+)vector. The pGEX-4T-1-Etk and pET-28a(+)-PFKFB4 recombinant plasmids were transformed into E. coli BL21(DE3) separately. PCR identification, double enzyme digesting, and sequencing were used to screen and identify the high expressing positive clone, which synthesized GST-Etk or His-PFKFB4 recombinant proteins. The purified GST-Etk fusion proteins were attached to Glutathione Sepharose (GE Healthcare), and then mixed and incubated with the purified His-PFKFB 4 protein at 4°C overnight. The eluted samples were detected by Western blot analysis using the GST antibody (TransGen) and His antibody (CWBio).

The adenoviral particles of tandem mRFP-GFP-LC3 were purchased from Hanbio. SCLC cells (approximately 1 CL 10³) were seeded into glass bottom cell culture dish (NEST). Twenty-four hours after mRFP-GFP-LC3 adenovirus infection, cells were transfected with siRNA targeting PFKFB4, Etk, or the corresponding negative control (siNC). The next day, cells were fixed with formaldehyde for 20 minutes. After being rinsed in PBS, the nuclei were stained with DAPI (Beyotime) for 5 minutes. Specially, suspension cell line H69 was analyzed directly without treated with formaldehyde and DAPI. The cells were then analyzed using LSM880 laser scanning confocal microscope (Zeiss). The number of green or red punctate fluorescence in each cell was counted in 10 different fields under the confocal laser scanning microscope.

Cells were fixed with 2.5% glutaraldehyde containing 0.1 mol/L sodium cacodylate, and treated with 1% osmium tetroxide. After dehydration, samples were embedded in araldite and then cut into thin sections that were stained with uranyl acetate and lead citrate. Digital images obtained using Philips CM-120 Transmission Electron Microscopy at 60 kV.

Tumor xenograft experiments were performed as described previously (37), and as detailed in the Supplementary Materials and Methods section.

Fresh tissues from a chemo-naïve SCLC patient undergoing surgical treatment were obtained and transported immediately to the animal facility in PBS at 4°C. Primary tissue sample was anonymized and obtained in accordance with the Zhujiang Hospital of Southern Medical University (Guangzhou, China) Institutional Review Board. Primary SCLC tissue specimens were minced with scissors into small (2–3 mm³) fragments. Under aseptic conditions, tissue fragments were implanted subcutaneously into the flanks of female B-NSG mouse (NOD-Prkdc^scid Il2rg^tm1/Bcgen, BIOCYTOGEN) by using a 10-gauge Trochar needle through a small incision on the animal's dorsal flank. Animal health was monitored daily. Once established, solid tumor xenografts were serially passaged using the same technique.

Data are represented as mean ± SD of at least three independent experiments. Independent samples t-test or one-way ANOVA was employed to analyze the possible differences between groups. The association between PFKFB4 and Etk expression was analyzed by Spearman rank correlation test. The association between PFKFB4, Etk expression, and clinicopathologic characteristics was explored by χ² test. Survival curves were estimated using the Kaplan–Meier analysis. Prognostic factors were evaluated by univariate and multivariate analyses (Cox proportional hazards model). Differences with P values of less than 0.05 were considered statistically significant. All statistical analyses were performed using SPSS 17.0 software.

To test our hypothesis that Etk regulates the chemoresistance of SCLC by modulating autophagy, we first examined whether chemoresistant SCLC cells exhibited increased autophagy. Autophagosomes in SCLC were evaluated using transmission electron microscopy (TEM) and confocal laser scanning microscopy (CLSM), as well as Western blotting to detect the conversion of LC3-I to active LC3-II protein (i.e., the LC3-II/LC3-I ratio) and P62 protein levels, two hallmarks of autophagy. As shown in Fig. 1A, TEM detected an increased number of autophagosomes in H69AR, H446DDP, and H82 cells than that in H69, H446, and H209. As shown in Fig. 1B, Western blotting demonstrated that compared with the drug-sensitive cells, H69AR, H446DDP, and H82 cells had an increased LC3-II/LC3-I ratio and a decreased P62 protein level, suggesting that the chemoresistant SCLC cells exhibited increased autophagy. Moreover, we expressed in the SCLC cells an mRFP-GFP-LC3 fusion protein using adenovirus vectors.

CLSM analysis showed an increased mRFP-GFP-LC3 signal in H69AR, H446DDP, and H82 compared with their parental sensitive cells, indicating an increase in autophagy concurrent with the development of chemoresistance (Fig. 1C). Together, the above results suggest that chemoresistant SCLC cells exhibit increased autophagy.

Etk was identified as an important chemoresistance-associated protein in SCLC in our previous study, which showed that knockdown of Etk impaired resistance of H69AR cells to chemotherapeutic drugs (18). Study by Guo and colleagues showed that CTN06, a dual inhibitor of Btk and Etk, induced autophagy in prostate cancer cells (19), implicating a role of Etk in modulating autophagy. Besides two pairs of chemoresistant SCLC cell lines, the other five SCLC cell lines including H82, H209, H526, H146, and H345 were tested by CCK-8 assays to screen their relative chemoresistance. H82 was regarded as the comparative chemoresistant cells with highest IC₅₀ values and H209 as chemosensitive cells with lowest IC₅₀ values (Supplementary Fig. S1A).

To determine that Etk plays a function in control of autophagy, we knocked down its expression in SCLC cells using siRNA and then analyzed the effects on autophagic flux. As shown in Supplementary Fig. S1C, Etk mRNA and protein expression is significantly elevated in H69AR, H446DDP, H82 cells compared with H69, H446, and H209 cells. The efficiency of siRNA knockdown in H69AR and H446DDP cells was confirmed using qRT-PCR and Western blotting (Supplementary Fig. S1D and S1E). Western blotting analysis and CLSM revealed that knockdown of Etk inhibited autophagy in drug-resistant SCLC cells (Fig. 1D and E).

To further examine the contribution of autophagy to the chemoresistance of SCLC, chloroquine, which inhibits autophagic cargo degradation, was used to block autophagy. Chloroquine was added to cells for 24 hours at a final concentration of 20 μmol/L. Western blotting revealed an increased LC3II/LC3I ratio in H69AR and H446DDP cells after chloroquine treatment (Fig. 1F). Moreover, the CCK-8 assays revealed that chloroquine significantly inhibited drug resistance of H69AR and H446DDP cells (Fig. 1G). Taken together, these results demonstrate that chloroquine and siRNAs targeting Etk could inhibit autophagy and enhance the sensitivity of resistant SCLC cells to chemotherapeutic agents.

To explore the molecular targets associated with Etk, a cDNA microarray analysis was conducted using H69AR cells transfected with siRNA targeting Etk. A total of 25 genes were downregulated, and 31 genes were upregulated by more than 1.5-fold in si-Etk–treated cells (Fig. 2A). Among those genes, PFKFB4 caught our attention because of its association with the tumor growth in lung cancer, prostate cancer, breast cancer, etc. (34, 35).

Consistent with the microarray data, knockdown of Etk indeed inhibited PFKFB4 expression at both the mRNA and protein level in H69AR, H446DDP, and H82 cells, whereas restoration of Etk led to the opposite changes in the chemosensitive cells (Fig. 2B and C; Supplementary Fig. S2A). The above results indicate that Etk signaling stimulates the expression of PFKFB4.

To further explore the function of Etk in control of PFKFB4, we examined whether the two proteins associate with each other. Indeed, our coimmunoprecipitation (co-IP) assay detected a significant amount of PFKFB4 protein being associated with Etk (Fig. 2D).

To evaluate whether these two proteins directly interacted with each other, we performed glutathione-S-transferase (GST)-pull down assay with purified GST-tagged Etk and His-tagged PFKFB4 Results from the pull-down assay showed that purified His-PFKFB4 protein was associated with GST-Etk, but not GST (Fig. 2E). Together, the results indicate that Etk directly interacts with PFKFB4 and that their expression is correlated in SCLC.

A recent study identified PFKFB4 as a novel autophagy regulator by a high-throughput shRNA screening with murine immortalized epithelial cells (36). To examine whether PFKFB4 regulates autophagy in SCLC cells, we knocked down its expression in SCLC cells and then analyzed the autophagic flux. The efficiency of siRNA knockdown in H69AR, H446DDP, and H82 cells was confirmed using qRT-PCR (Supplementary Fig. S3A and S3B). TEM revealed that the number of autophagic vesicles was decreased in H69AR and H446DDP cells transfected with siRNAs targeting PFKFB4 (Supplementary Fig. S3C). Similarly, Western blotting and CLSM analysis revealed that knockdown of PFKFB4 inhibited autophagy in drug-resistant SCLC cells (Fig. 2F and G), indicating that indeed in SCLC cells, PFKFB4 regulates autophagy.

To determine whether Etk controls autophagy through PFKFB4, we performed rescue experiments. Results obtained from the experiments demonstrated Etk overexpression increased autophagy; however, the increased authophagy was diminished by knockdown of PFKFB4 (Fig. 2H and I).

Next, to investigate whether PFKFB4 plays a role in chemoresistance in SCLC, we first analyzed PFKFB4 expression in chemoresistant and chemosensitive SCLC cell lines. Intriguingly, chemoresistant H69AR, H446DDP, and H82 cells expressed significantly higher levels of PFKFB4 than the chemosensitive H69, H446, and H209 cells, at both mRNA and protein level (Supplementary Fig. S4A). In order to assess whether PFKFB4 was functionally involved in the chemoresistance of SCLC, we altered PFKFB4 expression in the cells by shRNA (shPFKFB4-1, shPFKFB4-2)–mediated stable knockdown in H69AR, H446DDP, and H82 cell lines or PFKFB4 overexpression (LV5-PFKFB4) in chemosensitive SCLC cells (Supplementary Fig. S4B and S4C). Furthermore, we were able to rescue the shRNA knockdown of PFKFB4 by ectopic expression of PFKFB4 cDNA that was insensitive to shPFKFB4 (Supplementary Fig. S4F). We then examined the effects of PFKFB4 expression alteration on the chemoresistance in SCLC cells. CCK-8 assays were performed to measure the sensitivity of SCLC cells to various chemotherapeutic drugs (i.e., adriamycin, ADM; cisplatin, CDDP; etoposide, VP-16). The results revealed that the IC₅₀ values (i.e., the concentration at which growth is inhibited by 50%) were increased in chemosensitive SCLC cells with PFKFB4 overexpression (Fig. 3A), whereas the IC₅₀ values of H69AR, H446DDP, and H82 cells were significantly decreased after knockdown of PFKFB4 (Fig. 3B; Supplementary Fig. S4G).

To further confirm that PFKFB4 is a major downstream effector of Etk in SCLC chemoresistance, CCK-8 assays were used to evaluate the effects of knockdown of PFKFB4 in Etk-overexpressing cells on chemo-drug sensitivity. The results showed that, compared with the empty vector controls, cells with Etk overexpression exhibited increased IC₅₀ values for the chemotherapeutic drugs tested (Fig. 3C), and knockdown of PFKFB4 in H69 cells could rescue the increase of the IC₅₀ values mediated by Etk upregulation (Fig. 3C; Supplementary Fig. S2D and S2E).

Moreover, to determine whether PFKFB4 confers chemoresistance in vivo, we subcutaneously transplanted H446 and H446DDP cells with altered PFKFB4 expression into nude mice. PFKFB4 expression in the tumor xenografts was measured by IHC (Supplementary Fig. S4D and S4E). PFKFB4 knockdown significantly reduced the tumor volumes after chemotherapy treatment (CDDP and VP-16); in contrast, the tumor volumes were obviously larger in the PFKFB4 overexpression group than those in the corresponding control groups (Fig. 3D and E). These results suggested that PFKFB4 could affect the chemosensitivity of SCLC cells in vivo.

We next evaluated the effect of PFKFB4 on apoptotic cell death upon cell exposure to chemotherapeutic drugs. Flow cytometry analysis of Annexin V demonstrated that PFKFB4 knockdown in H69AR and H446DDP cells led to increased apoptosis after treatment with ADM, CDDP, or VP-16, whereas PFKFB4 overexpression in H69 and H446 cells decreased apoptosis (Fig. 3F and G; Supplementary Fig. S4H and S4I).

To determine the clinicopathologic significance of PFKFB4 expression in SCLC and clinical correlation between Etk and PFKFB4, 68 cancer specimens and 10 normal lung tissues were analyzed using IHC. The intensity of PFKFB4 staining in the chemoresistant SCLC specimens was much stronger than that in the drug-sensitive group, and weak expression was detected in normal lung alveolar cells (Fig. 4A). In addition, positive PFKFB4 expression was detected in 26 of the 32 (81.25%) chemoresistant SCLC samples, compared with 20 of the 36 (55.56%) drug-sensitive SCLC tissues and 2 of the 10 normal alveolar epithelial cells (20%; Fig. 4B). Similarly, elevated Etk expression was detected in the drug-sensitive group compared with the drug-resistant group (Fig. 4C). The rate of Etk positivity was much higher in anticancer drug–refractory SCLC specimens (87.5%) than in drug-sensitive specimens (52.78%) and normal bronchial epithelium (10%; Fig. 4D). As shown in Supplementary Table S1, compared with the drug-sensitive group, the chemoresistant group had a significantly higher frequency of positive PFKFB4 expression (P = 0.006), which was also higher in the extensive stages group, compared with the earlier stages one (P = 0.01).

However, there were no significant differences with respect to gender or age (≤56 years and >56 years; Supplementary Table S1). In terms of Etk expression in SCLC specimens, the higher Etk expression was correlated with a poor chemotherapy response and extensive disease stage (Supplementary Table S1). Furthermore, we compared Etk and PFKFB4 expression using IHC in SCLC tissues prior to initial treatment and found that PFKFB4 levels were positively correlated to Etk expression (Fig. 4E), which was in agreement with their expression in peripheral blood using qRT-PCR (Supplementary Fig. S2B and S2C). Kaplan–Meier survival analysis demonstrated high levels of PFKFB4 expression were correlated with a shorter overall survival in SCLC patients (Fig. 4F). The multivariate analysis revealed that PFKFB4 was an independent prognostic factor (P = 0.004; Supplementary Table S2). In addition, we tested circulating PFKFB4 levels in peripheral blood, which was drawn from SCLC patients prior to initiating therapy. The result indicated that PFKFB4 expression was higher in the drug-resistant group (n = 19) than in the drug-sensitive group (n = 24; Fig. 4G). Collectively, these results indicate that PFKFB4 overexpression was correlated with poor survival, worse response to chemotherapy, and later stages in SCLC.

Ibrutinib (PCI-32765), an irreversible phosphorylation inhibitor of Btk, inactivates the B-cell antigen receptor (BCR) signaling pathway, which is vital for the survival of malignant B cells (39). Oral administration of ibrutinib is highly effective in controlling mantle cell lymphoma and certain types of chronic lymphocytic leukemia (40). As a member of a non-receptor tyrosine kinase family, Etk shares a similar catalytic tyrosine kinase domain with Btk and is partially inhibited by ibrutinib. In light of the high clinical efficacy of ibrutinib in treating B-cell malignancy, we hypothesized that ibrutinib may affect the chemosensitivity in SCLC cells by suppressing Etk activation. Because ibrutinib inhibits phosphorylation of the target, we first evaluated expression levels of Etk and tyrosine-phosphorylated Etk (pEtk) in SCLC cells (Supplementary Fig. S1B; Fig. 5A). Using H69AR (which express higher levels of Etk and pEtk) and H69 cells as models, we added ibrutinib at doses ranging from 0.1 μmol/L to 5 μmol/L for 24 hours. The Western blotting analysis revealed that ibrutinib modestly inhibited Etk expression and that 1 μmol/L ibrutinib significantly decreased pEtk expression in H69AR cells (Fig. 5B left), whereas no significant changes were observed in H69 cells (Fig. 5B right). The CCK-8 assays revealed that the IC₅₀ values of the chemotherapeutic drugs were significantly decreased in H69AR cells treated with 1 μmol/L ibrutinib (Fig. 5C left) However, no significant difference of the IC₅₀ values was observed in H69 cells treated with ibrutinib (Fig. 5C right).

In addition, to determine whether ibrutinib modulates chemoresistance in vivo, we established and used a patient-derived xenograft (PDX) model with tumor tissues from a SCLC patient and also a subcutaneous xenograft models of H69 and H69AR. The tumor-carrying mice were administered either ibrutinib (25 mg/kg daily) in their drinking water, or chemotherapeutic drugs (CDDP+VP-16) by intraperitoneal injection, or the combination. The clinicopathologic characteristics of the patient that donated SCLC specimen to perform PDX experiment is listed in Supplementary Table S3. The results showed a smaller tumor volume in the ibrutinib or chemotherapeutic-treated mice, but combination of the two treatments reduced the tumor volume significantly in the H69AR tumor model (Fig. 5D and E, left). Interestingly, whereas no significant changes were observed in the H69 tumors (Fig. 5D and E, right), IHC analysis revealed significantly reduced expression of Etk and pEtk in the xenografts of ibrutinib-treated group (Fig. 5F). Significantly, the PDX tumors displayed similar results. As shown in Fig. 5G and H, combined treatment of ibrutinib and chemotherapy significantly decreased the tumor volume. Expression of Etk and pEtk was inhibited with ibrutinib administration (Fig. 5J). IHC analysis showed similar expression of Syn and CD56 and histomorphology between the SCLC patient specimen and the corresponding PDX xenograft tumors (Fig. 5I). Collectively, these results suggested a potentially synergistic effect of ibrutinib with chemotherapy in SCLC tumors.

To explore the mechanism of action of ibrutinib, we performed rescue experiments to determine whether ibrutinib works by inhibiting Etk or also Btk. As shown in Fig. 6A and B, the expression of Btk, especially phosphor-Btk (pBtk), is very low in the SCLC cells examined. Importantly, ibrutinib treatment at doses ranging from 0.1 μmol/L to 5 μmol/L had no significant effect on Btk and pBtk expression. In contrast, ibrutinib effectively inhibited Etk and Y566-phosphor-Etk (pEtk) expression in the SCLC cells, whereas restoration of Etk rescues the effect of ibrutinib (Fig. 6C). Phosphorylation of Etk at Y566 has been characterized as a hallmark of Etk activation. Therefore, the selective inhibition by ibrutinib of Etk phosphorylation at Y566, but not Btk, in SCLC cells strongly suggests that ibrutinib targets Etk function in SCLC cells.

To further demonstrate that the cell growth inhibition effect by ibrutinib is mediated by targeting Etk, we next performed CCK-8 assays. As shown in Fig. 6D, ibrutinib strongly sensitized H69AR cells to the three different chemotherapy drugs (i.e., ADM, CDDP, and VP-16) indicated by the drug IC₅₀ decreases in cells treated with ibrutinib (1 μmol/L). Remarkably, Etk ectopic expression could effectively diminish the IC₅₀ reduction to the three chemo-drugs. Taking together, the results suggest that ibrutinib suppresses chemoresistance by inhibiting Etk activation in SCLC.

Etk was found to be an important chemoresistance-associated protein by mediating Bcl-X_L and inhibited apoptosis via Bcl-2 of SCLC in our previous study (17, 18). However, the molecular mechanisms underlying chemoresistance mediated by Etk in SCLC remained elusive. In this study, we demonstrated for the first time that Etk interacts with PFKFB4 to promote the chemoresistance of SCLC by modulating autophagy (Fig. 6E).

To further understand the biological function of Etk in chemoresistance of SCLC, we examined the literatures. A study showed that CTN06, a dual inhibitor of Btk and Etk, promoted autophagy in prostate cancer, which also exhibited increased sensitivity to docetaxel and implied a negative correlation between Etk and autophagy (19). Thus, we hypothesized that Etk modulated chemoresistance of SCLC by autophagy. To confirm this, we compared the autophagic flux in SCLC cells and found that chemoresistant cells exhibited increased autophagy and knockdown of Etk inhibited autophagy. Then we performed a microarray analysis to screen molecules downstream of Etk and figured out that PFKFB4, as a downstream signal, might modulate chemoresistance of SCLC by regulating autophagy.

Majority of studies about PFKFB4 focus on glycolytic flux modulation (30–35). Recently, Strohecker and colleagues identified PFKFB4 as a novel autophagy regulator by a high-throughput shRNA screen after deleting the autophagy cargo receptor and substrate p62/SQSTM1; PFKFB4 was initially found as an autophagy stimulator in this article; however, the following tests showed that knockdown of PFKFB4 increased autophagic flux by suppressing the accumulation of reactive oxygen species (ROS) and p62 (36). In SCLC, we identified that knockdown Etk or PFKFB4 significantly reduced autophagic flux, which suggests that inhibiting autophagy by targeting either Etk or PFKFB4 may be a potential strategy in attenuating the chemoresistance of SCLC patients. Although it is difficult to reconcile the results of our study with those of the two discordant studies described above, we believe that the combined use of TEM, CLSM, and Western blotting to monitor autophagy, as well as the application of siRNAs to downregulate Etk and PRKRB4 in chemoresistant SCLC cell lines, provide adequate data to support our claims that downregulation of Etk or PFKFB4 could significantly inhibit autophagy in SCLC. In addition, the studies mentioned above that used prostate cancer cells and non–small cell lung cancer cells as the experiment models (19); thus, the discordant results suggest that in SCLC, PFKFB4 may modulate autophagy by mechanisms other than suppressing either ROS or p62 accumulation. Hence, it is critical to elucidate the molecule mechanisms by which PFKFB4 contributes to autophagy and present effective regimens to inhibit autophagy for improving chemosensitivity in SCLC.

Furthermore, to fully understand the relation between Etk and PFKFB4 in the mRNA and protein level and to find effective inhibitors targeting Etk and PFKFB4, additional studies are required to identify the specific binding domains of each protein and measure the kinase activity of these two enzymes. Our data indicated that PFKFB4 modulated chemoresistance both in vivo and in vitro, and correlated with poor survival and a weak chemotherapy response in SCLC. which is consistent with other reports that noted PFKFB4 overexpression promotes cancer cell survival in other cancer types (27–29).

Ibrutinib (PCI-32765) is highly effective in controlling some types of B-cell malignancies by inhibiting Btk. On the basis of the effective application of ibrutinib for leukemia patients, an idea arose that ibrutinib may help attenuate chemoresistance in SCLC patients because Etk shares similar domains with Btk and Etk is partially inhibited by ibrutinib. Our study revealed that ibrutinib modestly inhibited Etk expression and that an appropriate concentration of ibrutinib significantly decreased pEtk expression in vivo and in vitro. Ibrutinib administration did not significantly affect tumor growth in a nude mouse model, which is consistent with the high IC₅₀ values of ibrutinib observed in SCLC cells (data not shown). However, the combination therapy of ibrutinib and anticancer drugs significantly reduced the tumor volumes, which suggests a synergistic effect of ibrutinib to chemotherapy in SCLC cells. This study implied the potential use of ibrutinib in SCLC. To further investigate molecular mechanisms about how ibrutinib enhances the effect of antineoplastic drugs in killing SCLC cells, additional tests are required.

Lack of preclinical tools for predicting clinical activity of novel therapeutic strategies in cancer restricts the development of progress in oncology. More and more evidence shows that PDX has been a promising powerful tool to overcome this shortcoming (5, 41–44). To further support that ibrutinib restores chemosensitivity of SCLC, we used a PDX model and showed the similar results as SCLC cell line xenografts experiments, which suggest the potential clinical application of ibrutinib in SCLC.

In summary, our results demonstrate for the first time that Etk promote the chemoresistance of SCLC by interacting with PFKFB4. Furthermore, we identified that knocking down either Etk or PFKFB4 significantly reduced autophagic flux while increasing chemosensitivity in SCLC. Our study implied that ibrutinib may synergistically promote chemosensitivity in SCLC cells. Therefore, we conclude that Etk and PFKFB4 are promising prognostic biomarkers of the chemotherapy response and that inhibiting autophagy by targeting PFKFB4 may be a potential strategy for treating patients with drug-resistant SCLC.

No potential conflicts of interest were disclosed.

Conception and design: Q. Wang, F. Zeng, L. Guo

Development of methodology: Q. Wang, Y. Sun

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): F. Zeng, Y. Sun, W. Huang

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Q. Wang, F. Zeng, Q. Qiu, X. Huang

Writing, review, and/or revision of the manuscript: Q. Wang, F. Zeng, J. Zhang, J. Huang, L. Guo

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Q. Wang, F. Zeng, W. Huang

Study supervision: L. Guo

The SCLC Human SCLC cell lines NCI-H82, H209, H345, H146 and H526 were obtained as a generous gift from Dr. Ji Lin of MD Anderson Cancer Center. The flow cytometry analysis was supported by Guangdong Provincial Key Laboratory of Malignant Tumor Epigenitic and Gene Regulation, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University. This work was supported by grants from the National Natural Science Foundation of China (nos. 81172241 and 81572244), the Clinical Research Initiative Project of Southern Medical University (LC2016ZD029), and Guangdong Natural Science Foundation (Special fund for Scientific and Technological Development; 2017A030313644, 2016A030313822).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.