The Transcription Factor AP4 Promotes Oncogenic Phenotypes and Cisplatin Resistance by Regulating LAPTM4B Expression

Breast cancer is the most frequently diagnosed malignancy in women worldwide and the second leading cause of cancer-related mortality among females worldwide (1). Despite early detection and enhanced management, breast cancer is expected to account for 15% of total new cancers in China (2). Identification of novel prognostic markers and a thorough understanding of the marker's mechanism will provide a basis for designing therapeutic strategies for breast cancer patients.

LAPTM4B is a newly identified oncogene (NM_018407, Gene ID: 55353) and was first cloned in human hepatocellular carcinoma (HCC) in 2000. LAPTM4B protein has been found to be widely expressed in normal human tissues and upregulated in various types of carcinomas (3). It could promote the proliferation of tumor cells, boost invasion and metastasis, resist apoptosis, initiate autophagy, and assist drug resistance (4). In addition, overexpression of LAPTM4B was significantly correlated with poor prognosis in breast cancer (5), gallbladder cancer (6, 7), ovarian cancer (8), HCC (9–11), gastric cancer (12), and cervical cancer (13) etc. It is noteworthy that the LAPTM4B gene has two alleles named LAPTM4B*1 and LAPTM4B*2 (GenBank numbers AY219176 and AY219177, respectively). Allele *1 contains only one copy of a 19-bp sequence at the 5'UTR in the first exon, whereas the sequence in allele *2 is duplicated and tandemly repeated (14). Previous study showed that LAPTM4B *2/2-type allele was significantly associated with the susceptibility of various adenocarcinoma including breast cancer (15–17), HCC (18, 19), non–small lung cancer (20), gastric cancer (21), cervical cancer (22), colorectal cancer (23), lymphoma (24), gallbladder carcinoma (25), ovarian carcinoma (26), and malignant melanoma (27). Therefore, we hypothesized that if there are some specific transcription factors binding to the LAPTM4B polymorphism region that play important roles in regulating LAPTM4B expression.

In this study, we analyzed the LAPTM4B polymorphism region in the first exon and found the transcription factor AP4 conserved binding sites in the 19-bp polymorphism region. The expression levels of AP4 and LAPTM4B and their prognostic significance were demonstrated in breast cancer. AP4 promoted breast cancer proliferation, migration, invasion, and cisplatin resistance partially by directly upregulating LAPTM4B expression. These effects were associated with cell cycle, EMT, and PI3K/AKT signaling pathways. The results provide strong evidence supporting the expression and functional coupling between AP4 and LAPTM4B and help to explore new therapeutic targets for breast cancer.

The 5′-genomic sequences of LAPTM4B of different species were gained from GenBank. Promoter prediction and analysis for potential transcription factors binding to the polymorphism region of LAPTM4B promoter were performed using UCSC and PROMO (TRANSFAC 8.3) website online programs (http://www.genome.ucsc.edu/ and http://alggen.lsi.upc.es/cgi-bin/promo_v3/promo/promoinit.cgi?dirDB=TF_8.3).

The TF profiling plate arrays (Signosis) were used to search for transcription factors binding to the promoter region of LAPTM4B*1 and LAPTM4B*2 alleles. The promoter probes for the two alleles were located in −200∼+301bp and −200∼+320bp, respectively. The nucleotide sequences are numbered with the transcription start site as +1.

Human breast cell lines MCF10A, MCF7, and ZR-75-1 were purchased from Cellcook (Guangzhou, China). MDA-MB-231 and T47D cell lines were kindly provided by Dr. Shou Cheng-Chao (Peking University Cancer Hospital & Institute, Beijing, China) and were authenticated by short tandem repeat analysis (Genetic Testing Biotechnology Corporation) in February 2017. The cells were cultured according to the ATCC recommended conditions. They were transfected with siRNA, shRNA, or corresponding empty vectors as described in Supplementary Table S1.

The tissue samples were obtained from 146 breast cancer patients who were surgically treated in the Beijing Cancer Hospital from January 2008 to December 2010. All the patients underwent primary mastectomies plus axillary lymph node dissection. None of the patients had received adjuvant chemotherapy, immunotherapy, or radiotherapy before. Survival information was available for all patients until March 4, 2017. This study was conducted in accordance with the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of Peking University School of Oncology. The informed consent was obtained from all participants.

The expressions of AP4 and LAPTM4B in breast cancer tissues were examined by IHC. The slides were incubated with anti-AP-4 polyclonal antibody (dilution 1:200, Abcam ab28512) and anti-LAPTM4B polyclonal antibody (dilution 1:200, Bioss Inc, bs-6542R) at 4°C overnight. Stained tissue sections were evaluated separately by two pathologists blinded to the clinical parameters. The percentage of positive-staining tumor cells was scored as follows: 0 (<5% positive tumor cells), 1 (5%–50% positive tumor cells), and 2 (>50% positive tumor cells). Staining intensity was scored as follows: 0 (no staining or only weak staining), 1 (moderate staining), and 2 (strong staining). The sum of the staining intensity score and the percentage score was used to define the AP-4 and LAPTM4B expression levels: 0–2, low expression; 3–4, high expression.

Stable AP4 (GV341, Ubi-MCS-SV40-puromycin) and LAPTM4B overexpression (GV358, Ubi-MCS-3FLAG-SV40-EGFP-IRES-puromycin) and control cell lines were created through lentivirus infection. Stable knockdown of AP4 was also established by infection of lentivirus particles (GV248, hU6-MCS-Ubiquitin-EGFP-IRES-puromycin) carrying AP4 shRNA and the target sequence of shAP4 was shown in Supplementary Table S1. They were both constructed by GeneChem. To generate the LAPTM4B gene promoter luciferase reporter plasmid, the truncated LAPTM4B gene upstream sequence was cloned into pGL3-Basic plasmid (Promega). The six truncated plasmids (pGL3-PF1–6) were conserved by our laboratory (28). AP4-binding site inside the LAPTM4B polymorphism region was site-directed mutated by SOE-PCR using modified reverse primer spanning the AP4-binding site (SBS Genetech Co., Ltd). All AP4 specific siRNAs were synthesized by RiboBio and the sequences are shown in Supplementary Table S1.

For the binding of AP4 to the LAPTM4B promoter, breast cancer cells were planted into 24-well plates in 1 × 10⁵ per well, and incubated for 12 hours in complete medium. Then MCF7, T47D, ZR-75-1, and MDA-MB-231 were cotransfected with 500 ng of LAPTM4B promoter plasmid and 20 ng phRL-CMV plasmid in triplicate. Transfection was carried out using Lipofectamine 3000 transfection reagent (Invitrogen) according to the manufacturer's instructions. At 48 hours after transfection, luciferase activity was measured using the dual-luciferase reporter assay system (Promega). The luciferase value of pGL3-promoter (positive control) was set to 100% and the other constructs were compared with it.

MCF7 and MDA-MB-231 cells and their derivatives were cultured as described above. For cross-linking, formaldehyde was added to a final concentration of 1%. The reaction was stopped by addition of glycine at a final concentration of 0.125 mol/L. Chromatin was sheared to an average fragment size of 200–1,000 bp by sonication. The sheared chromatin was then immunoprecipitated by using goat polyclonal AP-4 antibody (sc-18593X, Santa Cruz Biotechnology) or normal goat IgG (CR2, Sino Biological Inc). Identification of the captured LAPTM4B promoter fragments was performed by qPCR analysis by using the promoter primers. The sequence of primers was shown in Supplementary Table S1.

The annealed double-strand oligonucleotide derived from the LAPTM4B promoter containing either the wild-type (5′-GGA GCT CCA GCA GCT GGC TGG AGC C-3′) or mutated AP-4 binding site (5′-GGA GCT CCA Gtc aac tGC TGG AGC C-3′) was 3′ end-labeled by biotin. The AP-4-binding site was underlined and the mutated nucleotides were denoted in lowercase. Consensus wild-type and mutant AP-4 probes were synthesized by Viagene and nuclear extracts were prepared by Nuclear and Cytoplasmic Extraction reagents (Viagene). Five micrograms of nuclear protein was mixed for 30 minutes at room temperature with biotin-labeled oligonucleotide using LightShift Chemiluminescent EMSA Kit (Pierce) according to the manufacturer's instructions.

For competition experiments, 100- and 200-fold molar excess of unlabeled competitor probes were preincubated with nuclear extract for 30 minutes on ice before biotin-labeled probes were added. For the super-shift assay, 1 μg and 2 μg of goat anti-human AP-4 polyclonal antibody were added to the binding reaction. Normal goat IgG was the control.

Cells were washed twice with ice-cold PBS and lysed using RIPA extraction reagent (TIANGEN). Whole-cell lysates were boiled in lysis buffer containing 5% SDS. Protein concentration was determined by BCA protein assay kit (Pierce). Samples were subjected to SDS-PAGE and electrophoretically transferred to nitrocellulose membranes. The membranes were then incubated with primary antibody at 4°C overnight (antibodies were depicted in Supplementary Table S2) and secondary antibody at room temperature for 1 hour. Signals were detected using enhanced chemiluminescence reagents (Pierce). In all experiments, β-actin was used as an internal control.

Total RNA was isolated from cells using TRIzol (Invitrogen) according to manufacturer's instructions. Reverse transcription of 2 μg total RNA was carried out using oligo (dT16) primer. Quantitative real-time PCR was performed on a Roche 480 using SYBR Green Real-time PCR Master Mix (TIANGEN). The sequences of the PCR primers were shown in Supplementary Table S1. All reactions were run in triplicate.

The cell proliferation assay was performed with Cell Counting Kit-8 (CCK-8; Dojindo). Cells were seeded into 96-well plates at 1 × 10³ cells per well and then incubated for 5 days. Each time point, 10 μL CCK-8 solution was added to each well and incubated for 2 hours at 37°C. Then the results were measured with a microplate reader at 450 nm in accordance with the manufacturer's instructions. In colony formation assay, the cells were seeded on 60-mm dishes. After 2 weeks, the colonies were fixed in 4% formaldehyde and stained with 1% crystal violet. Finally, positive colony formation (>50 cells/colony) was counted. Each experiment was repeated in triplicate.

The cells were collected and fixed in 70% cold ethanol for at least 12 hours at 4 °C. Cells were stained with 50 μg/mL propidium iodide (BD Biosciences) at room temperature for 15 minutes in dark. Cell cycle was performed by FACSCalibur system (BD Biosciences) and analyzed by ModFit 3.0 software (BD Biosciences).

The migration and invasion of breast cancer cells were detected using 24-well transwell cell culture chambers (8.0-μm pore size, Costar). For invasion assay, the insert membranes were precoated with Matrigel (BD Biosciences). The cells were incubated at 37°C for 24 hours (migration assay) and 48 hours (invasion assay). The nonmigratory or noninvasive cells in the top chambers were removed with cotton swabs. The migrated and invaded cells on the bottom membrane surface were fixed in in 4% formaldehydel, air-dried, stained with 1% crystal violet, and then counted under a microscope.

MCF7 and MDA-MB-231 cells were incubated with different concentration of cisplatin and collected after 48-hour incubation. Apoptosis was analyzed using an PE-Annexin V Apoptosis Detection kit (BD Biosciences) with FACSCalibur flow cytometer (BD Biosciences).

The MDA-MB-231 cell populations with stable AP4 knockdown, LAPTM4B overexpression, AP4 knockdown+LAPTM4B overexpression, and control vectors were constructed by lentiviral infection and puromycin selection. A total of 20 six-week-old female SCID mice purchased from Beijing HFK Bioscience Co., Ltd were randomly distributed into four groups (5 mice per group). Each mice was injected subcutaneously into the flanks with 100-μL PBS containing 2 × 10⁶ tumor cells. The tumor volume was determined twice every week by measuring the length (l) and width (w) and then calculating the volume as 1/2(w²×l). All animal studies were conducted with the approval of Medical Experimental Animal Care Commission of Peking University School of Oncology.

All the statistical analyses were evaluated using the Statistical Software Package for the Social Sciences (SPSS software version 19.0, SPSS), and P values <0.05 were considered to be statistically significant. Data were reported as the means ± SD. Spearman rank correlation was utilized to evaluate the relationship between AP-4 and LAPTM4B expression levels of breast cancer tissues. Survival analysis was performed using the Kaplan–Meier method. A Student t test was used to assess differences between two groups. Differences in cell growth curves and in vivo tumorigenicity were confirmed with repeated-measures ANOVA.

We used the promoter-binding TF profiling plate arrays to explore the transcription factors which could bind to LAPTM4B promoter region in breast cancer cell. Genomic DNA fragments of two alleles of LAPTM4B (LAPTM4B*1: −200bp∼+301bp and LAPTM4B*2: −200bp∼+320bp) were used as TF-bound probes in MCF7 cells (Supplementary Fig. S1). As a result, we screened 10 transcription factors that may bind to the the promoter of LAPTM4B*1 and 24 factors binding to LAPTM4B*2 (Supplementary Fig. S2). Then, we predicted the potential transcription factors binding to 19-bp polymorphism region of LAPTM4B alleles by computer analysis (Supplementary Table S3, S4). As a result, only AP-4 and ELK were selected after matching the result of TF profiling plate arrays. Therefore, AP4 which got the higher relative score was chosen for further study.

We first examined AP-4 and LAPTM4B protein levels in four breast cancer cell lines (MCF7, ZR-75-1, T47D, and MDA-MB-231) and one immortalized breast epithelial cell line MCF10A by Western blot analysis (Fig. 1A). MCF10A exhibited the lowest expression levels of both AP4 and LAPTM4B, while MCF7 and MDA-MB-231 got the higher expression than the other cell lines (Fig. 1B). Thus, LAPTM4B protein expression may be correlated with AP4 expression in these breast epithelial cell lines. Furthermore, we analyzed the mRNA levels of AP4 and LAPTM4B in 1021 breast cancer patients from the TCGA (The Cancer Genome Atlas) database and found that LAPTM4B expression significantly correlates with AP4 at the mRNA level (Fig. 1C). In our study, we examined the expression levels of AP-4 and LAPTM4B in 146 breast cancer tissues by IHC (Table 1; Fig. 1D). As a result, AP-4 was significantly positively correlated with LAPTM4B expression in breast cancer tissues (Table 1). Survival analysis indicated that patients with high levels of both AP4 and LAPTM4B had the poorest outcome in terms of overall survival and disease-free survival (Fig. 1E).

To further test whether LAPTM4B is a direct target of AP4, we analyzed the sequence upstream of the LAPTM4B gene. Figure 2A is model chart of the potential AP4-binding sites in the LAPTM4B polymorphism region and luciferase reporter constructs, demonstrating that there is only one AP4-binding site in the polymorphism region of LAPTM4B*1 allele but include two sites of LAPTM4B*2 due to the repetition of 19 bp. The core sequence of the truncated LAPTM4B promoter and the mutated AP4-binding sites were shown in Fig. 2B. Dual-luciferase reporter assay showed the fragment of +10∼+292 (pGL3-PF6) had the highest transcriptional activity among the seven 5′-deleted constructs (Fig. 2C). The transcriptional activities of mutated constructs with AP4-binding site mutation of LAPTM4B*1 allele (pGL3-PF6 mut) were significantly decreased than the wild-type. The transcriptional activities of constructs with only one AP4-binding site mutation of LAPTM4B*2 allele (pGL3-PF7 mut1, pGL3-PF7 mut2) were significantly increased, but became the lowest when the two binding sites of AP4 (pGL3-PF7 mutD) were mutated at the same time (Fig. 2C). The results suggest that AP4 could regulate the transcriptional activity of LAPTM4B gene promoter and may affect the transcription activity of LAPTM4B*2 allele more than LAPTM4B*1.

EMSA and super-shift assay showed that transcription factor AP4 could bind to the LAPTM4B polymorphism region specifically in MCF7 cells (Fig. 2D and E). ChIP assay was performed to assess the effect on the direct binding of AP4 to the LAPTM4B promoter and the result showed that anti-AP4 antibody, not goat IgG, specifically (15-fold and 8-fold difference, respectively) immunoprecipitated the LAPTM4B gene promoter in MDA-MB-231 with LAPTM4B*2/2 genotype (29), and MCF7 with LAPTM4B*1/1 genotype (Fig. 2F and G). Taken together, the results suggest that more AP4 protein binds to the LAPTM4B*2 allele than LAPTM4B*1 in its promoter region, which may result that AP4 affected the transcription activity of LAPTM4B*2 allele much more.

To investigate the regulation of LAPTM4B by AP4, an AP4 shRNA vector or AP4 overexpression vector was introduced into MCF7 and MDA-MB-231 cells. AP4 knockdown dramatically reduced the LAPTM4B mRNA and protein levels in both cell lines (Fig. 2H). The results were also confirmed by three siRNA transfection (Supplementary Fig. S3). AP4 overexpression consistently and dramatically upregulated LAPTM4B expression at mRNA and protein levels (Fig. 2I).

To explore the roles of LAPTM4B in AP4-mediated breast cancer progression, a rescue experiment was then designed. We knocked down AP4 using shRNA but overexpressed LAPTM4B in MCF7 and MDA-MB-231 cell lines (Fig. 3A). The CCK-8 assay showed that AP4 shRNA noticeably reduced breast cancer cell viability and proliferation; however, LAPTM4B overexpression markedly rescued cell proliferation arrest observed with AP4 depletion (Fig. 3B). The inhibition of colony formation by AP4 shRNA was also antagonized by LAPTM4B overexpression in both cell lines (Fig. 3C). In addition, cell distribution in the cell cycle was determined by flow cytometry. Cells with AP4 knockdown displayed significantly higher frequency at G₀–G₁ phase in MCF7 and MDA-MB-231 cell lines and a lower frequency at S-phase in MCF7 while overexpression of LAPTM4B partially restored the influence on the G₁–S transition in AP4-shRNA cells (Fig. 3D and E). Furthermore, cellular levels of some cell-cycle regulators showed that c-Myc, cyclin E1, and cyclin D1 were significantly decreased in AP4-shRNA cells, but the cyclin-dependent kinase inhibitor p27Kip1(CDKN1B), p21 (CDKN1A), and p16 (CDKN2A) were increased with AP4 knockdown. The effect of these protein levels could also be partly rescued by LAPTM4B overexpression. However, p21 was not affected by LAPTM4B overexpression in MDA-MB-231 cell lines (Fig. 3F).

Our in vivo data showed that knockdown of AP4 significantly inhibited MDA-MB-231-bearing xenograft growth and overexpression of LAPTM4B accelerated xenograft growth (Fig. 3G). The weight and volume decreased in AP4 knockdown tumors (Fig. 3G and H). Consistent with the in vitro data, AP4-inhibited tumor growth was significantly reversed by LAPTM4B overexpression (Fig. 3G and H). These data collectively suggest that AP4 could promote cell proliferation at least partly by upregulating LAPTM4B.

The effect of LAPTM4B on AP4-mediated cell migration and invasion was then studied. The transwell assays showed that AP4 shRNA noticeably reduced cell migration and invasion and LAPTM4B overexpression significantly increased the number of migrated and invaded cells in MCF7 and MDA-MB-231 cell lines (Fig. 4A and B). The inhibition of cell migration and invasion induced by AP4 shRNA was markedly rescued by LAPTM4B overexpression in both cell lines (Fig. 4A and B). Furthermore, AP4 depletion resulted in downregulation of the mesenchymal marker vimentin and N-cadherin expression and upregulation of E-cadherin protein levels in MCF7 and MDA-MB-231 cell lines (Fig. 4C). The expression levels of metastasis-related genes MMP-2 and MMP-9 were also decreased in AP-shRNA cells (Fig. 4C). These effects could be partially restored by LAPTM4B overexpression (Fig. 4C). Therefore, AP4 promotes cell migration and invasion partly by upregulating LAPTM4B through EMT signals and metastasis-related genes.

MDR (multidrug resistance, MDR) is a major clinical obstacle in the treatment of cancers. To determine whether AP4 and LAPTM4B can change the chemotherapy sensitivity of breast cancer cells, the inhibition values for cisplatin in MCF7 and MDA-MB-231 cell lines were determined. As shown in Fig. 5A, AP4 shRNA could significantly enhance the chemosensitivity of breast cancer cells to cisplatin, compared with the control groups. Forced overexpression of LAPTM4B reversed the promotion of the chemosensitivity to cisplatin induced by AP4 shRNA (Fig. 5A). LAPTM4B has been shown to motivate multidrug resistance by activation the PI3K/AKT signaling pathway (30). In our study, we detected the effects of PI3K-specific inhibitor on AP4- and LAPTM4B-associated MDR. The results showed that the effects of AP4 and LAPTM4B on MDR were eliminated in the presence of LY294002 (Fig. 5B).

In addition, the relationship between AP4, LAPTM4B, and the activation of PI3K/AKT signaling pathway was explored. As shown in Fig. 5C, lower levels of phosphorylations of AKT, GSK3β, appeared in AP4 knockdown cells compared with control cells, but no difference of the total AKT protein level was found. The alterations of these proteins involved in PI3K/AKT signaling pathway could also be partly rescued by LAPTM4B restoration. The results also showed that treatment with LY294002 greatly reduced the p-AKT and p-GSK3β in either AP4 knockdown or its controls, and eliminated the elevation of p-AKT and p-GSK3β in LAPTM4B -overexpressing cells (Fig. 5C), implicating the involvement of PI3K/AKT signaling in AP4 and LAPTM4B-motivated MDR. As described above, AP4 and LAPTM4B could influence c-MYC expression in MCF7 and MDA-MB-231 cell lines. We further examined whether the regulation of c-MYC is through PI3K/AKT signaling. As shown in Fig. 5C, c-MYC was significantly decreased in the presence of LY294002 and the regulation by AP4 and LAPTM4B was eliminated.

Chemotherapeutic drugs could induce a series of cellular response that impact on tumor cell apoptosis. As shown in Fig. 5D and E, AP4 knockdown cells were more sensitive to apoptosis in the control group (no cisplatin treatment) or cisplatin group. The effects were rescued significantly by LAPTM4B restoration in MCF7 and MDA-MB-231 cell lines. Moreover, our results herein showed that in the presence of cisplatin, the levels of active forms of caspase-3 and nuclear PARP enhanced in AP4-shRNA cells, but attenuated dramatically with LAPTM4B overexpression (Fig. 5F).

Therefore, it is confirmed that one of the mechanisms for AP4 and LAPTM4B motivation cisplatin resistance is antiapoptosis by activating PI3K/AKT signaling pathway.

LAPTM4B, an oncogene, is highly expressed in many solid tumors and inversely associated with overall survival and disease-free survival of patients with breast cancer (31), gastric cancer (32), gallbladder cancer (33), HCC (34), ovarian cancer (35), and cervical cancer (36). Moreover, the polymorphism region in the 5′-UTR of LAPTM4B gene was certified to be associated with tumor susceptibility including breast cancer (15–17). However, the mechanism of the transcriptional regulation of LAPTM4B has been rarely studied. In this study, we identified that transcription factor AP4 could bind to the polymorphism region of LAPTM4B gene and noticeably regulate its expression in breast cancer. Moreover, AP4 could promote breast cancer proliferation, migration, and drug resistance partly through upregulation of LAPTM4B.

Transcription factor AP4 is a member of the basic helix-loop-helix (bHLH) protein family. It was initially shown to activate late viral gene expression from the SV40 enhancer (37). It has been reported that AP4 is upregulated in many types of tumors and linked to poor prognosis in cancer patients (38, 39). We also demonstrated herein that AP4 and LAPTM4B were coexpressed in breast epithelial cell lines and breast tissues, and patients with higher levels of AP4 and LAPTM4B were likely to associated with shorter overall and disease-free survival. In addition, AP4 could bind to the polymorphism region of LAPTM4B promoter and promote LAPTM4B expression in breast cancer cells. Interestingly, luciferase assay showed that the transcriptional activity of LAPTM4B*1 promoter was significantly decreased with AP4-binding site mutation. However, the activity was increased when a single AP4-binding site was mutated in LAPTM4B*2 promoter. It concludes that there may be trans-acting factors playing negative regulatory roles in the polymorphism region of LAPTM4B*2 and further research is needed to clarify it. We further showed that AP4 was capable of inducing cell growth and migration by upregulating LAPTM4B in breast cancer. Conversely, a previous study reported that AP4 could suppress MCF7 cell proliferation by enhancing SHP1 expression and JNK activation (40). It indicates that AP4 may play a dual role in cell proliferation via different pathways. Similar to our results, AP4 has been reported to promote cell growth in HCC (41), colorectal carcinoma (42), and gastric cancer (43). Moreover, it has been proposed that AP4 could activate cell migration and EMT mediated by p53 in MDA-MB-231 cells (44) and mediate a c-MYC–induced EMT in colorectal cancer (45).

Mechanistically, AP4 transcriptionally upregulated LAPTM4B expression to activate c-MYC, p21, and EMT signaling pathway to enhance cell growth and migration in the study. Interestingly, it has been shown that c-MYC could directly regulate AP4 expression and c-MYC-AP4-p21 cascade plays an important role in maintaining cells in a proliferative, progenitor-like state of HCC and colorectal carcinoma (41, 42). Taken together, these results suggest that c-MYC–AP4–LAPTM4B–c-MYC axis may form a feedback loop to participate in cell growth and metastasis.

Recent study revealed that downregulation of AP4 effectively triggered apoptosis and enhanced chemosensitivity of human gastric cancer cells (43). Amplification of LAPTM4B and YWHAZ contributed to anthracycline resistance and recurrence of breast cancer (46). Moreover, LAPTM4B was a marker of resistance to neoadjuvant chemotherapy in HER2-negative breast cancer (47). One of the mechanisms for MDR involves in alternations in the apoptotic response associated with PI3K/AKT signaling. The PPRP motif in the N-terminus of LAPTM4B-35 protein directly binding to PI3K p85 α subunit could strongly phosphorylate AKT and then motivate MDR in HeLa cells (30). In this study, we showed that silencing AP4 induced apoptosis and sensitized cells to cisplatin via activating PI3K/AKT pathway partly by suppressing LAPTM4B. The findings may provide a promising novel strategy for sensitizing chemical therapy of cancers. In addition, AKT could regulate c-MYC through GSK3β in response to mitochondrial stress (48). In our study, we found that AP4 and LAPTM4B may influence c-MYC expression through PI3K/AKT signaling with PI3K inhibitor.

However, we did not find any difference on the regulation of LAPTM4B alleles between MCF7 (LAPTM4B*1/1) and MDA-MB-231 (LAPTM4B*2/2) cell lines. It does not mean that there is no regulation distinction due to different characteristics of cell itself. Thus, the roles of AP4 on the two alleles in the same cell line should be further studied.

In conclusion, this is the first report demonstrating that AP4 is positively correlated with LAPTM4B expression and predicts unfavorable overall and disease-free survival in breast cancer. AP4 tightly controls LAPTM4B gene transcription by binding to the polymorphism region of LAPTM4B promoter. In addition, knockdown of AP4 inhibits proliferation, migration, invasion, and chemotherapy resistance through suppression of LAPTM4B in breast cancer (Fig. 6). Collectively, our data suggest that the newly identified AP4/LAPTM4B axis may play an important role in breast cancer progression and serve as potential therapeutic targets for clinical management.

No potential conflicts of interest were disclosed.

Conception and design: L. Wang, Y. Meng, Q.Y. Zhang

Development of methodology: L. Wang, Y. Meng, Q.Y. Zhang

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): L. Wang, Y. Meng

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L. Wang, Y. Meng

Writing, review, and/or revision of the manuscript: L. Wang, Q.Y. Zhang

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L. Wang, Y. Meng, J.J. Xu

Study supervision: J.J. Xu, Q.Y. Zhang

This study was supported by National Natural Science Foundation of China (No. 81572910; awarded to Q.Y. Zhang).

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.