Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Identification of a Modified HOXB9 mRNA in Breast Cancer

Identification of a Modified HOXB9 mRNA in Breast Cancer Hindawi Journal of Oncology Volume 2020, Article ID 6065736, 10 pages https://doi.org/10.1155/2020/6065736 Research Article Ayako Nakashoji, Tetsu Hayashida , Yuko Kawai, Masayuki Kikuchi, Rurina Watanuki, Takamichi Yokoe, Tomoko Seki, Maiko Takahashi, Kazuhiro Miyao, Shigeo Yamaguchi, and Yuko Kitagawa Department of Surgery, Keio University School of Medicine, Shinanomachi 35 Shinjuku-ku, Tokyo 160-0016, Japan Correspondence should be addressed to Tetsu Hayashida; tetsu@keio.jp Received 3 August 2019; Revised 29 December 2019; Accepted 7 January 2020; Published 13 February 2020 Academic Editor: Rossana Berardi Copyright © 2020 Ayako Nakashoji et al. +is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. First identified as a developmental gene, HOXB9 is also known to be involved in tumor biological processes, and its aberrant expression correlates with poor prognosis of various cancers. In this study, we isolated a homeodomain-less, novel HOXB9 variant (HOXB9v) from human breast cancer cell line-derived mRNA. We confirmed that the novel variant was produced from variationless HOXB9 genomic DNA. RT-PCR of mRNA isolated from clinical samples and reanalysis of publicly available RNA- seq data proved that the new transcript is frequently expressed in human breast cancer. Exogenous HOXB9v expression sig- nificantly enhanced the proliferation of breast cancer cells, and gene ontology analysis indicated that apoptotic signaling was suppressed in these cells. Considering that HOXB9v lacks key domains of homeobox proteins, its behavior could be completely different from that of the previously described variationless HOXB9. Because none of the previous studies on HOXB9 have considered the presence of HOXB9v, further research analyzing the two transcripts individually is warranted to re-evaluate the true role of HOXB9 in cancer. promotes metastasis by activating the WNT signaling 1. Introduction pathway [5, 6]. In breast cancer, the gene induces the ex- Homeobox (HOX) genes were initially characterized as pression of proangiogenic factors, increasing the cell mo- developmental genes, which code for transcription factors tility and supporting epithelial-mesenchymal transition that play critical roles in embryogenesis. Evolutionarily, they (EMT) [7, 8]. HOXB9 also promotes the growth of colon are highly conserved and share a high degree of homology, cancer by activating IL6 signaling, inducing the secretion of especially within the same paralog groups. All 39 mam- angiogenic factors and increasing proliferation of tumor malian HOX genes consist of two exons and a single intron. cells [9]. Similar observations are found in ovarian cancer +e homeobox domain, encoded in the second exon [1], and hepatocellular carcinoma [10, 11]. +us, HOXB9 acti- includes the DNA binding site. +e diverse and specific vates the WNT signaling pathway and enhances the ac- transcriptional activities of the HOX proteins often depend quisition of capabilities critical to the transformation of on key cofactors including PBX, MEIS, and PREP, which normal cells to cancer, including EMTand the growth of new interact with hexapeptide motifs of HOX proteins [2]. vasculature within the tumor microenvironment. HOX genes play key roles in both solid and hemato- HOXB9 is also directly associated with cancer-induced logical malignancies, including cancers of the colon, breast, patient mortality. +e duration of disease-free and overall prostate, lung, brain, thyroid, ovary, bladder, kidney, skin, survival of patients with HOXB9-positive breast cancer is and blood [3, 4]. HOXB9, the ninth paralog in the HOX-B significantly shorter compared with patients with HOXB9- cluster, is associated with the growth and progression of negative breast cancer [12]. Increased HOXB9 expression multiple cancers. In lung adenocarcinoma, HOXB9 significantly correlates with decreased overall survival for 2 Journal of Oncology patients with colorectal cancer [9]. Patients of laryngeal Minilys homogenizer (Bertin Instruments, Bretonneux, squamous cell carcinoma, hepatocellular carcinoma, glioma, France) using 2.4 mm metal beads prior to mRNA extrac- tion. Total RNA was converted to cDNA using the High- and endometrial cancer also present poor outcomes or tumor progression, resulting from aberrant HOXB9 ex- Capacity RNA-to-cDNA Kit (+ermo Fischer Scientific). pression [13–16]. +e recent elucidation of the critical and diverse roles 2.4. HOXB9 Cloning. HOXB9 transcripts and genomic HOXB9 plays in various cancers have led us to explore the DNA were amplified by PCR and subsequently cloned into mechanism of this gene’s role in cancer progression. In the the pME-HA vector (Lucigen, Middleton, WI, USA) using present study, we identified and characterized a HOXB9 the Expresso CMV Cloning & Expression system (Lucigen). variant (HOXB9v) of mRNA from human breast cancer cell +e primers used were as follows: sense 5′ GAAGGAGA- lines. +e sequence of HOXB9v largely differs from the TACCACCATGTCCATTTCTGGGACGCTTAGC 3′ and previously known HOXB9 normal transcript (HOXB9n), antisense 5′ GGGCACGTCATACGGATACTCTTTGCCC and we found it lacks some important domains of HOX TGCTCCTTATT 3′. genes. Based on these findings, we inferred its role and After transformation into Competent Quick DH5α cells function were different from HOXB9n. +is study aimed to (Toyobo, Osaka, Japan) and culturing in kanamycin-con- confirm the presence of HOXB9v in clinical breast cancer taining LB plates, at least 4 colonies were selected for each specimens and investigate the role of HOXB9v in breast sample. cancer progression. 2. Materials and Methods 2.5. Sequence Analysis. Sequences were analyzed using the BigDye Terminator V3.1 Cycle Sequencing kit (+ermo 2.1. Cell Culture. We cultured eight human breast cancer cell Fisher Scientific) and Applied Biosystems 3500 Genetic lines (MCF7, MDA-MB-231, MDA-MB-468, Hs578T, analyzer (+ermo Fisher Scientific). +e GENETYX-MAC HCC38, BT-474, BT-549, and SKBR3), a human colon Ver.19 software (GENETYX, Osaka, Japan) was used for cancer cell line (WiDR), and a mouse breast cancer cell line homology alignment. (4T1) in DMEM supplemented with 10% FBS (+ermo Fisher Scientific, Waltham, MA, USA) with the addition of antibiotic and antimycotic agent (antibiotic-antimycotic 2.6. RT-PCR Analysis. +e expressions of HOXB9n and HOXB9v were detected by RT-PCR using the following mixed stock solution, Nacalai Tesque, Inc., Kyoto, Japan). primers: sense, 5′ TGTCCATTTCTGGGACGCTT 3′; an- T47D cells were grown in RPMI1640 medium with 10% FBS tisense, 5′ CTACGGTCCCTGGTGAGGTA 3′. +e genomic and antibiotic and antimycotic agents. MCF10A cells were DNA of HOXB9 was detected by PCR using the following grown in DMEM/F12 (1 :1) medium with 5% horse serum primers: sense, 5′ CGAGAGAGCTGCAAGTCGAT 3′; (+ermo Fischer Scientific), 20 ng/mL EGF (PeproTech, antisense, 5′ CTGCCGTCCGTCTACCAC 3′. +e primers Rocky Hill, NJ, USA), 0.5 mg/mL hydrocortisone, 100 ng/ for genomic DNA were designed against exon 1 and the mL cholera toxin, 10 µg/mL insulin (all Sigma-Aldrich, St. intron region of the HOXB9 gene to ensure that the pair will Louis, MO, USA), and antibiotic and antimycotic agents. specifically amplify only genomic DNA and not cDNA Cells were maintained at 37 C in a humidified 5% CO derived from mRNA. +e conditions applied for amplifi- incubator. cation were as follows: 94 C for 1 minute, followed by 35 ° ° ° cycles at 95 C for 5 seconds, 55 C for 5 seconds, and 72 C for 2.2. Patients and Samples. Clinical specimens of human 5 seconds, and run on the Life ECO thermal cycler breast cancer (n � 14) were collected from patients with (Hangzhou Bioer Technology, Hangzhou, China) using the primary operable breast cancer who underwent total or SapphireAmp Fast PCR Master Mix (Takara Bio, Shiga, partial mastectomy between July and November 2018 in Japan) or KOD-Plus-Neo (Toyobo). Keio University Hospital (Tokyo, Japan). Patient-matched healthy breast epithelium samples (n � 6) were collected from healthy breast tissue from patients who underwent 2.7. Public Data Reanalysis. +e RNA sequence data set (GSE119937) in FASTQ format was downloaded via SRA total mastectomy. Ethics approval for the present study was provided by the Ethics Committee at the Keio University (SRP161704) using the SRA Toolkit (version 2.3.4-2). RNA- seq reads were aligned by STAR (version 2.6.1b) against the School of Medicine (approval number: 20180090), and the study was performed in accordance with the provisions of hg38 reference genome. All reads mapped on the HOXB9 the Declaration of Helsinki (as revised in Fortaleza, Brazil, gene were visually confirmed by taking snapshots in IGV October 2013). All included patients gave informed consent. (version 2.4.15). 2.3. mRNA and Genomic DNA Extraction from Cell Lines and 2.8. Establishment of Stable MCF7 Cell Lines Overexpressing Clinical Specimens. Total RNA and genomic DNA were HOXB9n or HOXB9v. +e HOXB9n and HOXB9v se- extracted using the RNeasy Mini Kit and QIAmp DNA Mini quences were amplified by PCR and subsequently cloned Kit (Qiagen, Hilden, Germany) following the manufacturer’s into the pBiT3.1-N [CMV/HiBiT/Blast] expression vector instructions. Clinical specimens were homogenized using a (Promega, Tokyo, Japan) at the Xhol and BamHI sites using Journal of Oncology 3 Ligation high Ver.2 (Toyobo). +e primers used for am- results were analyzed by TAC 4.0 software (+ermo Fisher plification were as follows: sense, 3′ ATACCTCGAGG Scientific). Genes having a false discovery rate (FDR) under 0.05 and upregulated in HOXB9v samples were considered TCCATTTCTGG 5′; antisense, 3′ CACGTCATACGGAT CCTCTTTG 5′. MCF7 cells were transfected with the as upregulated differentially expressed genes (up-DEGs). HiBiT-tagged HOXB9n or HOXB9v vector using the Viafect DAVID Bioinformatics Resources 6.8 [17] was used for gene Transfection reagent (Promega) and selected with 10 µg/mL ontology and pathway analysis of DEGs. R ver. 3.5.0 software of blasticidin for at least 4 weeks. +e expression of HiBiT- was used to draw the heatmap of 37 genes included in GO: tagged HOXB9n and HOXB9v proteins was detected using 0043069 (negative regulation of programmed cell death). the Nano-Glo HiBiT Blotting System (Promega) as per the manufacturer’s instructions. 3. Results 3.1. Identification of a Novel Transcript HOXB9v, Which Lacks 2.9. Transient Overexpression of HOXB9n and HOXB9v in Important Domains of HOX Gene. We isolated total RNA MDA-MB-468 Cells. MDA-MB-468 cells were transfected from human breast cancer cell lines (MCF7, T47D, MDA- with aforementioned HiBiT-tagged HOXB9n or HOXB9v MB-231, MDA-MB-468, HCC38, BT-474, BT-549, Hs578T, vector using the jetPRIME (Polyplus-transfection, Illkirch, and SKBR3), a normal human mammary gland cell line France) as per the manufacturer’s instructions. (MCF10A), and a human colon cancer cell line (WiDR). We cloned and sequenced the HOXB9 gene from the nucleic 2.10. Quantitative Real-Time PCR. Quantitative real-time acids derived from these cell lines. We isolated a novel PCR was run on ViiA7 (+ermo Fisher Scientific) using Fast variant of the HOXB9 transcript from the MCF7, T47D, SYBR Green Master Mix (+ermo Fisher Scientific). Pre- MDA-MB-231, MDA-MB-468, Hs578T, HCC38, and incubation was performed for 20 seconds at 95 C and MCF10A cell lines. +e sequence homology of the new amplification for 41 cycles (1 second of denaturation at 95 C transcript is shown in Figures 1(a) and S1. and 20 seconds of annealing and extension at 60 C), followed We will refer to the new truncated transcript as HOXB9v, by melt-curve analysis. GAPDH served as an internal con- to distinguish from the full-length HOXB9 transcript trol, and QuantStudio Real-Time PCR Software v1.2 (HOXB9n). A 100-base deletion in exon 1 in the new transcript leads to a frameshift and the formation of a stop (+ermo Fisher Scientific) was used for quantification. +e relative standard curve method was used for linear regres- codon (TAG), which truncates the protein coding at AA167. sion analysis of unknown samples, and data are presented as Figure 1(b) is a schematic diagram of HOXB9 transcripts fold change between samples. +e primers used were as showing the exons and the deleted lesion. +e encoded follows: HOXB9: sense, 5′ CGGTGGCTGTCGTGAAATT protein will therefore possibly lack the homeobox domain, 3′; antisense, 5′ CGAGACAATCACCCCCAAAG 3′; GAP DNA binding domain, and the hexapeptide motif, a major DH: sense, 5′ATCATCCCTGCCTCTACTGG 3′; antisense, player in cofactor interactions (Figure 1(c)). +e HOXB9v 5′ TTTCTAGACGGCAGGTCAGGT 3′. sequence has been submitted to GenBank under Accession No. LC466645. We next performed PCR analyses to verify the presence 2.11. Cell Proliferation Assay. Cell proliferation in three- of the HOXB9v transcript and identify genomic DNA dimensional (3D) culture was measured using the 24-well variations of the HOXB9 gene in human breast cancer cell Bio-Assembler kit and NanoShuttle-PL (Greiner Bio-One, lines. +e primer target region included the deletion site of Kremsmunster, ¨ Austria). Further, 20,000 cells/well were HOXB9v and the amplicon for HOXB9n was 643 bp and incubated for 48 hours before taking photomicrographs. 543 bp for HOXB9v. We detected both HOXB9n and Cell proliferation in flat culture was measured using Cell HOXB9v transcripts in breast cancer cell lines (Figure 2(a)). Count Reagent SF (Nacalai Tesque). Briefly, 5000 cells/well +e PCR products were sequenced and were confirmed that in 96-well microtiter plates (Sumilon, Sumitomo Bakelite, each band corresponded to the exact sequence of the Tokyo Japan) were incubated for 5 days or after 24 hour HOXB9n or HOXB9v (Figure 2(b)). By PCR of breast cancer continuous exposure to either 30 nM HXR9 or 30 nM CXR9 cell lines’ genomic DNA, no bands indicative of genomic for 4 days. +e absorbance in the wells was measured on days DNA variations in HOXB9 were detected. +e amplicon of 1, 3, and 5 using a Sunrise Rainbow-RC (TECAN, variation-less HOXB9 genomic DNA was 569 bp Mannedorf, ¨ Switzerland) microplate spectrophotometer at (Figure 2(c)). Sequencing of the PCR products confirmed 450 nm, using 600 nm as reference. +e HXR9 that they had no variations or deletions in genomic DNA (WYPWMKKHHRRRRRRRRR-) and control CXR9 (Figure S2). +ese findings show that the two transcripts (WYPAKKHHRRRRRRRRR) peptides were synthesized by (HOXB9n and HOXB9v) were produced from variation-less Eurofins Genomics K. K. (Tokyo, Japan). HOXB9 genomic DNA. To determine the presence of HOXB9v transcripts in 2.12. Microarray and Differential Expression Analyses. human breast cancer samples, we performed PCR analysis. Total RNA was isolated using the RNeasy Mini kit (Qiagen). HOXB9v was commonly detected from clinical breast cancer Microarray was performed using the human Clariom S assay samples (Figure 3(a)), regardless of their hormone receptor (+ermo Fisher Scientific) by the GeneChip Scanner 3000 and HER2 status. However, HOXB9v was not detected in 7G system (Affymetrix, Santa Clara, CA, USA), and the normal mammary gland samples (Figure 3(b)). 4 Journal of Oncology HOXB9 gDNA 445 TACATCC AGCCCCAGGGCGTCCCGCCGGCCGAGAGCAGGTACCTCCGCACCTGGCTGGAGCCGGCGCCGCGCGGCGAAGCGGCCCCGGGGCAGGGCCAGGCGGCGGTGAAGGC GGAGCCGCTGCTGGGCG 574 HOXB9n mRNA 241 TACATCC AGCCCCAGGGCGTCCCGCCGGCCGAGAGCAGGTACCTCCGCACCTGGCTGGAG CCGGCGCCGCGCGGCGAAGCGGCCCCGGGGCAGGGCCAGGCGGCGGTGAAGGC GGAGCCGCTGCTGGGCG 370 HOXB9v mRNA 241 TACATCC TGCCCC – GGGG – CCGCTGCTGGGCG 270 (a) Exon I Exon II ATG TAA HOXB9n 1 205 721 722 955 2701 ATG TAG HOXB9v 1 205 462 562 721 722 2701 (b) Hexapeptide & linker HOXB9n N-terminal flexible region Homeodomain Nucleotide 1 528 555 753 AA 1 176 185 251 Frameshift HOXB9v Nucleotide 1 257 357 501 653 AA 1 85fs 167 (c) Figure 1: Structure and sequence of HOXB9n and HOXB9v. (a) Sequences of genomic DNA HOXB9, mRNA HOXB9n, and mRNA HOXB9v (black and grey highlighting indicates homology between sequences). (b) Schematic diagram of HOXB9n (upper) and HOXB9v (lower) transcripts showing the exons, splicing regions, and the deleted region. In HOXB9v, a 100 bp deletion in exon 1 leads to a frameshift and a stop codon formation (TAA). (c) Protein structure of HOXB9n and HOXB9v. Transcription from the start codon (ATG) to the stop codon (TAA) results in the translation of a full-length HOXB9n protein (upper). +e 100 bp deletion in HOXB9v (shown in black) leads to a frameshift from AA85 (shown in grey) and truncation of protein coding by a stop codon (TAG) at AA167, which results in HOXB9v protein without hexapeptide and homeodomain. Stop codon. MDA- MDA- MCF7 T47D Hs578T SKBR3 MCF10A HOXB9n HOXB9v MB-231 MB-468 700bp –– –– 700bp 600bp –– –– 600bp 500bp –– –– 500bp (a) (b) MDA- MDA- MCF7 T47D Hs578T MCF10A MB-231 MB-468 600bp –– 500bp –– 400bp –– (c) Figure 2: Detection of mRNA and genomic DNA of HOXB9 in cell lines. (a) Both HOXB9n and HOXB9v transcripts were detected in breast cancer cell line mRNA (cDNA). (b) Sequencing confirmation of PCR product of Figure 2(a); SKBR3 (HOXB9n, upper column) and MCF10A (HOXB9v, lower column). (c) No genomic variation was detected in genomic DNA of breast cancer cell line genomic DNA. HOXB9v transcripts are commonly found in human breast cancer specimens. Journal of Oncology 5 Sample# 1 2 3 45678 9 10 11 12 13 14 HR/HER2 +/+ +/– +/– +/– +/– –/– +/– –/– +/– +/– –/– +/– +/– –/– (a) Sample# 125 11 12 14 HOXB9n HOXB9v (b) Figure 3: Detection of HOXB9v in clinical samples. (a) HOXB9n and HOXB9v were detected in breast cancer clinical samples. (b) HOXB9v was not detected in normal mammary gland samples. Hormone receptor status and HER2 status of each cancer sample are shown beneath each sample number. We assigned the same sample number if cancer and normal gland samples were acquired from the same patient. SRR7826774 T3 N1 IDC (primary) ER – PgR – HER2–, ki67 60% SRR7826783 T1cN0 IDC (primary) ER + PgR – HER2–, ki67 5% SRR7826861 T1aN1 IDC (primary) ER + PgR + HER2–, ki67 0% Exon II Intron Exon I 3′ 5′ ATG HOXB9n TAA HOXB9 HOXB9v TAG ATG Figure 4: NGS data indicates the presence of HOXB9v in breast cancer. Showing RNA sequence breast cancer sample data (SRR782677) mapped on HOXB9 exon 1. +e blue band on the bottom shows the coding region (the thick part) and the 5′ UTR (the thin part) of exon 1. +ere is a region where data are sparsely mapped (indicated with the red arrow), which matches the deletion region in HOXBv. HOXB9v overexpressing MCF7 and MDA-MB-468 cells proliferated faster compared with HOXB9n overexpressing cells. +us, to further confirm the presence of HOXB9v in transiently overexpressed HOXB9n or HOXB9v in MDA- human breast cancer samples, we reanalyzed a publicly MB-468 cells, and the cell proliferation assay showed that available breast cancer RNA sequence data set HOXB9v overexpression increased MDA-MB468 cell (GSE119937) [18] and mapped the reads onto the HOXB9 growth (Figures 5(e) and 5(f)). gene sequence. We identified a region in exon 1 where the number of mapped reads was low in numerous samples; 3.2. HXR9 and CXR9 Treatment. HXR9, an 18-amino acid this region matched the deletion region of HOXB9v (Figure 4). peptide, competently inhibits the hexapeptide motif of HOX proteins and prevents HOX-PBX binding [19]. To determine the role of HOXB9v in breast cancer, we established stable MCF7 cell lines overexpressing HOXB9n HOXB9n and HOXB9v expressing MCF7 cells were treated with HXR9 or a control peptide, CXR9 or HOXB9v. Gene and protein expressions of HOXB9n and HOXB9v were verified in both cell lines (Figures 5(a) and (Figures 6(a) and 6(b)). Although HOXB9v lacks the hexapeptide motif, which is known to interact with PBX 5(b)). Cell proliferation assays in both 3D culture and flat culture showed that HOXB9v overexpression increased proteins [20], HXR9 significantly inhibited proliferation of both cell lines. MCF7 cell growth (Figures 5(c) and 5(d)). We also 6 Journal of Oncology log 10 HOXB9v HOXB9n ∗ (μm) 35kDa 1834.8 1601.0 25kDa 17kDa HOXB9v HOXB9n 0 HOXB9v HOXB9n p = 1.7247E – 06 (a) (b) (c) log 10 1.4 100000 0.5 1.2 10000 0.4 1000 0.3 0.8 ∗ 0.6 100 0.2 0.4 10 0.1 0.2 0 0 HOXB9v HOXB9n Day1 Day3 Day5 Day1 Day3 Day5 HOXB9n HOXB9n ∗ ∗ p = 0.0134 p = 4.40E – 07 HOXB9v HOXB9v (d) (e) (f) Figure 5: (a–d) Confirmation and proliferation of stable MCF7 cell lines that overexpress HOXB9n or HOXB9v. Levels of HOXB9n and HOXB9v (a) mRNA and (b) protein in stable cell lines. Proliferation of the HOXB9n and HOXB9v stable cell lines using a (c) 3D cell culture model and a (d) 2D cell culture model. (e-f) Confirmation and proliferation of MDA-MB-468 cells with transient overexpression of HOXB9n or HOXB9v. (e) Levels of HOXB9n and HOXB9v mRNA expression. (f) Proliferation of HOXB9n and HOXB9v overexpressing cells by 2D cell culture model. 3.3. Microarray and Gene Ontology Analysis. To explore the genomic expression heatmap comparing 37 genes in- reason behind the faster proliferation of HOXB9v over- volved in the apoptotic process (GO:0043069∼negative expressing cells, we performed microarray and gene on- regulation of programmed cell death) shows that apo- tology analyses using HOXB9n and HOXB9v-expressing ptosis is highly suppressed in HOXB9v expressing cells MCF7 cells. We chose 1056 genes as DEGs, and gene (Figure 7). ontology analysis showed that up-DEGs between HOXB9n and HOXB9v expressing cells presented sig- 4. Discussion nificant differences in pathways relevant to apoptosis suppression (GO:0060548∼negative regulation of cell In the present study, we identified a novel modified transcript death, GO:0043069∼negative regulation of programmed of HOXB9, in which a deletion in exon 1 causes a frameshift, cell death, and GO:0043066∼negative regulation of apo- formation of a stop codon, and truncation of protein coding. ptotic process) and steroid hormone response (GO:004 +is leads to a defect in the homeodomain and hexapeptide 8545∼response to steroid hormone, GO:0032870∼cellular regions, which are both crucial for HOX gene function. We response to hormone stimulus, and GO:0071383∼cellular confirmed that HOXB9v is widely present in human breast response to steroid hormone stimulus) (Table 1). +e cancer cell lines and clinical breast cancer samples, however not HOXB9 mRNA expression relative to control HOXB9 mRNA expression relative to control Journal of Oncology 7 1.6 1.6 1.4 1.4 1.2 1.2 1 1 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0 0 Day1 Day3 Day5 Day1 Day3 Day5 HOXB9n HXR9 HOXB9v HXR9 ∗ ∗ HOXB9n CXR9 p = 0.00015 HOXB9v CXR9 p = 0.00675 (a) (b) Figure 6: (a) HOXB9n-MCF7 cells and (b) HOXB9v-MCF7 cells were treated with HXR9 or with a control peptide, CXR9. HXR9 significantly inhibited the proliferation of both cell lines. Table 1: Gene ontology analysis by differential gene expression. Apoptosis was suppressed in HOXB9v-MCF7 cells compared with HOXB9n-MCF7 cells. Term P value Bonferroni − 07 Negative regulation of cell death 7.36E 0.002961 − 06 Response to steroid hormone 2.24E 0.008994 − 06 Cellular response to zinc ion 2.66E 0.01065 − 06 Negative regulation of programmed cell death 3.30E 0.013228 − 06 Cellular response to hormone stimulus 6.93E 0.027551 − 06 Negative regulation of apoptotic process 7.26E 0.028853 − 05 Cellular response to steroid hormone stimulus 1.15E 0.045436 − 05 Negative regulation of transcription, DNA-template 1.17E 0.045904 in normal gland samples. Detection of HOXB9v in MCF10A upstream from its stop codon; therefore, we infer it escapes may be attributed to the fact that this cell line is not karyo- from NMD. typically normal, while it maintains major characteristics of Nevertheless, the production of splice variants of normal breast epithelium [21]. We further confirmed that HOX genes lacking the homeodomain is likely a common HOXBv and the previously known HOXB9n were both pro- phenomenon. Murine Hoxb9, which shares sequence duced from variationless HOXB9 genomic DNA. Two distinct similarity with human HOXB9, is reported to generate a mRNA products from a single variationless genomic DNA may splice variant without a homeodomain. Other HOX indicate that HOXB9v is a novel splice variant of HOXB9. It genes, including human HOXA1, HOXB6, HOXA9, HOXA10, murine Hoxb6 and Meis1, and Xenopus may be assumed that HOXB9v has a typical splice site of GU as a 5′ donor site and AG as a 3′ acceptor site [22]; however, the XlHbox2, have also been reported to generate splice single-nucleotide transitions at 248 and 258 and the single- variants lacking homeodomains [24–28]. Interestingly, nucleotide deletion at 254 must be taken into account. Further HOXA9T, a homeodomain-less isoform of HOXA9, investigation and discussions are needed to decide whether which is structurally similar to HOXB9v, has been HOXB9v is a splice variant or an mRNA modified by a different demonstrated to act as an oncogene in leukemia without mechanism. Nonetheless, the modified transcript is certainly directly binding to DNA [24]. HOXB9 has been reported not a result of variations in genomic DNA. Additionally, to promote tumorigenesis in various types of cancers; HOXB9v results in a different stop codon site from that of however, no previous research on HOXB9 has shown a HOXB9n and may become a target of nonsense-mediated mRNA variant of HOXB9. Considering that the structure mRNA decay (NMD), an mRNA surveillance mechanism that of the HOXB9v protein differs from that of the previously studied HOXB9n protein and lacks important domains eliminates premature translation-termination codons. How- ever, NMD is known to be initiated if an exon-junction such as homeodomain and hexapeptide motif, its complex is present more than 50–55 bases downstream of the function is likely to differ from that of HOXB9n. Further stop codon [23]. HOXB9v forms its exon-junction complex study is required to re-evaluate the role of HOXB9 in 8 Journal of Oncology Scaled –1 –2 Sample IER3 CAV1 IL6ST EFNA1 TGFB3 BNIP3 PRDX5 SYCP2 PDCD4 SHH RGL2 CREB3L1 BCL6 CEACAM5 DDAH2 THBS1 MUC1 AR SOCS2 MET ANXA1 NR4A1 BAD NTSR1 POR FOXP1 ASCL1 DHRS2 CORO1A UCP2 AAMDC HIPK2 BNIP3L THOC6 TFAP2A TNFAIP3 BMP7 Sample HOXB9n HOXB9v Figure 7: Microarray data heatmap of HOXB9n-MCF7 and HOXB9v-MCF7. Genomic heatmap compares expression of 37 genes which are involved in suppression of apoptotic process (GO:0043069), with red and green color intensities indicating high and low expressions, respectively. Programmed cell death is suppressed in HOXB9v-MCF7 cells. cancer. HOXB9n and HOXB9v should be assessed have been reported to regulate apoptosis in cancer separately. [29, 30], and HOX-regulated apoptosis is a general In the growth assay, MCF7 and MDA-MB-468 cells mechanism used during development to maintain expressing HOXB9v at high levels presented more rapid metameric patterns [31, 32]. However, the role of HOXB9 in apoptosis is yet to be investigated. proliferation than did those expressing HOXB9v. It is meaningful that similar results were observed in two cell +e binding selectivity of HOX proteins is influenced by lines: one with hardly any expression of HOXB9n and cofactors, including members of the PBX, MEIS, and PREP HOXB9v (MCF7) and the other with high expression of families [2]. Additionally, HOX-PBX interactions involve a both (MDA-MB-468). Fostered proliferation by HOXB9v short HOX protein motif, the hexapeptide, located upstream may be attributed to the suppressed apoptosis observed in of the homeodomain [33]. HXR9 is an 18-amino acid our microarray studies. Gene ontology analysis of peptide and suppresses tumor proliferation by inhibiting HOXB9v-expressing cells compared to HOXB9n- HOX-PBX binding [19]. In our proliferation assay, HOXB9v expressing cells indicated significant upregulation of expressing cell lines were sensitive to the HOX-PBX in- pathways related to apoptosis suppression, further hibitor HXR9, even though HOXB9v lacks the hexapeptide. underscoring its role in this regard. Several HOX genes +is may be because HOXB9v, in consort with HOXB9n or Journal of Oncology 9 [6] O.-S. Kwon, E. Oh, J.-R. Park et al., “GalNAc-T14 promotes other HOX proteins, indirectly promotes HOX-PBX metastasis through Wnt dependent HOXB9 expression in binding. lung adenocarcinoma,” Oncotarget, vol. 6, no. 39, pp. 41916–41928, 2015. 5. Conclusions [7] T. Hayashida, F. Takahashi, N. Chiba et al., “HOXB9, a gene overexpressed in breast cancer, promotes tumorigenicity and We report a modified HOXB9 mRNA variant that results in a lung metastasis,” Proceedings of the National Academy of homeodomain defect. We confirmed its presence in human Sciences, vol. 107, no. 3, pp. 1100–1105, 2010. breast cancer cell lines and breast cancer clinical samples and [8] B. Shrestha, K. I. Ansari, A. Bhan, S. Kasiri, I. Hussain, and also revealed that HOXB9v may promote breast cancer S. S. Mandal, “Homeodomain-containing protein HOXB9 proliferation by suppression of apoptosis. Further research is regulates expression of growth and angiogenic factors, fa- warranted to analyze HOXB9v and HOXB9n individually cilitates tumor growth in vitro and is overexpressed in breast and re-evaluate the true role of HOXB9 in cancer. cancer tissue,” Ae FEBS Journal, vol. 279, no. 19, pp. 3715–3726, 2012. [9] Y. Hoshino, T. Hayashida, A. Hirata et al., “Bevacizumab Data Availability terminates homeobox B9-induced tumor proliferation by silencing microenvironmental communication,” Molecular +e HOXB9v sequence has been submitted to GenBank Cancer, vol. 13, no. 1–14, p. 102, 2014. under Accession No. LC466645. +e RNA sequence data set [10] S. Y. Wu, R. Rupaimoole, F. Shen et al., “A miR-192-EGR1- (GSE119937) used in public data reanalysis in the fastq HOXB9 regulatory network controls the angiogenic switch in format is available via SRA (SRP161704). Other datasets used cancer,” Nature Communications, vol. 7, no. 1, Article ID and/or analyzed during the current study are available from 11169, 2016. [11] L. Sha, L. Dong, L. Lv, X. Bai, and X. Ji, “HOXB9 promotes the corresponding author on reasonable request. epithelial-to-mesenchymal transition via transforming growth factor-β1 pathway in hepatocellular carcinoma cells,” Conflicts of Interest Clinical and Experimental Medicine, vol. 15, no. 1, pp. 55–64, +e authors declare that there are no conflicts of interest [12] H. Seki, T. Hayashida, H. Jinno et al., “HOXB9 expression regarding the publication of this paper. promoting tumor cell proliferation and angiogenesis is as- sociated with clinical outcomes in breast cancer patients,” Acknowledgments Annals of Surgical Oncology, vol. 19, no. 6, pp. 1831–1840, +e authors greatly thank Dr. Sasaki Takafumi, Keio University [13] A. Brock, S. Krause, H. Li et al., “Silencing HoxA1 by School of Medicine, for technical advice. +e authors also intraductal injection of siRNA lipidoid nanoparticles prevents thank Dr. Shunsuke Kato for scientific advice and Mr. Kazuya mammary tumor progression in mice,” Science Translational Medicine, vol. 6, no. 217, p. 217ra2, 2014. Takakuwa for help in bioinformatics analysis. +is work was [14] L. Fang, L. Xu, and L. Zou, “Overexpressed homeobox B9 supported by JSPS KAKENHI (Grant number 19K09057). regulates oncogenic activities by transforming growth factor- β1 in gliomas,” Biochemical and Biophysical Research Com- Supplementary Materials munications, vol. 446, no. 1, pp. 272–279, 2014. [15] C. Sun, C. Han, P. Wang, Y. Jin, Y. Sun, and L. Qu, “HOXB9 Figure S1: full sequences of HOXB9n and HOXB9v mRNA expression correlates with histological grade and prognosis in (black highlight indicates the homology between sequences). LSCC,” BioMed Research International, vol. 2017, Article ID Figure S2: sequence results of PCR products of breast cancer 3680305, 10 pages, 2017. cell-line derived genomic DNA. (Supplementary Materials) [16] J. Wan, H. Liu, Q. Feng, J. Liu, and L. Ming, “HOXB9 pro- motes endometrial cancer progression by targeting E2F3,” Cell Death & Disease, vol. 9, no. 5, p. 509, 2018. References [17] D. W. Huang, B. T. Sherman, Q. Tan et al., “DAVID bio- informatics resources: expanded annotation database and [1] T. R. J. Lappin, D. G. Grier, A. +ompson, and H. L. Halliday, novel algorithms to better extract biology from large gene “HOX genes: seductive science, mysterious mechanisms,” lists,” Nucleic Acids Research, vol. 35, no. 2, pp. W169–W175, Ulster Medical Journal, vol. 75, no. 1, pp. 23–31, 2006. [2] C. B. Moens and L. Selleri, “Hox cofactors in vertebrate [18] K. S. Keene, T. King, E. S. Hwang et al., “Molecular deter- development,” Developmental Biology, vol. 291, no. 2, minants of post-mastectomy breast cancer recurrence,” Npj pp. 193–206, 2006. Breast Cancer, vol. 4, no. 1, p. 34, 2018. [3] S. Bhatlekar, B. M. Fields, and B. M. Boman, “HOX genes and their role in the development of human cancers,” Journal of [19] R. Morgan, A. Boxall, K. J. Harrington et al., “Targeting the HOX/PBX dimer in breast cancer,” Breast Cancer Research Molecular Medicine, vol. 92, no. 8, pp. 811–823, 2014. [4] R. Vikram, K. S. M. Ramachandran, and K. S. M. Abdul, and Treatment, vol. 136, no. 2, pp. 389–398, 2012. [20] N. A. Laronde-leblanc and C. Wolberger, “Structure of “Functional significance of long non-coding RNAs in breast cancer,” Breast Cancer, vol. 21, no. 5, pp. 515–521, 2014. HoxA9 and Pbx1 bound to DNA: hox hexapeptide and DNA recognition anterior to posterior,” Genes and Development, [5] D. X. Nguyen, A. C. Chiang, X. H.-F. Zhang et al., “WNT/TCF signaling through LEF1 and HOXB9 mediates lung adeno- vol. 17, no. 16, pp. 2060–2072, 2003. [21] H. D. Soule, T. M. Maloney, S. R. Wolman et al., “Isolation carcinoma metastasis,” Journal of End-To-End-Testing, vol. 138, no. 1, pp. 51–62, 2009. and characterization of a spontaneously immortalized human 10 Journal of Oncology breast epithelial cell line, MCF-10,” Cancer Research, vol. 50, no. 18, pp. 6075–6086, 1990. [22] Y. Wang, J. Liu, B. Huang et al., “Mechanism of alternative splicing and its regulation,” Biomedical Reports, vol. 3, no. 2, pp. 152–158, 2015. [23] A. Nickless, J. M. Bailis, and Z. You, “Control of gene ex- pression through the nonsense-mediated RNA decay path- way,” Cell & Bioscience, vol. 7, no. 1, p. 26, 2017. [24] C. R. Stadler, N. Vegi, M. A. Mulaw et al., “+e leukemo- genicity of Hoxa9 depends on alternative splicing,” Leukemia, vol. 28, no. 9, pp. 1838–1843, 2014. [25] S. Fujimoto, K. Araki, O. Chisaka, M. Araki, K. Takagi, and K. I. Yamamura, “Analysis of the murine Hoxa-9 cDNA: an alternatively spliced transcript encodes a truncated protein lacking the homeodomain,” Gene, vol. 209, no. 1-2, pp. 77–85, [26] Y. S. Hong, S. Y. Kim, A. Bhattacharya, D. R. Pratt, W. K. Hong, and M. A. Tainsky, “Structure and function of the HOX A1 human homeobox gene cDNA,” Gene, vol. 159, no. 2, pp. 209–214, 1995. [27] W.-F. Shen, K. Detmer, T. A. Simonitch-Eason, C. Lawrence, and C. Largman, “Alternative splicing of the HOX 2.2 ho- meobox gene in human hematopoietic cells and murine embryonic and adult tissues,” Nucleic Acids Research, vol. 19, no. 3, pp. 539–545, 1991. [28] C. V. Wright, K. W. Cho, A. Fritz, E. M. Burglin, ¨ and E. M. De Robertis, “A Xenopus laevis gene encodes both homeobox-containing and homeobox-less transcripts,” Ae EMBO Journal, vol. 6, no. 13, pp. 4083–4094, 1987. [29] N. Shah and S. Sukumar, “+e Hox genes and their roles in oncogenesis,” Nature Reviews Cancer, vol. 10, no. 5, pp. 361–371, 2010. [30] H. Kocak, S. Ackermann, B. Hero et al., “Hox-C9 activates the intrinsic pathway of apoptosis and is associated with spon- taneous regression in neuroblastoma,” Cell Death & Disease, vol. 4, no. 4, Article ID e586, 2013. [31] R. Arya and K. White, “Cell death in development: signaling pathways and core mechanisms,” Seminars in Cell & Devel- opmental Biology, vol. 39, pp. 12–19, 2015. [32] K. Domsch, I. Papagiannouli, and I. Lohmann, “+e HOX- apoptosis regulatory interplay in development and disease,” Current Topics in Developmental Biology, vol. 114, pp. 121– 158, 2015. [33] P. M. J. In der Rieden, G. Mainguy, J. M. Woltering, and A. J. Durston, “Homeodomain to hexapeptide or PBC-in- teraction-domain distance: size apparently matters,” Trends in Genetics, vol. 20, no. 2, pp. 76–79, 2004. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Oncology Hindawi Publishing Corporation

Identification of a Modified HOXB9 mRNA in Breast Cancer

Download PDF
10 pages

Loading next page...
 
/lp/hindawi-publishing-corporation/identification-of-a-modified-hoxb9-mrna-in-breast-cancer-g0teVzaZEr

References (35)

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2020 Ayako Nakashoji et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1687-8450
eISSN
1687-8469
DOI
10.1155/2020/6065736
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Oncology Volume 2020, Article ID 6065736, 10 pages https://doi.org/10.1155/2020/6065736 Research Article Ayako Nakashoji, Tetsu Hayashida , Yuko Kawai, Masayuki Kikuchi, Rurina Watanuki, Takamichi Yokoe, Tomoko Seki, Maiko Takahashi, Kazuhiro Miyao, Shigeo Yamaguchi, and Yuko Kitagawa Department of Surgery, Keio University School of Medicine, Shinanomachi 35 Shinjuku-ku, Tokyo 160-0016, Japan Correspondence should be addressed to Tetsu Hayashida; tetsu@keio.jp Received 3 August 2019; Revised 29 December 2019; Accepted 7 January 2020; Published 13 February 2020 Academic Editor: Rossana Berardi Copyright © 2020 Ayako Nakashoji et al. +is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. First identified as a developmental gene, HOXB9 is also known to be involved in tumor biological processes, and its aberrant expression correlates with poor prognosis of various cancers. In this study, we isolated a homeodomain-less, novel HOXB9 variant (HOXB9v) from human breast cancer cell line-derived mRNA. We confirmed that the novel variant was produced from variationless HOXB9 genomic DNA. RT-PCR of mRNA isolated from clinical samples and reanalysis of publicly available RNA- seq data proved that the new transcript is frequently expressed in human breast cancer. Exogenous HOXB9v expression sig- nificantly enhanced the proliferation of breast cancer cells, and gene ontology analysis indicated that apoptotic signaling was suppressed in these cells. Considering that HOXB9v lacks key domains of homeobox proteins, its behavior could be completely different from that of the previously described variationless HOXB9. Because none of the previous studies on HOXB9 have considered the presence of HOXB9v, further research analyzing the two transcripts individually is warranted to re-evaluate the true role of HOXB9 in cancer. promotes metastasis by activating the WNT signaling 1. Introduction pathway [5, 6]. In breast cancer, the gene induces the ex- Homeobox (HOX) genes were initially characterized as pression of proangiogenic factors, increasing the cell mo- developmental genes, which code for transcription factors tility and supporting epithelial-mesenchymal transition that play critical roles in embryogenesis. Evolutionarily, they (EMT) [7, 8]. HOXB9 also promotes the growth of colon are highly conserved and share a high degree of homology, cancer by activating IL6 signaling, inducing the secretion of especially within the same paralog groups. All 39 mam- angiogenic factors and increasing proliferation of tumor malian HOX genes consist of two exons and a single intron. cells [9]. Similar observations are found in ovarian cancer +e homeobox domain, encoded in the second exon [1], and hepatocellular carcinoma [10, 11]. +us, HOXB9 acti- includes the DNA binding site. +e diverse and specific vates the WNT signaling pathway and enhances the ac- transcriptional activities of the HOX proteins often depend quisition of capabilities critical to the transformation of on key cofactors including PBX, MEIS, and PREP, which normal cells to cancer, including EMTand the growth of new interact with hexapeptide motifs of HOX proteins [2]. vasculature within the tumor microenvironment. HOX genes play key roles in both solid and hemato- HOXB9 is also directly associated with cancer-induced logical malignancies, including cancers of the colon, breast, patient mortality. +e duration of disease-free and overall prostate, lung, brain, thyroid, ovary, bladder, kidney, skin, survival of patients with HOXB9-positive breast cancer is and blood [3, 4]. HOXB9, the ninth paralog in the HOX-B significantly shorter compared with patients with HOXB9- cluster, is associated with the growth and progression of negative breast cancer [12]. Increased HOXB9 expression multiple cancers. In lung adenocarcinoma, HOXB9 significantly correlates with decreased overall survival for 2 Journal of Oncology patients with colorectal cancer [9]. Patients of laryngeal Minilys homogenizer (Bertin Instruments, Bretonneux, squamous cell carcinoma, hepatocellular carcinoma, glioma, France) using 2.4 mm metal beads prior to mRNA extrac- tion. Total RNA was converted to cDNA using the High- and endometrial cancer also present poor outcomes or tumor progression, resulting from aberrant HOXB9 ex- Capacity RNA-to-cDNA Kit (+ermo Fischer Scientific). pression [13–16]. +e recent elucidation of the critical and diverse roles 2.4. HOXB9 Cloning. HOXB9 transcripts and genomic HOXB9 plays in various cancers have led us to explore the DNA were amplified by PCR and subsequently cloned into mechanism of this gene’s role in cancer progression. In the the pME-HA vector (Lucigen, Middleton, WI, USA) using present study, we identified and characterized a HOXB9 the Expresso CMV Cloning & Expression system (Lucigen). variant (HOXB9v) of mRNA from human breast cancer cell +e primers used were as follows: sense 5′ GAAGGAGA- lines. +e sequence of HOXB9v largely differs from the TACCACCATGTCCATTTCTGGGACGCTTAGC 3′ and previously known HOXB9 normal transcript (HOXB9n), antisense 5′ GGGCACGTCATACGGATACTCTTTGCCC and we found it lacks some important domains of HOX TGCTCCTTATT 3′. genes. Based on these findings, we inferred its role and After transformation into Competent Quick DH5α cells function were different from HOXB9n. +is study aimed to (Toyobo, Osaka, Japan) and culturing in kanamycin-con- confirm the presence of HOXB9v in clinical breast cancer taining LB plates, at least 4 colonies were selected for each specimens and investigate the role of HOXB9v in breast sample. cancer progression. 2. Materials and Methods 2.5. Sequence Analysis. Sequences were analyzed using the BigDye Terminator V3.1 Cycle Sequencing kit (+ermo 2.1. Cell Culture. We cultured eight human breast cancer cell Fisher Scientific) and Applied Biosystems 3500 Genetic lines (MCF7, MDA-MB-231, MDA-MB-468, Hs578T, analyzer (+ermo Fisher Scientific). +e GENETYX-MAC HCC38, BT-474, BT-549, and SKBR3), a human colon Ver.19 software (GENETYX, Osaka, Japan) was used for cancer cell line (WiDR), and a mouse breast cancer cell line homology alignment. (4T1) in DMEM supplemented with 10% FBS (+ermo Fisher Scientific, Waltham, MA, USA) with the addition of antibiotic and antimycotic agent (antibiotic-antimycotic 2.6. RT-PCR Analysis. +e expressions of HOXB9n and HOXB9v were detected by RT-PCR using the following mixed stock solution, Nacalai Tesque, Inc., Kyoto, Japan). primers: sense, 5′ TGTCCATTTCTGGGACGCTT 3′; an- T47D cells were grown in RPMI1640 medium with 10% FBS tisense, 5′ CTACGGTCCCTGGTGAGGTA 3′. +e genomic and antibiotic and antimycotic agents. MCF10A cells were DNA of HOXB9 was detected by PCR using the following grown in DMEM/F12 (1 :1) medium with 5% horse serum primers: sense, 5′ CGAGAGAGCTGCAAGTCGAT 3′; (+ermo Fischer Scientific), 20 ng/mL EGF (PeproTech, antisense, 5′ CTGCCGTCCGTCTACCAC 3′. +e primers Rocky Hill, NJ, USA), 0.5 mg/mL hydrocortisone, 100 ng/ for genomic DNA were designed against exon 1 and the mL cholera toxin, 10 µg/mL insulin (all Sigma-Aldrich, St. intron region of the HOXB9 gene to ensure that the pair will Louis, MO, USA), and antibiotic and antimycotic agents. specifically amplify only genomic DNA and not cDNA Cells were maintained at 37 C in a humidified 5% CO derived from mRNA. +e conditions applied for amplifi- incubator. cation were as follows: 94 C for 1 minute, followed by 35 ° ° ° cycles at 95 C for 5 seconds, 55 C for 5 seconds, and 72 C for 2.2. Patients and Samples. Clinical specimens of human 5 seconds, and run on the Life ECO thermal cycler breast cancer (n � 14) were collected from patients with (Hangzhou Bioer Technology, Hangzhou, China) using the primary operable breast cancer who underwent total or SapphireAmp Fast PCR Master Mix (Takara Bio, Shiga, partial mastectomy between July and November 2018 in Japan) or KOD-Plus-Neo (Toyobo). Keio University Hospital (Tokyo, Japan). Patient-matched healthy breast epithelium samples (n � 6) were collected from healthy breast tissue from patients who underwent 2.7. Public Data Reanalysis. +e RNA sequence data set (GSE119937) in FASTQ format was downloaded via SRA total mastectomy. Ethics approval for the present study was provided by the Ethics Committee at the Keio University (SRP161704) using the SRA Toolkit (version 2.3.4-2). RNA- seq reads were aligned by STAR (version 2.6.1b) against the School of Medicine (approval number: 20180090), and the study was performed in accordance with the provisions of hg38 reference genome. All reads mapped on the HOXB9 the Declaration of Helsinki (as revised in Fortaleza, Brazil, gene were visually confirmed by taking snapshots in IGV October 2013). All included patients gave informed consent. (version 2.4.15). 2.3. mRNA and Genomic DNA Extraction from Cell Lines and 2.8. Establishment of Stable MCF7 Cell Lines Overexpressing Clinical Specimens. Total RNA and genomic DNA were HOXB9n or HOXB9v. +e HOXB9n and HOXB9v se- extracted using the RNeasy Mini Kit and QIAmp DNA Mini quences were amplified by PCR and subsequently cloned Kit (Qiagen, Hilden, Germany) following the manufacturer’s into the pBiT3.1-N [CMV/HiBiT/Blast] expression vector instructions. Clinical specimens were homogenized using a (Promega, Tokyo, Japan) at the Xhol and BamHI sites using Journal of Oncology 3 Ligation high Ver.2 (Toyobo). +e primers used for am- results were analyzed by TAC 4.0 software (+ermo Fisher plification were as follows: sense, 3′ ATACCTCGAGG Scientific). Genes having a false discovery rate (FDR) under 0.05 and upregulated in HOXB9v samples were considered TCCATTTCTGG 5′; antisense, 3′ CACGTCATACGGAT CCTCTTTG 5′. MCF7 cells were transfected with the as upregulated differentially expressed genes (up-DEGs). HiBiT-tagged HOXB9n or HOXB9v vector using the Viafect DAVID Bioinformatics Resources 6.8 [17] was used for gene Transfection reagent (Promega) and selected with 10 µg/mL ontology and pathway analysis of DEGs. R ver. 3.5.0 software of blasticidin for at least 4 weeks. +e expression of HiBiT- was used to draw the heatmap of 37 genes included in GO: tagged HOXB9n and HOXB9v proteins was detected using 0043069 (negative regulation of programmed cell death). the Nano-Glo HiBiT Blotting System (Promega) as per the manufacturer’s instructions. 3. Results 3.1. Identification of a Novel Transcript HOXB9v, Which Lacks 2.9. Transient Overexpression of HOXB9n and HOXB9v in Important Domains of HOX Gene. We isolated total RNA MDA-MB-468 Cells. MDA-MB-468 cells were transfected from human breast cancer cell lines (MCF7, T47D, MDA- with aforementioned HiBiT-tagged HOXB9n or HOXB9v MB-231, MDA-MB-468, HCC38, BT-474, BT-549, Hs578T, vector using the jetPRIME (Polyplus-transfection, Illkirch, and SKBR3), a normal human mammary gland cell line France) as per the manufacturer’s instructions. (MCF10A), and a human colon cancer cell line (WiDR). We cloned and sequenced the HOXB9 gene from the nucleic 2.10. Quantitative Real-Time PCR. Quantitative real-time acids derived from these cell lines. We isolated a novel PCR was run on ViiA7 (+ermo Fisher Scientific) using Fast variant of the HOXB9 transcript from the MCF7, T47D, SYBR Green Master Mix (+ermo Fisher Scientific). Pre- MDA-MB-231, MDA-MB-468, Hs578T, HCC38, and incubation was performed for 20 seconds at 95 C and MCF10A cell lines. +e sequence homology of the new amplification for 41 cycles (1 second of denaturation at 95 C transcript is shown in Figures 1(a) and S1. and 20 seconds of annealing and extension at 60 C), followed We will refer to the new truncated transcript as HOXB9v, by melt-curve analysis. GAPDH served as an internal con- to distinguish from the full-length HOXB9 transcript trol, and QuantStudio Real-Time PCR Software v1.2 (HOXB9n). A 100-base deletion in exon 1 in the new transcript leads to a frameshift and the formation of a stop (+ermo Fisher Scientific) was used for quantification. +e relative standard curve method was used for linear regres- codon (TAG), which truncates the protein coding at AA167. sion analysis of unknown samples, and data are presented as Figure 1(b) is a schematic diagram of HOXB9 transcripts fold change between samples. +e primers used were as showing the exons and the deleted lesion. +e encoded follows: HOXB9: sense, 5′ CGGTGGCTGTCGTGAAATT protein will therefore possibly lack the homeobox domain, 3′; antisense, 5′ CGAGACAATCACCCCCAAAG 3′; GAP DNA binding domain, and the hexapeptide motif, a major DH: sense, 5′ATCATCCCTGCCTCTACTGG 3′; antisense, player in cofactor interactions (Figure 1(c)). +e HOXB9v 5′ TTTCTAGACGGCAGGTCAGGT 3′. sequence has been submitted to GenBank under Accession No. LC466645. We next performed PCR analyses to verify the presence 2.11. Cell Proliferation Assay. Cell proliferation in three- of the HOXB9v transcript and identify genomic DNA dimensional (3D) culture was measured using the 24-well variations of the HOXB9 gene in human breast cancer cell Bio-Assembler kit and NanoShuttle-PL (Greiner Bio-One, lines. +e primer target region included the deletion site of Kremsmunster, ¨ Austria). Further, 20,000 cells/well were HOXB9v and the amplicon for HOXB9n was 643 bp and incubated for 48 hours before taking photomicrographs. 543 bp for HOXB9v. We detected both HOXB9n and Cell proliferation in flat culture was measured using Cell HOXB9v transcripts in breast cancer cell lines (Figure 2(a)). Count Reagent SF (Nacalai Tesque). Briefly, 5000 cells/well +e PCR products were sequenced and were confirmed that in 96-well microtiter plates (Sumilon, Sumitomo Bakelite, each band corresponded to the exact sequence of the Tokyo Japan) were incubated for 5 days or after 24 hour HOXB9n or HOXB9v (Figure 2(b)). By PCR of breast cancer continuous exposure to either 30 nM HXR9 or 30 nM CXR9 cell lines’ genomic DNA, no bands indicative of genomic for 4 days. +e absorbance in the wells was measured on days DNA variations in HOXB9 were detected. +e amplicon of 1, 3, and 5 using a Sunrise Rainbow-RC (TECAN, variation-less HOXB9 genomic DNA was 569 bp Mannedorf, ¨ Switzerland) microplate spectrophotometer at (Figure 2(c)). Sequencing of the PCR products confirmed 450 nm, using 600 nm as reference. +e HXR9 that they had no variations or deletions in genomic DNA (WYPWMKKHHRRRRRRRRR-) and control CXR9 (Figure S2). +ese findings show that the two transcripts (WYPAKKHHRRRRRRRRR) peptides were synthesized by (HOXB9n and HOXB9v) were produced from variation-less Eurofins Genomics K. K. (Tokyo, Japan). HOXB9 genomic DNA. To determine the presence of HOXB9v transcripts in 2.12. Microarray and Differential Expression Analyses. human breast cancer samples, we performed PCR analysis. Total RNA was isolated using the RNeasy Mini kit (Qiagen). HOXB9v was commonly detected from clinical breast cancer Microarray was performed using the human Clariom S assay samples (Figure 3(a)), regardless of their hormone receptor (+ermo Fisher Scientific) by the GeneChip Scanner 3000 and HER2 status. However, HOXB9v was not detected in 7G system (Affymetrix, Santa Clara, CA, USA), and the normal mammary gland samples (Figure 3(b)). 4 Journal of Oncology HOXB9 gDNA 445 TACATCC AGCCCCAGGGCGTCCCGCCGGCCGAGAGCAGGTACCTCCGCACCTGGCTGGAGCCGGCGCCGCGCGGCGAAGCGGCCCCGGGGCAGGGCCAGGCGGCGGTGAAGGC GGAGCCGCTGCTGGGCG 574 HOXB9n mRNA 241 TACATCC AGCCCCAGGGCGTCCCGCCGGCCGAGAGCAGGTACCTCCGCACCTGGCTGGAG CCGGCGCCGCGCGGCGAAGCGGCCCCGGGGCAGGGCCAGGCGGCGGTGAAGGC GGAGCCGCTGCTGGGCG 370 HOXB9v mRNA 241 TACATCC TGCCCC – GGGG – CCGCTGCTGGGCG 270 (a) Exon I Exon II ATG TAA HOXB9n 1 205 721 722 955 2701 ATG TAG HOXB9v 1 205 462 562 721 722 2701 (b) Hexapeptide & linker HOXB9n N-terminal flexible region Homeodomain Nucleotide 1 528 555 753 AA 1 176 185 251 Frameshift HOXB9v Nucleotide 1 257 357 501 653 AA 1 85fs 167 (c) Figure 1: Structure and sequence of HOXB9n and HOXB9v. (a) Sequences of genomic DNA HOXB9, mRNA HOXB9n, and mRNA HOXB9v (black and grey highlighting indicates homology between sequences). (b) Schematic diagram of HOXB9n (upper) and HOXB9v (lower) transcripts showing the exons, splicing regions, and the deleted region. In HOXB9v, a 100 bp deletion in exon 1 leads to a frameshift and a stop codon formation (TAA). (c) Protein structure of HOXB9n and HOXB9v. Transcription from the start codon (ATG) to the stop codon (TAA) results in the translation of a full-length HOXB9n protein (upper). +e 100 bp deletion in HOXB9v (shown in black) leads to a frameshift from AA85 (shown in grey) and truncation of protein coding by a stop codon (TAG) at AA167, which results in HOXB9v protein without hexapeptide and homeodomain. Stop codon. MDA- MDA- MCF7 T47D Hs578T SKBR3 MCF10A HOXB9n HOXB9v MB-231 MB-468 700bp –– –– 700bp 600bp –– –– 600bp 500bp –– –– 500bp (a) (b) MDA- MDA- MCF7 T47D Hs578T MCF10A MB-231 MB-468 600bp –– 500bp –– 400bp –– (c) Figure 2: Detection of mRNA and genomic DNA of HOXB9 in cell lines. (a) Both HOXB9n and HOXB9v transcripts were detected in breast cancer cell line mRNA (cDNA). (b) Sequencing confirmation of PCR product of Figure 2(a); SKBR3 (HOXB9n, upper column) and MCF10A (HOXB9v, lower column). (c) No genomic variation was detected in genomic DNA of breast cancer cell line genomic DNA. HOXB9v transcripts are commonly found in human breast cancer specimens. Journal of Oncology 5 Sample# 1 2 3 45678 9 10 11 12 13 14 HR/HER2 +/+ +/– +/– +/– +/– –/– +/– –/– +/– +/– –/– +/– +/– –/– (a) Sample# 125 11 12 14 HOXB9n HOXB9v (b) Figure 3: Detection of HOXB9v in clinical samples. (a) HOXB9n and HOXB9v were detected in breast cancer clinical samples. (b) HOXB9v was not detected in normal mammary gland samples. Hormone receptor status and HER2 status of each cancer sample are shown beneath each sample number. We assigned the same sample number if cancer and normal gland samples were acquired from the same patient. SRR7826774 T3 N1 IDC (primary) ER – PgR – HER2–, ki67 60% SRR7826783 T1cN0 IDC (primary) ER + PgR – HER2–, ki67 5% SRR7826861 T1aN1 IDC (primary) ER + PgR + HER2–, ki67 0% Exon II Intron Exon I 3′ 5′ ATG HOXB9n TAA HOXB9 HOXB9v TAG ATG Figure 4: NGS data indicates the presence of HOXB9v in breast cancer. Showing RNA sequence breast cancer sample data (SRR782677) mapped on HOXB9 exon 1. +e blue band on the bottom shows the coding region (the thick part) and the 5′ UTR (the thin part) of exon 1. +ere is a region where data are sparsely mapped (indicated with the red arrow), which matches the deletion region in HOXBv. HOXB9v overexpressing MCF7 and MDA-MB-468 cells proliferated faster compared with HOXB9n overexpressing cells. +us, to further confirm the presence of HOXB9v in transiently overexpressed HOXB9n or HOXB9v in MDA- human breast cancer samples, we reanalyzed a publicly MB-468 cells, and the cell proliferation assay showed that available breast cancer RNA sequence data set HOXB9v overexpression increased MDA-MB468 cell (GSE119937) [18] and mapped the reads onto the HOXB9 growth (Figures 5(e) and 5(f)). gene sequence. We identified a region in exon 1 where the number of mapped reads was low in numerous samples; 3.2. HXR9 and CXR9 Treatment. HXR9, an 18-amino acid this region matched the deletion region of HOXB9v (Figure 4). peptide, competently inhibits the hexapeptide motif of HOX proteins and prevents HOX-PBX binding [19]. To determine the role of HOXB9v in breast cancer, we established stable MCF7 cell lines overexpressing HOXB9n HOXB9n and HOXB9v expressing MCF7 cells were treated with HXR9 or a control peptide, CXR9 or HOXB9v. Gene and protein expressions of HOXB9n and HOXB9v were verified in both cell lines (Figures 5(a) and (Figures 6(a) and 6(b)). Although HOXB9v lacks the hexapeptide motif, which is known to interact with PBX 5(b)). Cell proliferation assays in both 3D culture and flat culture showed that HOXB9v overexpression increased proteins [20], HXR9 significantly inhibited proliferation of both cell lines. MCF7 cell growth (Figures 5(c) and 5(d)). We also 6 Journal of Oncology log 10 HOXB9v HOXB9n ∗ (μm) 35kDa 1834.8 1601.0 25kDa 17kDa HOXB9v HOXB9n 0 HOXB9v HOXB9n p = 1.7247E – 06 (a) (b) (c) log 10 1.4 100000 0.5 1.2 10000 0.4 1000 0.3 0.8 ∗ 0.6 100 0.2 0.4 10 0.1 0.2 0 0 HOXB9v HOXB9n Day1 Day3 Day5 Day1 Day3 Day5 HOXB9n HOXB9n ∗ ∗ p = 0.0134 p = 4.40E – 07 HOXB9v HOXB9v (d) (e) (f) Figure 5: (a–d) Confirmation and proliferation of stable MCF7 cell lines that overexpress HOXB9n or HOXB9v. Levels of HOXB9n and HOXB9v (a) mRNA and (b) protein in stable cell lines. Proliferation of the HOXB9n and HOXB9v stable cell lines using a (c) 3D cell culture model and a (d) 2D cell culture model. (e-f) Confirmation and proliferation of MDA-MB-468 cells with transient overexpression of HOXB9n or HOXB9v. (e) Levels of HOXB9n and HOXB9v mRNA expression. (f) Proliferation of HOXB9n and HOXB9v overexpressing cells by 2D cell culture model. 3.3. Microarray and Gene Ontology Analysis. To explore the genomic expression heatmap comparing 37 genes in- reason behind the faster proliferation of HOXB9v over- volved in the apoptotic process (GO:0043069∼negative expressing cells, we performed microarray and gene on- regulation of programmed cell death) shows that apo- tology analyses using HOXB9n and HOXB9v-expressing ptosis is highly suppressed in HOXB9v expressing cells MCF7 cells. We chose 1056 genes as DEGs, and gene (Figure 7). ontology analysis showed that up-DEGs between HOXB9n and HOXB9v expressing cells presented sig- 4. Discussion nificant differences in pathways relevant to apoptosis suppression (GO:0060548∼negative regulation of cell In the present study, we identified a novel modified transcript death, GO:0043069∼negative regulation of programmed of HOXB9, in which a deletion in exon 1 causes a frameshift, cell death, and GO:0043066∼negative regulation of apo- formation of a stop codon, and truncation of protein coding. ptotic process) and steroid hormone response (GO:004 +is leads to a defect in the homeodomain and hexapeptide 8545∼response to steroid hormone, GO:0032870∼cellular regions, which are both crucial for HOX gene function. We response to hormone stimulus, and GO:0071383∼cellular confirmed that HOXB9v is widely present in human breast response to steroid hormone stimulus) (Table 1). +e cancer cell lines and clinical breast cancer samples, however not HOXB9 mRNA expression relative to control HOXB9 mRNA expression relative to control Journal of Oncology 7 1.6 1.6 1.4 1.4 1.2 1.2 1 1 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0 0 Day1 Day3 Day5 Day1 Day3 Day5 HOXB9n HXR9 HOXB9v HXR9 ∗ ∗ HOXB9n CXR9 p = 0.00015 HOXB9v CXR9 p = 0.00675 (a) (b) Figure 6: (a) HOXB9n-MCF7 cells and (b) HOXB9v-MCF7 cells were treated with HXR9 or with a control peptide, CXR9. HXR9 significantly inhibited the proliferation of both cell lines. Table 1: Gene ontology analysis by differential gene expression. Apoptosis was suppressed in HOXB9v-MCF7 cells compared with HOXB9n-MCF7 cells. Term P value Bonferroni − 07 Negative regulation of cell death 7.36E 0.002961 − 06 Response to steroid hormone 2.24E 0.008994 − 06 Cellular response to zinc ion 2.66E 0.01065 − 06 Negative regulation of programmed cell death 3.30E 0.013228 − 06 Cellular response to hormone stimulus 6.93E 0.027551 − 06 Negative regulation of apoptotic process 7.26E 0.028853 − 05 Cellular response to steroid hormone stimulus 1.15E 0.045436 − 05 Negative regulation of transcription, DNA-template 1.17E 0.045904 in normal gland samples. Detection of HOXB9v in MCF10A upstream from its stop codon; therefore, we infer it escapes may be attributed to the fact that this cell line is not karyo- from NMD. typically normal, while it maintains major characteristics of Nevertheless, the production of splice variants of normal breast epithelium [21]. We further confirmed that HOX genes lacking the homeodomain is likely a common HOXBv and the previously known HOXB9n were both pro- phenomenon. Murine Hoxb9, which shares sequence duced from variationless HOXB9 genomic DNA. Two distinct similarity with human HOXB9, is reported to generate a mRNA products from a single variationless genomic DNA may splice variant without a homeodomain. Other HOX indicate that HOXB9v is a novel splice variant of HOXB9. It genes, including human HOXA1, HOXB6, HOXA9, HOXA10, murine Hoxb6 and Meis1, and Xenopus may be assumed that HOXB9v has a typical splice site of GU as a 5′ donor site and AG as a 3′ acceptor site [22]; however, the XlHbox2, have also been reported to generate splice single-nucleotide transitions at 248 and 258 and the single- variants lacking homeodomains [24–28]. Interestingly, nucleotide deletion at 254 must be taken into account. Further HOXA9T, a homeodomain-less isoform of HOXA9, investigation and discussions are needed to decide whether which is structurally similar to HOXB9v, has been HOXB9v is a splice variant or an mRNA modified by a different demonstrated to act as an oncogene in leukemia without mechanism. Nonetheless, the modified transcript is certainly directly binding to DNA [24]. HOXB9 has been reported not a result of variations in genomic DNA. Additionally, to promote tumorigenesis in various types of cancers; HOXB9v results in a different stop codon site from that of however, no previous research on HOXB9 has shown a HOXB9n and may become a target of nonsense-mediated mRNA variant of HOXB9. Considering that the structure mRNA decay (NMD), an mRNA surveillance mechanism that of the HOXB9v protein differs from that of the previously studied HOXB9n protein and lacks important domains eliminates premature translation-termination codons. How- ever, NMD is known to be initiated if an exon-junction such as homeodomain and hexapeptide motif, its complex is present more than 50–55 bases downstream of the function is likely to differ from that of HOXB9n. Further stop codon [23]. HOXB9v forms its exon-junction complex study is required to re-evaluate the role of HOXB9 in 8 Journal of Oncology Scaled –1 –2 Sample IER3 CAV1 IL6ST EFNA1 TGFB3 BNIP3 PRDX5 SYCP2 PDCD4 SHH RGL2 CREB3L1 BCL6 CEACAM5 DDAH2 THBS1 MUC1 AR SOCS2 MET ANXA1 NR4A1 BAD NTSR1 POR FOXP1 ASCL1 DHRS2 CORO1A UCP2 AAMDC HIPK2 BNIP3L THOC6 TFAP2A TNFAIP3 BMP7 Sample HOXB9n HOXB9v Figure 7: Microarray data heatmap of HOXB9n-MCF7 and HOXB9v-MCF7. Genomic heatmap compares expression of 37 genes which are involved in suppression of apoptotic process (GO:0043069), with red and green color intensities indicating high and low expressions, respectively. Programmed cell death is suppressed in HOXB9v-MCF7 cells. cancer. HOXB9n and HOXB9v should be assessed have been reported to regulate apoptosis in cancer separately. [29, 30], and HOX-regulated apoptosis is a general In the growth assay, MCF7 and MDA-MB-468 cells mechanism used during development to maintain expressing HOXB9v at high levels presented more rapid metameric patterns [31, 32]. However, the role of HOXB9 in apoptosis is yet to be investigated. proliferation than did those expressing HOXB9v. It is meaningful that similar results were observed in two cell +e binding selectivity of HOX proteins is influenced by lines: one with hardly any expression of HOXB9n and cofactors, including members of the PBX, MEIS, and PREP HOXB9v (MCF7) and the other with high expression of families [2]. Additionally, HOX-PBX interactions involve a both (MDA-MB-468). Fostered proliferation by HOXB9v short HOX protein motif, the hexapeptide, located upstream may be attributed to the suppressed apoptosis observed in of the homeodomain [33]. HXR9 is an 18-amino acid our microarray studies. Gene ontology analysis of peptide and suppresses tumor proliferation by inhibiting HOXB9v-expressing cells compared to HOXB9n- HOX-PBX binding [19]. In our proliferation assay, HOXB9v expressing cells indicated significant upregulation of expressing cell lines were sensitive to the HOX-PBX in- pathways related to apoptosis suppression, further hibitor HXR9, even though HOXB9v lacks the hexapeptide. underscoring its role in this regard. Several HOX genes +is may be because HOXB9v, in consort with HOXB9n or Journal of Oncology 9 [6] O.-S. Kwon, E. Oh, J.-R. Park et al., “GalNAc-T14 promotes other HOX proteins, indirectly promotes HOX-PBX metastasis through Wnt dependent HOXB9 expression in binding. lung adenocarcinoma,” Oncotarget, vol. 6, no. 39, pp. 41916–41928, 2015. 5. Conclusions [7] T. Hayashida, F. Takahashi, N. Chiba et al., “HOXB9, a gene overexpressed in breast cancer, promotes tumorigenicity and We report a modified HOXB9 mRNA variant that results in a lung metastasis,” Proceedings of the National Academy of homeodomain defect. We confirmed its presence in human Sciences, vol. 107, no. 3, pp. 1100–1105, 2010. breast cancer cell lines and breast cancer clinical samples and [8] B. Shrestha, K. I. Ansari, A. Bhan, S. Kasiri, I. Hussain, and also revealed that HOXB9v may promote breast cancer S. S. Mandal, “Homeodomain-containing protein HOXB9 proliferation by suppression of apoptosis. Further research is regulates expression of growth and angiogenic factors, fa- warranted to analyze HOXB9v and HOXB9n individually cilitates tumor growth in vitro and is overexpressed in breast and re-evaluate the true role of HOXB9 in cancer. cancer tissue,” Ae FEBS Journal, vol. 279, no. 19, pp. 3715–3726, 2012. [9] Y. Hoshino, T. Hayashida, A. Hirata et al., “Bevacizumab Data Availability terminates homeobox B9-induced tumor proliferation by silencing microenvironmental communication,” Molecular +e HOXB9v sequence has been submitted to GenBank Cancer, vol. 13, no. 1–14, p. 102, 2014. under Accession No. LC466645. +e RNA sequence data set [10] S. Y. Wu, R. Rupaimoole, F. Shen et al., “A miR-192-EGR1- (GSE119937) used in public data reanalysis in the fastq HOXB9 regulatory network controls the angiogenic switch in format is available via SRA (SRP161704). Other datasets used cancer,” Nature Communications, vol. 7, no. 1, Article ID and/or analyzed during the current study are available from 11169, 2016. [11] L. Sha, L. Dong, L. Lv, X. Bai, and X. Ji, “HOXB9 promotes the corresponding author on reasonable request. epithelial-to-mesenchymal transition via transforming growth factor-β1 pathway in hepatocellular carcinoma cells,” Conflicts of Interest Clinical and Experimental Medicine, vol. 15, no. 1, pp. 55–64, +e authors declare that there are no conflicts of interest [12] H. Seki, T. Hayashida, H. Jinno et al., “HOXB9 expression regarding the publication of this paper. promoting tumor cell proliferation and angiogenesis is as- sociated with clinical outcomes in breast cancer patients,” Acknowledgments Annals of Surgical Oncology, vol. 19, no. 6, pp. 1831–1840, +e authors greatly thank Dr. Sasaki Takafumi, Keio University [13] A. Brock, S. Krause, H. Li et al., “Silencing HoxA1 by School of Medicine, for technical advice. +e authors also intraductal injection of siRNA lipidoid nanoparticles prevents thank Dr. Shunsuke Kato for scientific advice and Mr. Kazuya mammary tumor progression in mice,” Science Translational Medicine, vol. 6, no. 217, p. 217ra2, 2014. Takakuwa for help in bioinformatics analysis. +is work was [14] L. Fang, L. Xu, and L. Zou, “Overexpressed homeobox B9 supported by JSPS KAKENHI (Grant number 19K09057). regulates oncogenic activities by transforming growth factor- β1 in gliomas,” Biochemical and Biophysical Research Com- Supplementary Materials munications, vol. 446, no. 1, pp. 272–279, 2014. [15] C. Sun, C. Han, P. Wang, Y. Jin, Y. Sun, and L. Qu, “HOXB9 Figure S1: full sequences of HOXB9n and HOXB9v mRNA expression correlates with histological grade and prognosis in (black highlight indicates the homology between sequences). LSCC,” BioMed Research International, vol. 2017, Article ID Figure S2: sequence results of PCR products of breast cancer 3680305, 10 pages, 2017. cell-line derived genomic DNA. (Supplementary Materials) [16] J. Wan, H. Liu, Q. Feng, J. Liu, and L. Ming, “HOXB9 pro- motes endometrial cancer progression by targeting E2F3,” Cell Death & Disease, vol. 9, no. 5, p. 509, 2018. References [17] D. W. Huang, B. T. Sherman, Q. Tan et al., “DAVID bio- informatics resources: expanded annotation database and [1] T. R. J. Lappin, D. G. Grier, A. +ompson, and H. L. Halliday, novel algorithms to better extract biology from large gene “HOX genes: seductive science, mysterious mechanisms,” lists,” Nucleic Acids Research, vol. 35, no. 2, pp. W169–W175, Ulster Medical Journal, vol. 75, no. 1, pp. 23–31, 2006. [2] C. B. Moens and L. Selleri, “Hox cofactors in vertebrate [18] K. S. Keene, T. King, E. S. Hwang et al., “Molecular deter- development,” Developmental Biology, vol. 291, no. 2, minants of post-mastectomy breast cancer recurrence,” Npj pp. 193–206, 2006. Breast Cancer, vol. 4, no. 1, p. 34, 2018. [3] S. Bhatlekar, B. M. Fields, and B. M. Boman, “HOX genes and their role in the development of human cancers,” Journal of [19] R. Morgan, A. Boxall, K. J. Harrington et al., “Targeting the HOX/PBX dimer in breast cancer,” Breast Cancer Research Molecular Medicine, vol. 92, no. 8, pp. 811–823, 2014. [4] R. Vikram, K. S. M. Ramachandran, and K. S. M. Abdul, and Treatment, vol. 136, no. 2, pp. 389–398, 2012. [20] N. A. Laronde-leblanc and C. Wolberger, “Structure of “Functional significance of long non-coding RNAs in breast cancer,” Breast Cancer, vol. 21, no. 5, pp. 515–521, 2014. HoxA9 and Pbx1 bound to DNA: hox hexapeptide and DNA recognition anterior to posterior,” Genes and Development, [5] D. X. Nguyen, A. C. Chiang, X. H.-F. Zhang et al., “WNT/TCF signaling through LEF1 and HOXB9 mediates lung adeno- vol. 17, no. 16, pp. 2060–2072, 2003. [21] H. D. Soule, T. M. Maloney, S. R. Wolman et al., “Isolation carcinoma metastasis,” Journal of End-To-End-Testing, vol. 138, no. 1, pp. 51–62, 2009. and characterization of a spontaneously immortalized human 10 Journal of Oncology breast epithelial cell line, MCF-10,” Cancer Research, vol. 50, no. 18, pp. 6075–6086, 1990. [22] Y. Wang, J. Liu, B. Huang et al., “Mechanism of alternative splicing and its regulation,” Biomedical Reports, vol. 3, no. 2, pp. 152–158, 2015. [23] A. Nickless, J. M. Bailis, and Z. You, “Control of gene ex- pression through the nonsense-mediated RNA decay path- way,” Cell & Bioscience, vol. 7, no. 1, p. 26, 2017. [24] C. R. Stadler, N. Vegi, M. A. Mulaw et al., “+e leukemo- genicity of Hoxa9 depends on alternative splicing,” Leukemia, vol. 28, no. 9, pp. 1838–1843, 2014. [25] S. Fujimoto, K. Araki, O. Chisaka, M. Araki, K. Takagi, and K. I. Yamamura, “Analysis of the murine Hoxa-9 cDNA: an alternatively spliced transcript encodes a truncated protein lacking the homeodomain,” Gene, vol. 209, no. 1-2, pp. 77–85, [26] Y. S. Hong, S. Y. Kim, A. Bhattacharya, D. R. Pratt, W. K. Hong, and M. A. Tainsky, “Structure and function of the HOX A1 human homeobox gene cDNA,” Gene, vol. 159, no. 2, pp. 209–214, 1995. [27] W.-F. Shen, K. Detmer, T. A. Simonitch-Eason, C. Lawrence, and C. Largman, “Alternative splicing of the HOX 2.2 ho- meobox gene in human hematopoietic cells and murine embryonic and adult tissues,” Nucleic Acids Research, vol. 19, no. 3, pp. 539–545, 1991. [28] C. V. Wright, K. W. Cho, A. Fritz, E. M. Burglin, ¨ and E. M. De Robertis, “A Xenopus laevis gene encodes both homeobox-containing and homeobox-less transcripts,” Ae EMBO Journal, vol. 6, no. 13, pp. 4083–4094, 1987. [29] N. Shah and S. Sukumar, “+e Hox genes and their roles in oncogenesis,” Nature Reviews Cancer, vol. 10, no. 5, pp. 361–371, 2010. [30] H. Kocak, S. Ackermann, B. Hero et al., “Hox-C9 activates the intrinsic pathway of apoptosis and is associated with spon- taneous regression in neuroblastoma,” Cell Death & Disease, vol. 4, no. 4, Article ID e586, 2013. [31] R. Arya and K. White, “Cell death in development: signaling pathways and core mechanisms,” Seminars in Cell & Devel- opmental Biology, vol. 39, pp. 12–19, 2015. [32] K. Domsch, I. Papagiannouli, and I. Lohmann, “+e HOX- apoptosis regulatory interplay in development and disease,” Current Topics in Developmental Biology, vol. 114, pp. 121– 158, 2015. [33] P. M. J. In der Rieden, G. Mainguy, J. M. Woltering, and A. J. Durston, “Homeodomain to hexapeptide or PBC-in- teraction-domain distance: size apparently matters,” Trends in Genetics, vol. 20, no. 2, pp. 76–79, 2004.

Journal

Journal of Oncology – Hindawi Publishing Corporation

Published: Feb 13, 2020

There are no references for this article.