Treatment for sex hormone-insensitive GREB1-positive tumors

Abstract

The purpose of the present invention is to provide a novel tumor treatment agent. In GREB1-positive tumors that do not exhibit sex hormone sensitivity, GREB1 is a downstream target gene of the Wnt / β-catenin signal. Through the use as an active ingredient of a substance that can inhibit GREB1 expression, the growth of GREB1-positive tumor cells that do not exhibit sex hormone sensitivity can be effectively inhibited and such a tumor can be treated.

Description

[Technical Field]

[0001] The present invention relates to a therapeutic agent for sex hormone-insensitive GREB1 (Growth regulation by estrogen in breast cancer)-positive tumors (GREB1-positive tumors that are not sensitive to sex hormones), and further to a method for detecting hepatoblastoma, hepatocellular carcinoma, malignant melanoma, and neuroblastoma using the expression level of GREB1 as an indicator. [Background technology]

[0002] Hepatoblastoma is a malignant tumor of the liver in children. It often develops by the age of three and accounts for approximately 90% of malignant liver tumors in children under four years of age. In Japan, hepatoblastoma affects 40-30 people per year, and globally, the incidence rate is approximately one in one million, making it a rare cancer. The etiology of hepatoblastoma is unknown, but nearly 70% of cases have deletions or activating mutations in the region containing exon 3, which is involved in the degradation of β-catenin. Furthermore, hepatoblastoma frequently occurs in families with familial adenomatous polyposis coli (FAP), which accumulates β-catenin. Therefore, activation of Wnt signaling is thought to be involved in the pathogenesis.

[0003] In hepatoblastoma, the prognosis is influenced by whether or not complete resection is performed, so treatment strategies are selected based on postoperative staging. Stage I disease, where complete resection is performed, has a favorable prognosis, but in stage IV disease, where distant metastasis occurs, the 5-year disease-free survival rate drops to 36%. Traditionally, hepatoblastoma has been treated with a combination of chemotherapy and surgical resection. While cisplatin-containing regimens have been established as standard chemotherapy, serious side effects such as bone marrow suppression and renal dysfunction have become a problem.

[0004] On the other hand, GREB1 was discovered as one of the genes induced by estrogen in breast cancer cells (see Non-Patent Document 1). It has been reported that its expression is directly induced by estrogen receptor α (ERα), and that it is highly expressed in ERα-positive breast cancer cells but not in ERα-negative cells. ERα has been reported to bind to the promoter region of the GREB1 gene, express GREB1, and activate its transcriptional activity by directly interacting with ERα (see Non-Patent Document 2). In fact, it has also been reported that the expression level of GREB1 mRNA correlates with the expression level of ERα in breast cancer (see Non-Patent Document 2). Furthermore, it has been found that knockdown of GREB1 suppresses breast cancer cell proliferation, while overexpression of GREB1 promotes breast cancer cell proliferation (see Non-Patent Document 3). Furthermore, it has been reported that GREB1 is not expressed in ERα-negative cells but is expressed in ERα-positive ovarian cancer cells, and that GREB1 plays an important role in ovarian cancer cell proliferation (see Non-Patent Document 4). Because the GREB1 promoter region contains an androgen response element, GREB1 is induced by androgen in androgen receptor (AR)-positive prostate cancer cells, but not in AR-negative cells (see Non-Patent Document 5). Thus, GREB1 is a mediator of sex hormone-sensitive tumor growth.

[0005] However, it remains unclear whether GREB1 expression is involved in tumorigenesis other than sex hormone-sensitive tumors. [Prior art documents] [Non-patent literature]

[0006] [Non-Patent Document 1] Cancer Res. Nov 15;60(22):6367-75. [Non-patent document 2] Cell Rep. 2013 Feb 21;3(2):342-9. [Non-patent document 3] Breast Cancer Res Treat. 2005 Jul;92(2):141-9. [Non-patent document 4] Int J Cancer. 2014 Sep 1;135(5):1072-84. [Non-patent document 5] Prostate. 2006 Jun 1;66(8):886-94. Summary of the Invention [Problem to be solved by the invention]

[0007] An object of the present invention is to provide a novel tumor therapeutic agent. Another object of the present invention is to provide a testing method for predicting the presence or absence of hepatoblastoma, hepatocellular carcinoma, malignant melanoma, and neuroblastoma. [Means for solving the problem]

[0008] The present inventors conducted extensive research to solve the above-mentioned problems and found that hepatoblastoma, liver cancer, melanoma, and neuroblastoma cells, whose growth is not stimulated by sex hormones such as estrogen and androgen, express GREB1. They also found that GREB1 is involved in the growth of tumor cells with this characteristic (i.e., sex hormone-insensitive GREB1-positive tumors). Specifically, among sex hormone-insensitive GREB1-positive tumors, they found that GREB1 is a target gene of the Wnt / β-catenin signaling pathway in hepatoblastoma and hepatocellular carcinoma, that GREB1 is a target gene of the transcription factor MITF (Microphthalmia-associated transcription factor) in melanoma, and that GREB1 is gene-amplified in neuroblastoma. They also found that the growth of sex hormone-insensitive GREB1-positive tumors is suppressed by siRNA and antisense oligonucleotides against GREB1. Furthermore, immunochemical studies have revealed that GREB1 is expressed in a tumor tissue-specific manner in hepatoblastoma, hepatocellular carcinoma, malignant melanoma, and neuroblastoma. The present invention was completed based on these findings and further investigations.

[0009] That is, the present invention provides the following aspects. Item 1. A therapeutic agent for sex hormone-insensitive GREB1-positive tumors, the active ingredient of which is a substance that suppresses the expression of GREB1. Item 2. The therapeutic agent according to Item 1, wherein the substance is at least one nucleic acid drug selected from the group consisting of siRNA, shRNA, dsRNA, antisense nucleic acid, and ribozyme against GREB1. Item 3. The therapeutic agent according to Item 1 or 2, wherein the tumor is hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma. Item 4. A testing method for predicting the presence or absence of hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma, comprising: A testing method comprising a step of measuring the expression level of GREB1 in liver tissue, skin tissue, or nerve tissue collected from a subject. Item 5. The testing method according to Item 4, wherein the expression level of GREB1 in the liver tissue, skin tissue, or nerve tissue is measured by tissue immunization using an anti-GREB1 antibody. Item 6. A hepatoblastoma diagnostic agent comprising a reagent for detecting GREB1. Item 7. The test agent according to Item 6, wherein the reagent for detecting GREB1 is an anti-GREB1 antibody or a fragment thereof, or a primer capable of hybridizing with GREB1 mRNA or GREB1 cDNA. Item 8. A method for treating sex hormone-insensitive GREB1-positive tumors, comprising administering a therapeutically effective amount of a substance that suppresses the expression of GREB1 to a patient suffering from a sex hormone-insensitive GREB1-positive tumor. Item 9. The method of treatment according to Item 9, wherein the tumor is hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma. Item 10. Use of a substance that suppresses GREB1 expression for the manufacture of a therapeutic agent for sex hormone-insensitive GREB1-positive tumors. Item 11. The use according to Item 10, wherein the tumor is hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma. Item 12. A substance that inhibits the expression of GREB1, used in the treatment of sex hormone-insensitive GREB1-positive tumors. Item 13. The substance that suppresses the expression of GREB1 according to Item 12, wherein the tumor is hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma. [Effects of the Invention]

[0010] The therapeutic agents of the present invention were developed based on the new finding that GREB1 is a downstream target gene of the Wnt / β-catenin signal in sex hormone-insensitive GREB1-positive tumors. By targeting GREB1 and inhibiting its expression, the therapeutic agents can effectively inhibit tumor cell proliferation. Examples of sex hormone-insensitive GREB1-positive tumors include hepatoblastoma, hepatocellular carcinoma, malignant melanoma, and neuroblastoma. For example, hepatoblastoma is a rare cancer that is specific to children and occurs infrequently. Therefore, the development of chemotherapeutic agents and molecularly targeted drugs has not progressed sufficiently. However, one embodiment of the present invention makes it possible to provide an excellent molecular target drug for hepatoblastoma. Furthermore, analysis using gene expression databases has revealed that GREB1 expression is extremely limited in other major normal tissues. Therefore, the therapeutic agents of the present invention, which target GREB1, are also innovative in that they have few side effects.

[0011] Furthermore, according to the testing method of the present invention, by using the expression level of GREB1 in biopsy specimens or surgically resected specimens suspected of being tumorous diseases as an indicator, it can be applied to the diagnosis of early hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma, and in the case of surgically resected specimens, to the diagnosis of the extent of tissue progression of these tumors. [Brief explanation of the drawings]

[0012] [Figure 1](a) mRNA expression analysis of target genes downstream of Wnt / β-catenin signaling using hepatoblastoma cells. (b) TCF4-binding sequence (-443 to -448) located upstream of the human GREB1 gene, and the results of immunoprecipitation of chromatin from HepG2 cells with a specific antibody followed by PCR analysis of the TCF4-binding site (-443 to -448) using region-specific primers. (c) Real-time PCR analysis of GREB1 mRNA expression levels in HepG2 cells transfected with control siRNA or β-catenin siRNA (left), and the results of probing cell lysates with anti-GREB1, anti-Axin2, anti-β-catenin, and anti-HSP90 antibodies (right). (d) Immunostaining of hepatoblastoma tissue (n=11) with anti-GREB1 antibody and hematoxylin, and classification of the area stained by GREB1 immunostaining relative to the total area of ​​tumor lesions into three categories (<5%, 5-30%, and 30-95%). (e) Immunostaining of hepatoblastoma tissue specimens with anti-GREB1 antibody or anti-β-catenin antibody and hematoxylin, with solid boxes indicating areas with high β-catenin staining and dotted boxes indicating areas with low β-catenin staining. (f) Analysis of GREB1 mRNA expression levels in 50 tumor lesions and five non-tumor regions using 55 cases from the hepatoblastoma (GEO ID: gse75271) mRNA profile dataset. g shows the results of an analysis of the correlation between the expression levels of target genes of the Wnt / β-catenin pathway (Y-axis) and the expression level of the GREB1 gene (X-axis) using 55 cases in the mRNA profile dataset of hepatoblastoma (GEO ID: gse75271). [Figure 2]Figure 1a shows the results of measuring the expression level of GREB1 mRNA by real-time PCR after treating HepG2 cells with the estrogen receptor antagonist ICI-182,786. Figure 1b shows the results of measuring the expression level of GREB1 mRNA by real-time PCR after treating BMEL cells with the GSK3 inhibitor CHIR99021 (Wnt / β-catenin pathway activator) at a final concentration of 5 μM. Figure 1c shows the results of measuring the expression levels of GREB1 mRNA and β-catenin mRNA by real-time PCR after transfecting Huh6 cells with siRNA against β-catenin. Figure 1d shows the results of measuring the expression levels of GREB1 mRNA by real-time PCR in HepG2 cells and Huh6 cells. Figure 1e shows the results of measuring the expression levels of GREB1 and Axin2 mRNA by real-time PCR after treating MCF7 cells with the estrogen receptor antagonist ICI-182,786, CHIR99021, or a combination thereof. (f) Figure 1 shows the results of measuring GREB1 mRNA expression levels by real-time PCR after treating hepatocellular carcinoma cell lines HLE, SNU387, SNU449, and Huh7 with CHIR99021 at a final concentration of 5 μM. (g) Figure 2 shows the results of immunostaining hepatoblastoma tissue (n = 11) with anti-β-catenin antibody and hematoxylin, and the results of categorizing the area stained by β-catenin immunostaining relative to the total area of ​​tumor lesions into three categories (<5%, 5-30%, and 30-95%). (h) Figure 3 shows the results of immunostaining hepatoblastoma tissue specimens with anti-GREB1 antibody and hematoxylin. Solid boxes indicate areas with high GREB1 staining (solid and non-polarized), and dotted boxes indicate areas with low GREB1 staining (tubular and polarized). Panel i shows the results of an analysis of the correlation between the expression levels of target genes of the estrogen signaling pathway (Y-axis) and the expression level of the GREB1 gene (X-axis) using 55 cases in the mRNA profile dataset of hepatoblastoma (GEO ID: gse75271).j shows the results of analyzing the expression levels of GREB1 mRNA, GS mRNA, or LGR5 mRNA in cases with wild-type (n=3), point mutations (n=19), or deletions (n=28) of exons 3 and 4 of the β-catenin gene, using 55 cases in the mRNA profile dataset of hepatoblastoma (GEO ID: gse75271). [Figure 3](a) HepG2 cells or HepG2 cells expressing GREB1 (HepG2 / GREB1) were transfected with control siRNA or β-catenin siRNA, cultured in two dimensions (cultured in plastic dishes), and the cell counts were measured over time. (b) HepG2 cells were transfected with control siRNA or two different siRNAs (GREB1 #1 siRNA and GREB1 #2 siRNA), and the cell lysates were probed with anti-GREB1 antibody, anti-Axin2 antibody, anti-β-catenin antibody, and anti-HSP90 antibody. (c) Real-time PCR analysis of the levels of GREB1 mRNA, PRLR mRNA, and XBP1 mRNA in HepG2 cells transfected with control siRNA or siRNA (GREB1 #2 siRNA). Figure d shows the results of an analysis of the correlation between the expression levels of hepatoblastoma-specific marker genes (Y-axis) and the expression level of the GREB1 gene (X-axis) using 55 cases in the mRNA profile dataset of hepatoblastoma (GEO ID: gse75271). Figure e shows the results of measuring the cell count over time after two-dimensional culture (cultured in plastic dishes) of Huh6 cells expressing GFP or GFP-GREB1. Figure f shows the results of measuring the size of the spheres formed after three-dimensional culture (cultured in Matrigel) of mock- or GREB1-expressing Huh6 cells. Figure g shows the results of probing cell lysates of HepG2 cells, HepG2 cells in which GREB1 has been knocked out (GREB1 KO HepG2), or GREBI KO HepG2 cells expressing GFP-GREB1 (GREB1 KO HepG2 / GREB1) with anti-GREB1, anti-GFP, and anti-Histone H3 antibodies. FIG. 1h shows the results of measuring the cell number over time in two-dimensional culture (cultured in plastic dishes) of HepG2 cells (Control), HepG2 cells in which GREB1 was knocked out (GREB1 KO), or HepG2 cells in which GREB1 was expressed (GREB1 KO / GREB1).(i) shows the results of real-time PCR analysis of DLK1 mRNA and AFP mRNA levels in HepG2 cells, HepG2 cells with GREB1 knockout (GREB1 KO), and GREB1-expressing GREB1 KO HepG2 cells (GREB1 KO / GREB1). (j) shows the results of probing cell lysates from HepG2 cells, HepG2 cells with GREB1 knockout (GREB1 KO), and GREB1-expressing GREB1 KO HepG2 cells (GREB1 KO / GREB1) with anti-Cyclin A antibody, anti-phosphorylated Histone H3 (pHistone H3) antibody, and anti-Histone H3 antibody. [Figure 4](a) Mock- or GREB1-transfected HepG2 cells were transfected with control siRNA or two different siRNAs against GREB1 (GREB1 #1 siRNA and GREB1 #2 siRNA), cultured in two dimensions (cultured in plastic dishes), and the cell counts were measured over time. (b) HepG2 cells transfected with control siRNA or siRNA (GREB1 #2 siRNA) were cultured in three-dimensional Matrigel for five days, stained with phalloidin, and the sphere area and the percentage of polarized spheres with a lumen were calculated. (c) Real-time PCR analysis of the expression levels of GREB1 mRNA, DLK1 mRNA, AFP mRNA, and PEG3 mRNA in HepG2 cells transfected with control siRNA or siRNA against GREB1 (GREB1 #2 siRNA). (d) HepG2 cells transfected with control siRNA or two different siRNAs against GREB1 (GREB1 #1 siRNA and GREB1 #2 siRNA) were cultured in medium containing 0.1% FBS for 1 day, and the cell lysates were probed with anti-cyclin A antibody, anti-cyclin B antibody, anti-phosphorylated histone H3 antibody, anti-histone H3 antibody, anti-GREB1 antibody, and anti-HSP90 antibody. (e) Analysis of the correlation between cell proliferation marker genes (Y axis) and the expression level of the GREB1 gene (X axis) using 55 cases in the mRNA profile dataset of hepatoblastoma (GEO ID: gse75271). (f) HepG2 cells were transfected with control siRNA or two different siRNAs against GREB1 (GREB1 #1 siRNA and GREB1 #2 siRNA), cultured for 2 days in medium containing 0.1% FBS (with or without the caspase inhibitor Z-VAD), and stained with propidium iodide (PI) and Hoechst 33342 to determine the cell viability.(g) HepG2 cells were transfected with control siRNA or two different siRNAs against GREB1 (GREB1 #1 siRNA and GREB1 #2 siRNA) and cultured for 2 days, and the cell lysates were probed with anti-cleaved caspase 3 antibody, anti-PARP1 antibody, and anti-HSP90 antibody. [Figure 5](a) Schematic diagram of wild-type GREB1 and the GREB1ΔNLS mutant, and staining of X293T cells expressing the HA-FLAG-GREB1ΔNLS mutant with anti-FLAG antibody, phalloidin, and Hoechst33342. (b) Cell lysates (Input) and immunoprecipitates (IP) from X293T cells expressing HA-FLAG-GREB1 wild-type or HA-FLAG-GREB1ΔNLS mutant, and GFP, GFP-Smad3, GFP-Smad4, or GFP-Smad7 were probed with anti-HA or anti-GFP antibodies. (c) HepG2 cell lysates (Input) and immunoprecipitates (IP) from anti-Smad2 / 3 antibodies were probed with anti-Smad2 / 3 and anti-GREB1 antibodies. (d) Schematic diagram of Smad2 and Smad2 mutants (N and C), and the results of probing with anti-FLAG or anti-GFP antibodies the lysate (Input) of X293T cells expressing HA-FLAG-mGREB1 and GFP, or GFP-Smad2 (Full, mutant N, or mutant C) and the immunoprecipitates (IP) prepared with anti-GFP antibodies from the cells. (e) HepG2 cell lysate was precipitated with recombinant GST, recombinant GST-Smad2 / MH1, or recombinant GST-Smad2 / MH2, and the cell lysate (Input) and the precipitates prepared with glutathione-sepharose (pulldown) were stained with CBB or probed with anti-GREB1 and anti-Smad4 antibodies. (f) Schematic diagram of GREB1 mutants (1-666:N, NLS / 667-1333:M, and NLS / 1334-1954:C) and staining of X293T cells expressing GFP-GREB1 mutants with anti-GFP antibody and Hoechst 33342. (g) Lysates (Input) and immunoprecipitates (IP) of X293T cells expressing FLAG-Smad3 and GFP, GFP-GREB1, or GFP-GREB1 mutants were probed with anti-FLAG antibody or anti-GFP antibody. [Figure 6](a) shows the results of staining X293T cells expressing GFP-Smad3, GFP-Smad4, or GFP-Smad7 with anti-GFP antibody and Hoechst33342. (b) shows the results of probing Huh6 cell lysates (Input) expressing GFP-GREB1 and immunoprecipitates (IP) prepared with anti-Smad2 / 3 antibody with anti-GFP antibody. (c) shows the results of probing X293T cell lysates (Input) expressing GFP, GFP-Smad3, GFP-Smad4, or GFP-Smad7 and immunoprecipitates (IP) prepared with anti-GFP antibody with anti-β-catenin antibody, anti-c-Myc antibody, or anti-GFP antibody. d shows the results of precipitating the lysate of Huh6 cells expressing GFP-GREB1 with recombinant GST, recombinant GST-Smad2 / MH1, or recombinant GST-Smad2 / MH2, and then staining the cell lysate (Input) and the precipitate (pulldown) with glutathione sepharose with CBB or probing with anti-GFP antibody. e) Schematic diagram of GREB1 mutants (NSL / 667-195(ΔN), Δ667-1333(ΔM), and NSL / 1-133(ΔC)), and results of probing with anti-FLAG or anti-GFP antibodies lysates (Input) and immunoprecipitates (IP) from X293T cells expressing FLAG-Smad3 and GFP, GFP-GREB1, or GFP-GREB1 mutants (NSL / 667-195(ΔN), Δ667-1333(ΔM), and NSL / 1-1333(ΔC)) with anti-FLAG or anti-GFP antibodies. [Figure 7](a) is a diagram showing the results of analyzing PAI-1 mRNA and GADD45B mRNA expression levels in 50 tumor lesion sites and 5 non-tumor regions using 55 cases in the mRNA profile dataset of hepatoblastoma (GEO ID: gse75271). (b) is a diagram showing the results of analyzing the correlation between the expression levels of target genes of TGFβ signaling (Y axis) and the expression level of the GREB1 gene (X axis) using 55 cases in the mRNA profile dataset of hepatoblastoma (GEO ID: gse75271). (c) is a diagram showing the results of measuring the intracellular PAI-1 and SNAIL2 mRNA levels in HepG2 cells expressing GFP (HepG2 / GFP) or HepG2 cells expressing GFP-GREB1 (HepG2 / GFP-GREB1) transfected with control siRNA or GREB1 #2 siRNA. (d) HepG2 cells were transfected with control siRNA or GREB1 #2 siRNA and cultured in the presence or absence of a TGFβ receptor inhibitor (ALK5 inhibitor), and the intracellular PAI-1 and SNAIL2 mRNA levels were measured. (e) Lysates (Input) from HepG2 cells transfected with control siRNA or GREB1 #2 siRNA and immunoprecipitates (IP) from anti-Smad2 / 3 antibodies were probed with anti-p300, anti-GREB1, or anti-Smad2 / 3 antibodies. (f) Lysates (Input) from X293T cells expressing HA-FLAG-GREB1, GFP-Smad2 mutant (C), or GFP and immunoprecipitates (IP) from GFP antibodies were probed with anti-p300, anti-FLAG, or anti-GFP antibodies. (g) HepG2 cells transfected with control siRNA or GREB1 #2 siRNA were cultured in the presence or absence of TGFβ, and the exon 2 region of PAI-1 contained in the cell lysate (Input) and immunoprecipitate (IP) after culture was analyzed by PCR using region-specific primers.(h) shows the results of transfecting HepG2 cells with the indicated plasmids and measuring the expression levels of SNAIL2 mRNA, p15 mRNA, or Axin2 mRNA by real-time PCR. (i) shows the results of two-dimensionally culturing HepG2 cells transfected with control siRNA, TGFβ1 siRNA, GREB1 siRNA (GREB1 #2 siRNA), or a combination of TGFβ1 siRNA and GREB1 siRNA (GREB1 #2 siRNA), followed by measuring the cell count over time. (j) shows the results of two-dimensionally culturing HepG2 cells and HepG2 cells with Smad2 / 3 knockout (HepG2 / Smad2 / 3 KO) transfected with control siRNA or GREB1 siRNA (GREB1 #2 siRNA), followed by measuring the cell count over time. [Figure 8](a) Analysis of Axin2 mRNA and DKK1 mRNA expression levels in 50 tumor lesion sites and 5 non-tumor regions using 55 cases in the mRNA profile dataset of hepatoblastoma (GEO ID: gse75271). (b) Analysis of PAI-1 mRNA levels by real-time PCR in HepG2 cells (Control), HepG2 cells with GREB1 knockout (GREB1 KO), or GREB1-expressing GREB1 KO HepG2 cells (GREB1 KO / GREB1). (c) HepG2 cells transfected with control siRNA or siRNA against GREB1 (GREB1 #2 siRNA) were cultured in the presence or absence of 10 ng / mL TGFβ for 30 minutes, and then stained with anti-Smad2 / 3 antibody and Hoechst 33342. (d) HepG2 cells transfected with control siRNA or siRNA against GREB1 (GREB1 #2 siRNA) were cultured in the presence or absence of TGFβ, and the cell lysates (Input) and immunoprecipitates (IP) prepared with anti-Smad2 / 3 antibodies were probed with anti-Smad4 or anti-Smad2 / 3 antibodies. (e) HepG2 cells transfected with control siRNA or siRNA against GREB1 (GREB1 #2 siRNA) were cultured in the presence or absence of TGFβ, and the cell lysates were probed with anti-GREB1 antibodies, anti-phosphorylated Smad2 / 3 (pSmad2 / 3) antibodies, or anti-Smad2 / 3 antibodies. (f) HepG2 cells were expressing the TGFBR1 / T204D mutant and the levels of AFP mRNA and DLK1 mRNA were analyzed by real-time PCR. FIG. 1g shows the results of probing lysates of HepG2 cells and HepG2 cells in which Smad2 / 3 was knocked out (HepG2 / Smad2 / 3 KO) with anti-Smad2 / 3 antibody or anti-HSP90 antibody.FIG. 1h shows the results of analyzing the AFP mRNA level by real-time PCR in HepG2 cells and HepG2 cells in which Smad2 / 3 was knocked out (HepG2 / Smad2 / 3 KO) transfected with control siRNA or siRNA against GREB1 (GREB1 #2 siRNA). [Figure 9](a) shows the results of analyzing the gene expression levels of TGFβ1, TGFB2, and TGFB3 using HepG2 RNA sequence data from the Cancer Cell Line Encyclopedia (CCLE) mRNA profile dataset. (b) shows the results of analyzing the levels of PAI-1 mRNA, TGFB1 mRNA, and GREB1 mRNA by real-time PCR in HepG2 cells transfected with control siRNA, TGFβ1 siRNA, GREB1 siRNA (GREB1 #2 siRNA), or a combination of siRNAs containing TGFβ1 and GREB1 (GREB1 #2 siRNA). (c) shows the results of probing lysates of HepG2 cells transfected with control siRNA, TGFβ1 siRNA, GREB1 siRNA (GREB1 #2 siRNA), or a combination of siRNAs containing TGFβ1 and GREB1 (GREB1 #2 siRNA) with anti-GREB1 antibody, anti-TGFB1 antibody, or anti-HSP90 antibody. (d) HepG2 cells transfected with control siRNA, TGFβ1 siRNA, or a combination of TGFβ1 siRNA and GREB1 siRNA (GREB1 #2 siRNA) were cultured for 4 hours in the presence or absence of TGFβ at 0.01, 0.1, or 1 ng / ml, and then analyzed by real-time PCR for PAI-1 mRNA levels. (e) HepG2 cells transfected with control siRNA or siRNA against GREB1 (GREB1 #2 siRNA) were analyzed by real-time PCR for p15 mRNA levels, p21 mRNA levels, and p27 mRNA levels. (f) HepG2 cells and HepG2 cells with Smad2 / 3 knockout (HepG2 / Smad2 / 3 KO) were transfected with control siRNA or siRNA against GREB1 (GREB1 #2 siRNA), and then analyzed by real-time PCR for p15 mRNA levels.Figure 1g shows the results of two-dimensional culture (culture in plastic dishes) of HepG2 cells transfected with control siRNA, GREB1 siRNA (GREB1 #2 siRNA), p15 siRNA, or a combination of GREB1 (GREB1 #2 siRNA) and p15 siRNA, followed by measurement of cell number over time. Figure 1h shows the results of HepG2 cells transfected with control siRNA, GREB1 siRNA (GREB1 #2 siRNA), p15 siRNA, or a combination of GREB1 (GREB1 #2 siRNA) and p15 siRNA, cultured in 0.1% FBS-containing medium (with or without the caspase inhibitor Z-VAD) for 2 days, stained with propidium iodide (PI) and Hoechst 33342, and then measured for cell viability. Figure 1i shows the results of measurement of PAI-1 mRNA and GREB1 mRNA expression levels by real-time PCR after treatment of MCF7 cells with the estrogen receptor antagonist ICI-182,786. (j) shows the results of analyzing the levels of PAI-1 mRNA or GREB1 mRNA by real-time PCR after transfection of MCF7 cells with control siRNA or siRNA against GREB1 (GREB1 #2 siRNA). (k) shows the results of probing MCF7 cell lysates (Input) and immunoprecipitates (IP) with anti-Smad2 / 3 antibody and anti-GREB1 antibody. [Figure 10](a) is a diagram showing the results of immobilizing GFP-GREB1-expressing HepG2 cells and staining them with an anti-GFP antibody and Hoechst 33342. (b) is a diagram showing the results of immobilizing GFP-GREB1-expressing HepG2 cells and staining them with an anti-GFP antibody, an anti-fibrillarin antibody, an anti-SC35 antibody, an anti-PML antibody, or an anti-coilin antibody, and Hoechst 33342. (c) is a diagram showing the results of immobilizing GFP-GREB1- and FLAG-SMAD3-expressing HepG2 cells and staining them with an anti-GFP antibody, an anti-FLAG antibody, and Hoechst 33342. (d) is a diagram showing the results of incubating HepG2 cells in the presence of ethynyluridine (EU) for 30 minutes, followed by fixation and observation. [Figure 11] Figure 1a shows HepG2 cells cultured for 30 minutes in the presence or absence of 10 ng / mL TGFβ, fixed, and stained with anti-SMAD2 / 3 antibody, anti-GREB antibody, and Hoechst 33342. Figure 1b shows HepG2 cells cultured in the presence or absence of TGFβ, fixed, and stained with anti-GREB1 antibody, anti-phosphorylated SMAD2 / 3 (pSMAD2 / 3) antibody, and Hoechst 33342. Figure 1c shows HepG2 cells cultured in the presence or absence of TGFβ, fixed, and stained with mouse anti-GREB1 antibody and rabbit anti-SMAD2 / 3 antibody, followed by binding of a secondary antibody (PLA probe) (left), and the percentage of cells with punctate PLA signals in the nucleus (right). (d) shows the results of staining X293T cells expressing GFP-linked GREB1 mutants (1-666:N, NLS / 667-1333:M, and NLS / 1334-1954:C) with an anti-GFP antibody. (e) shows the results of incubating HepG2 cells expressing GFP-SMAD3 and either expressing or not expressing HA-FLAG-GREB1 in the presence of ethynyluridine (EU), followed by fixation and staining with an anti-GFP antibody, an anti-GREB1 antibody, and Hoechst33342. [Figure 12](a) Immunostaining of hepatoblastoma tissues (n=11) with anti-β-catenin antibody and hematoxylin, and classification of the area stained by β-catenin immunostaining relative to the total area of ​​tumor lesions into three categories (<5%, 5-30%, and 30-95%). (b) Liver tissue sections isolated from mice transfected with ΔN90β-catenin and YAPS127A (BY model), ΔN90β-catenin and c-Met (BM model), or ΔN90β-catenin, YAPS127A, and c-Met (BYM model) were stained with anti-GREB1 or anti-DLK1 antibody and hematoxylin, and the liver was examined. c shows the results of analyzing the mRNA levels of GREB1 and TASCSTD1 by real-time PCR after extracting total RNA from liver tumor nodules (BY: n=3; BM: n=7; BYM: n=11) of mice transfected with ΔN90β-catenin and YAPS127A (BY model), ΔN90β-catenin and c-Met (BM model), or ΔN90β-catenin, YAPS127A, and c-Met (BYM model). [Figure 13](a) shows the results of fixing Huh6 cells or HepG2 cells and staining them with an anti-YAP antibody and Hoechst33342. (b) shows the results of measuring the expression levels of GREB1 mRNA, YAP mRNA, and TAZ mRNA by real-time PCR in HepG2 cells or Huh6 cells transfected with control siRNA or siRNA against YAP / TAZ. (c) shows the results of measuring the expression levels of GREB1 mRNA, Axin2 mRNA, and Cyr61 mRNA by real-time PCR in Huh6 cells treated with CHIR99021, XMU-MP1, an Mst1 / 2 kinase inhibitor (YAP activator), or a combination thereof. (d) shows the results of measuring the expression levels of c-Met mRNA, Axin2 mRNA, GREB1 mRNA, ANKRD1 mRNA, and Cyr61 mRNA by real-time PCR in HepG2 cells transfected with control siRNA or siRNA against c-Met. e shows the results of probing lysates of HepG2 cells transfected with control siRNA or siRNA against c-Met with anti-c-Met antibody, anti-phosphorylated β-catenin (pY654) antibody, anti-β-catenin antibody, anti-Axin2 antibody, anti-GREB1 antibody, and anti-β-Actin antibody. [Figure 14](a) Schematic diagram of livers from mice transfected with ΔN90β-catenin, YAPS127A, and c-Met (BYM, control; C1–C7); mice treated with ΔN90β-catenin, YAPS127A, and c-Met together with GREB1 shRNA (BYM + GREB1 shRNA; BYM GREB1 KD mice, K1–K6); and untreated mice (NL mice). (b) Six tumor nodules were isolated from the BYM mice (C1–C7) and the expression levels of hepatoblastoma-related genes were measured. (c) Hematoxylin-eosin staining of liver tissue sections isolated from BYM mice with high GREB1 expression (C4) and low GREB1 expression (C3) was performed. Figure d shows the results of staining liver tissue sections isolated from BYM mice in the GREB1-high group (C4) and the GREB1-low group (C3) with anti-GREB1 or anti-DLK1 antibodies and hematoxylin. Figure e shows the liver weight / total body weight ratio and serum AFP levels for each mouse. Figure f shows the results of analyzing the mRNA levels of GREB1, DLK1, and TASCSTD1 by real-time PCR using total RNA extracted from tumor nodules in three BYM mice (GREB1-high group: C1, C4, and C6), four BYM mice (GREB1-low group: C2, C3, C5, and C7), and two BYM mice (BYM GREB1 KD mice; K2 and K4) administered with GREB1 shRNA. (g) Liver tissue sections from BYM mice (GREB1 high group: C4) and BYM mice administered with GREB1 shRNA (GREB1 shRNA: K2) were stained with hematoxylin and eosin or anti-GREB1 antibody and hematoxylin. (h) Liver tissue sections from BYM mice (GREB1 high group: C4) and BYM mice administered with GREB1 shRNA (GREB1 shRNA: K2) were stained with anti-N-cadherin antibody and Hoechst 33342 (left panel), and the ratio of N-cadherin expression levels in tumor lesions to those in non-tumor areas was calculated (bottom panel). [Figure 15](a) shows the results of measuring the expression levels of the introduced ΔN90β-catenin, YAPS127A, and c-Met in tumor nodules obtained from the livers of mice (BYM, control; C1 to C7) into which ΔN90β-catenin, YAPS127A, and c-Met had been introduced and untreated mice (NL mice). (b) shows the results of measuring the mRNA levels of GREB1 by real-time PCR in total RNA extracted from the livers of untreated mice (NL), tumor nodules (n=42) from mice (BYM) into which ΔN90β-catenin, YAPS127A, and c-Met had been introduced, and non-tumor tissues (n=8) from the BYM mice. Figure 1c shows the results of an analysis of the correlation between the expression levels (Y-axis) of DLK1 mRNA, TACSTD1 mRNA, GPC3 mRNA, MEG3 mRNA, and Axin2 mRNA and the expression level (X-axis) of the GREB1 gene in tumor nodules obtained from the BYM mice (C1 to C7). Figure 1d shows the results of hematoxylin-eosin staining of liver tissue sections isolated from BYM mice in the GREB1-high group (C4) and the GREB1-low group (C3). Figure 1e shows the results of probing lysates from HepG2 cells (Control), HepG2 cells in which GREB1 was knocked out (GREB1 KO), or HepG2 cells in which GREB1 and Smad2 / 3 were knocked out in combination (GREB1 KO + Smad2 / 3 KO) with anti-GREB1, anti-Smad2 / 3, and anti-HSP90 antibodies. FIG. 1f shows the results of measuring the cell number over time in two-dimensional culture (cultured in plastic dishes) of HepG2 cells (WT), HepG2 cells in which GREB1 was knocked out (GREB1 KO), or HepG2 cells in which GREB1 and Smad2 / 3 were knocked out in combination (GREB1 KO / Smad2 / 3 KO). [Figure 16](a) Wild-type HepG2 cells (control), HepG2 cells with GREB1 knockout (GREB1 KO), HepG2 cells with GREB1 knockout and GREB1 expression (GREB1KO / GREB1), HepG2 cells with GREB1 knockout and GREB1 expression (GREB1KO / GREB1ΔNLS), or cells with GREB1 and Smad2 / 3 knockout in combination (GREB1 KO / Smad2 / 3 KO) were subcutaneously transplanted into mice, and the appearance and weight of the xenograft tumor pieces were measured 28 days after transplantation. (b) HepG2 cells expressing GFP or GFP-GREB1 were transfected with control ASO or each GREB1 ASO and cultured in three-dimensional Matrigel. The cultured cells were stained with phalloidin and Hoechst 33342 to determine the area of ​​the spheres. (c) Matrigel containing HepG2 cells was transplanted into the livers of nude mice, and starting 3 days after transplantation, control ASO, GREB1 ASO-6921, or GREB1 ASO-7724 was subcutaneously administered twice a week. The appearance of the liver tumors was observed 27 days after transplantation, and tumor weights were calculated. (d) Sections of each liver tumor were stained with anti-Ki-67 antibody and hematoxylin, and the ratio of Ki-67-positive cells to the total number of hematoxylin-positive cells was calculated. (e) Sections of each liver tumor were stained with anti-cleaved-caspase 3 antibody and hematoxylin, and the ratio of cleaved-caspase 3-positive cells to the total number of hematoxylin-positive cells was calculated. (f) Real-time PCR analysis of the mRNA levels of GREB1 and PAI-1 in each liver tumor. [Figure 17](a) HepG2 cells were transfected with control ASO or each GREB1 ASO, and the cell lysates were probed with anti-GREB1 and anti-HSP90 antibodies. (b) Matrigel containing HepG2 cells was transplanted into the liver of nude mice. Starting 3 days after transplantation, control ASO, GREB1 ASO-6921, or GREB1 ASO-7724 was administered subcutaneously twice a week. On day 27 after transplantation, sections of non-tumor liver samples were stained with anti-cleaved caspase 3 antibody and hematoxylin to determine the percentage of cleaved caspase 3-positive cells. (c) ΔN90β-catenin, YAPS127A, and c-Met (BYM) were introduced into mice, and starting 3 days after introduction, a control ASO or an ASO targeting mouse GREB1 (mGREB1 ASO-5715) was subcutaneously administered twice a week. Six to seven weeks after transplantation, the appearance of liver tumors was observed and the ratio of liver weight to total body weight was calculated. (d) GREB1 mRNA levels in each liver tumor were analyzed by real-time PCR. [Figure 18] (A) is a scatter plot showing GREB1 mRNA expression levels in cancer cell lines obtained from The Cancer Cell Line Encyclopedia (CCLE) using the DepMap portal. (B) is a diagram showing the results of probing lysates of each cancer cell line with anti-GREB1 and anti-GAPDH antibodies. (C) is a diagram showing the results of immunostaining of neuroblastoma tissue specimens (n ​​= 13) with anti-GREB1 and anti-β-catenin antibodies and hematoxylin, and the results of classifying the area stained by GREB1 or β-catenin immunostaining relative to the total area of ​​tumor lesions into three categories (<5%, 5-30%, and 30-95%). (D) is a table showing the results of analyzing the relationship between GREB1 and β-catenin protein levels based on the results of the immunohistological analysis in (C). (E) is a diagram showing the results of measuring the cell number (mean ± SD) over time in two-dimensional cultures (in plastic dishes) of CHP212 cells transfected with control ASO or GREB1 ASO. [Figure 19] Figure 1A shows the results of immunostaining hepatocellular carcinoma tissue specimens (n=210) with anti-GREB1 antibody, anti-β-catenin antibody, and hematoxylin. Figure 1B shows the results of analyzing the relationship between the protein levels of GREB1 and β-catenin based on the results of the immunohistological analysis in Figure 1A. Figure 1C shows the results of probing lysates of each cancer cell line with anti-GREB1 antibody and anti-GAPDH antibody. Figure 1D shows the results of probing lysates of each cell line transfected with control siRNA or β-catenin siRNA with anti-GREB1 antibody, anti-β-catenin antibody, and anti-GAPDH antibody. Figure 1E shows the results of counting Hep3B or JHH7 cells transfected with control siRNA or GREB1#2 siRNA in two-dimensional culture (cultured in plastic dishes) over time. (F) Hep3B cells (control KO) and GREB1-knockout Hep3B cells (GREB1 KO) were subcutaneously implanted into nude mice (n=6). The left panel shows a representative appearance of xenograft tumors excised 45 days after implantation. The middle panel shows the time course of tumor size. The right panel shows tumor weight after 45 days. (G) JHH7 cells (control KO) and GREB1-knockout JHH7 cells (GREB1 KO) were subcutaneously implanted into nude mice (n=6). The left panel shows a representative appearance of xenograft tumors excised 14 days after implantation. The middle panel shows the time course of tumor size. The right panel shows tumor weight after 14 days. **P < 0.01; *P < 0.05. [Figure 20](A) shows the results of probing cell lysates with anti-GREB1 and anti-HSP90 antibodies. (B) The left panel shows a schematic diagram of human GREB1 isoforms (aa stands for amino acid). (B) The right panel shows the average exon expression level (OPKM; Observations Per Kilobases of exon / splice per Million aligned reads) of the human GREB1 gene in breast cancer and cutaneous melanoma obtained from the TCGA dataset using TCGA SpliceSeq (http: / / projects.insilico.us.com / TCGASpliceSeq). (C) A table showing the top 10 genes correlated with GREB1 expression in the cutaneous melanoma dataset obtained from TCGA using the R2: genomics analysis and visualization platform (http: / / r2.amc.nl). R is the Pearson correlation coefficient, and * indicates a known downstream target gene of MITF. D shows the results of immunostaining cutaneous malignant melanoma specimens with anti-GREB1 antibody, anti-MITF antibody, and hematoxylin. E shows the results of probing CLO679 cells (control) and CLO679 cells in which MITF has been knocked out (MITF KO) with anti-GREB1 antibody, anti-MITF antibody, and anti-clathrin antibody. F shows the results of measuring the cell number (mean ± SD) over time in SKMEL28 cells transfected with control siRNA or GREB1 siRNA, which were cultured in two dimensions (cultured in plastic dishes). G shows the results of measuring the cell number (mean ± SD) over time in SKMEL28 cells transfected with control ASO or GREB1 ASO, which were cultured in two dimensions (cultured in plastic dishes). DETAILED DESCRIPTION OF THE INVENTION

[0013] 1.Treatment for hormone-insensitive GREB1-positive tumors The therapeutic agent of the present invention is a pharmaceutical used to treat sex hormone-insensitive GREB1-positive tumors, and is characterized by containing as an active ingredient a substance that suppresses the expression of GREB1. The therapeutic agent of the present invention will be described in detail below.

[0014] [Active ingredient] The therapeutic agent of the present invention uses a substance that suppresses the expression of GREB1 as an active ingredient. GREB1 is a target gene of Wnt / β-catenin signaling and MITF in sex hormone-insensitive GREB1-positive tumor cells, and can undergo gene amplification, thereby promoting the proliferation of the tumor cells. By suppressing the expression of GREB1, the therapeutic agent of the present invention can effectively suppress the proliferation of the tumor.

[0015] The amino acid sequence and nucleotide sequence of GREB1 are also publicly known. For example, the amino acid sequence of human GREB1 (isoform a) is known as SEQ ID NO: 1, the nucleotide sequence of human GREB1 (isoform a) mRNA is known as SEQ ID NO: 2, and the nucleotide sequence of cDNA encoding human GREB1 (isoform a) is known as SEQ ID NO: 3.

[0016] In the present invention, the term "substance that suppresses GREB1 expression" is not particularly limited, as long as it is pharmaceutically acceptable and capable of suppressing GREB1 expression from DNA encoding GREB1 (GREB1 gene). Substances that suppress GREB1 expression may exert an inhibitory effect on GREB1 expression at any stage of the GREB1 gene, such as transcription, post-transcriptional regulation, translation, or post-translational modification. Specific examples of substances that suppress GREB1 expression include nucleic acid molecules that suppress the transcription of the GREB1 gene, such as decoy nucleic acids; RNA molecules or precursors thereof that exhibit RNA interference activity against GREB1 mRNA, such as siRNA, shRNA, and dsRNA; and nucleic acid drugs, such as miRNA, antisense nucleic acids (antisense DNA, antisense RNA), and ribozymes that suppress the translation of GREB1 mRNA. These nucleic acid molecules may be used alone or in combination of two or more. The nucleotide sequences of these nucleic acid molecules can be appropriately designed by those skilled in the art using known techniques based on the nucleotide sequence of the GREB1 gene.

[0017] Among these nucleic acid molecules, from the viewpoint of ease of clinical application, siRNA, shRNA, dsRNA, antisense nucleic acid, and ribozyme are preferred, and siRNA, shRNA, and antisense nucleic acid are more preferred.

[0018] Furthermore, the nucleic acid molecule may be subjected to various modifications generally applied to nucleic acids, as necessary, to confer resistance to nucleases and the like. Examples of such modifications include modifications of the sugar chain moiety, such as 2'-fluorolation and 2'-O-methylation; modifications of the base moiety; and modifications of the phosphate moiety, such as amination, lower alkylamination, acetylation, and phosphorothioate. Furthermore, the nucleic acid molecule may have an artificial nucleic acid (such as a cross-linked nucleic acid, peptide nucleic acid, or locked nucleic acid) introduced into a portion of the RNA molecule and / or DNA molecule.

[0019] For example, a suitable example of a crosslinked nucleic acid is a nucleotide having a structure shown in the following general formula (1). [ka]

[0020] In general formula (1), Base is a base corresponding to the base sequence, specifically a purin-9-yl group or a 2-oxo-1,2-dihydropyrimidin-1-yl group which may be substituted with a substituent. Specific examples of the substituent include a hydroxyl group, a linear alkyl group having 1 to 6 carbon atoms, a linear alkoxy group having 1 to 6 carbon atoms, a mercapto group, a linear alkylthio group having 1 to 6 carbon atoms, an amino group, a linear alkylamino group having 1 to 6 carbon atoms, and a halogen atom.

[0021] Specific examples of Base in general formula (1) include, when the base is A (adenine), a 6-aminopurin-9-yl group which may be substituted with a substituent; when the base is G (guanine), a 2-amino-6-hydroxypurin-9-yl group which may be substituted with a substituent; when the base is C (cytosine), a 2-oxo-4-amino-1,2-dihydropyrimidin-1-yl group which may be substituted with a substituent (for example, a 4-amino-5-methyl-2-oxo-1,2-dihydropyrimidin-1-yl group (including a 5-methylcytosin-1-yl group, a 2'-O-methylcytosin-1-yl group, etc.); and when the base is T (thymine), a 2-oxo-4-hydroxy-5-methyl-1,2-dihydropyrimidin-1-yl group which may be substituted with a substituent.

[0022] In general formula (1), R represents a hydrogen atom, an alkyl group having 1 to 7 carbon atoms, which may be branched or cyclic, an alkenyl group having 2 to 7 carbon atoms, which may be branched or cyclic, an aryl group having 3 to 12 carbon atoms, which may have a substituent and contain a heteroatom, or an aralkyl group having an aryl moiety having 3 to 12 carbon atoms, which may have a substituent and contain a heteroatom. Specific examples of the substituent that may be contained in the aryl group or aralkyl group include a hydroxyl group, a linear alkyl group having 1 to 6 carbon atoms, a linear alkoxy group having 1 to 6 carbon atoms, a mercapto group, a linear alkylthio group having 1 to 6 carbon atoms, an amino group, a linear alkylamino group having 1 to 6 carbon atoms, and a halogen atom. R is preferably a hydrogen atom, a methyl group, an ethyl group, an n-propyl group, an isopropyl group, a phenyl group, or a benzyl group, more preferably a hydrogen atom or a methyl group, and particularly preferably a methyl group. In this specification, a 2',4'-bridged nucleotide in which R in general formula (1) is a methyl group may be referred to as "AmNA."

[0023] Another example of a bridged nucleotide is a nucleotide having a structure shown in the following general formula (2): This nucleotide is a known 2',4'-bridged nucleotide, also known as a guanidine-bridged nucleic acid (WO 2014 / 046212). [ka] In the general formula (2), Base is the same as Base in the general formula (1). 1 , R 12 , and R 13 are the same or different and represent a hydrogen atom or an alkyl group having 1 to 7 carbon atoms which may be branched or cyclic, R 14 indicates a hydrogen atom.

[0024] Another example of a bridged nucleotide is a nucleotide having a structure shown in the following general formula (3): This nucleotide is a known 2',4'-bridged nucleotide, also known as a spirocyclopropane-bridged nucleic acid (WO 2015 / 125783). [ka]

[0025] In the general formula (3), Base is the same as Base in the general formula (1). 21 and R 22 are the same or different and are a hydrogen atom; an alkyl group having 1 to 7 carbon atoms which may be substituted with an aryl group having 3 to 12 carbon atoms which may contain a heteroatom and which may form a branched or cyclic group; or an aralkyl group having an aryl moiety having 3 to 12 carbon atoms which may contain a heteroatom; or R 21 and R 22 together form the group -(CH2) n - [wherein n is an integer of 2 to 5].

[0026] Other examples of bridged nucleotides include nucleotides having the structure shown in the following general formula (4) or (4'): These nucleotides are known 2',4'-bridged nucleotides, also known as ethyleneoxy-bridged nucleic acids (WO 2016 / 017422). [ka]

[0027] In the general formulas (4) and (4'), Base is the same as Base in the general formula (1). 3 represents an oxygen atom or a sulfur atom.

[0028] In the general formulas (4) and (4′), R 31 and R 32are the same or different and represent a hydrogen atom; a hydroxyl group; an alkyl group having 1 to 7 carbon atoms which may be branched or cyclic; an alkoxy group having 1 to 7 carbon atoms which may be branched or cyclic; or an amino group. In the case of general formula (4), R 31 and R 32 are taken together to form the group =C(R 35 )R 36 [In the formula, R 35 and R 36 may be the same or different and represent a hydrogen atom, a hydroxyl group, a mercapto group, an amino group, a linear or branched alkoxy group having 1 to 6 carbon atoms, a linear or branched alkylthio group having 1 to 6 carbon atoms, a cyanoalkoxy group having 1 to 6 carbon atoms, or a linear or branched alkylamino group having 1 to 6 carbon atoms.

[0029] In the general formulas (4) and (4′), R 33 represents a hydrogen atom, an alkyl group having 1 to 7 carbon atoms which may be branched or cyclic, an alkoxy group having 1 to 7 carbon atoms which may be branched or cyclic, or a linear or branched alkylthio group having 1 to 6 carbon atoms.

[0030] In general formula (4), R 34 represents a hydrogen atom, an alkyl group having 1 to 7 carbon atoms which may be branched or cyclic, an alkoxy group having 1 to 7 carbon atoms which may be branched or cyclic, or a linear or branched alkylthio group having 1 to 6 carbon atoms.

[0031] Further, other examples of bridged nucleotides include the structures shown below: Base and R in the following structural formula are the same as Base and R in the general formula (1) above. [ka]

[0032] In the nucleic acid molecules used in the present invention, chemical modifications may be made to some of the nucleotides and / or internucleoside linkages, or may be made to all of the nucleotides and / or internucleoside linkages.

[0033] When the nucleic acid molecule used in the present invention is an antisense nucleic acid, a preferred example of chemical modification is that it contains at least one, preferably 1 to 10, more preferably 2 to 8, even more preferably 2 to 6, and particularly preferably 5 2',4'-bridged nucleotides. Furthermore, a preferred example of a 2',4'-bridged nucleotide is that the first to third nucleotides from the 5'-end and the second and third nucleotides from the 3'-end are 2',4'-bridged nucleotides (preferably AmNA).

[0034] A preferred example of chemical modification applied to the nucleic acid molecule used in the present invention is that at least one of the internucleoside bonds is a phosphorothioate bond. In addition, a preferred example of a phosphorothioate bond is that, per 100% of the total number of internucleoside bonds, preferably 50% or more, more preferably 80% or more, even more preferably 90% or more, and particularly preferably 100% (all internucleoside bonds) are phosphorothioate bonds.

[0035] When the nucleic acid molecule used in the present invention is an siRNA against human GREB1, specific examples thereof include an siRNA comprising a sense strand consisting of the nucleotide sequence shown in SEQ ID NO: 4 and an antisense strand consisting of the nucleotide sequence shown in SEQ ID NO: 5; and an siRNA comprising a sense strand consisting of the nucleotide sequence shown in SEQ ID NO: 6 and an antisense strand consisting of the nucleotide sequence shown in SEQ ID NO: 7. In SEQ ID NOs: 4 to 6, positions 1 to 19 represent the RNA strand, and positions 20 and 21 represent overhangs (deoxythymidine).

[0036] Furthermore, when the nucleic acid molecule used in the present invention is an antisense nucleic acid (antisense DNA, ASO) against human GREB1, specific examples thereof include chemically modified ASOs consisting of the following sequences A to D. 5(Y)^G(Y)^A(Y)^a^t^g^g^c^a^g^g^a^5(Y)^A(Y)^g (Sequence A: used as ASO-6434 in the examples) G(Y)^T(Y)^5(Y)^t^g^t^t^t^c^a^a^g^T(Y)^A(Y)^a (Sequence B: used as ASO-6921 in the examples) T(Y)^5(Y)^T(Y)^a^g^t^t^c^t^c^a^t^5(Y)^A(Y)^a (Sequence C: used as ASO-6968 in the examples) A(Y)^T(Y)^T(Y)^g^a^g^g^g^t^a^g^g^5(Y)^A(Y)^a (Sequence D: used as ASO-7724 in the examples)

[0037] In the sequences A to D, "G(Y)" represents guanine in the structure of AmNA (a 2',4'-bridged nucleotide in which R is a methyl group in the general formula (1)), "A(Y)" represents adenine in the structure of AmNA, "T(Y)" represents thymine in the structure of AmNA, "5(Y)" represents 5-methylcytosine in the structure of AmNA, lowercase letters "a, t, c, g" represent unmodified DNA, and "^" represents a phosphorothioate bond.

[0038] Furthermore, when the nucleic acid molecule used in the present invention is an shRNA against mouse Greb1, a specific example of DNA encoding the shRNA is the base sequence shown in SEQ ID NO:8.

[0039] Furthermore, when the nucleic acid molecule used in the present invention is an RNA molecule, it may be designed to be produced in vivo. Specifically, DNA encoding the RNA molecule may be inserted into an expression vector for mammalian cells. Examples of such expression vectors include viral vectors such as retrovirus, lentivirus, adenovirus, adeno-associated virus, herpes virus, and Sendai virus, as well as animal cell expression plasmids.

[0040] [Target diseases] Since GREB1 is specifically expressed in sex hormone-insensitive GREB1-positive tumor cells and promotes the proliferation of these tumor cells, the therapeutic agent of the present invention can inhibit the proliferation of these tumor cells by suppressing the expression of GREB1. Therefore, the therapeutic agent of the present invention is used to treat hormone-insensitive GREB1-positive tumors.

[0041] In the present invention, a GREB1-positive tumor refers to a tumor formed by tumor cells in which GREB1 expression is observed. Whether or not a tumor is GREB1-positive can be confirmed by tissue immunization of collected tumor lesion tissue. Specifically, collected tumor lesion tissue is immunostained with an anti-GREB1 antibody, and if GREB1 expression is observed in 5% or more of the tumor lesion site, the tumor is determined to be GREB1-positive. Suitable examples of tumors to be treated with the therapeutic agent of the present invention include tumors in which 20% or more of the tumor lesion site has high GREB1 expression, and particularly preferably tumors in which 50% or more of the tumor lesion site has high GREB1 expression.

[0042] In addition, whether a tumor is GREB1-positive can also be determined by recovering RNA from the collected tumor tissue and measuring it by quantitative PCR. In this case, the presence or absence of GREB1 expression can be determined using the tumor and a non-tumor site of the same tissue from the same case as indicators. Specifically, if the amount of GREB1 in a cell lysate of tumor tissue is higher than the amount of GREB1 in a cell lysate of a tumor and a non-tumor site of the same tissue from the same case, the tumor is determined to be GREB1-positive.

[0043] Furthermore, in the present invention, a sex hormone-insensitive GREB1-positive tumor (a hormone-insensitive GREB1-positive tumor) refers to a tumor formed by GREB1-positive tumor cells that do not have the "property of promoting growth through increased GREB1 expression via hormone receptors in response to stimulation by sex hormones such as estrogen and androgen." Sex hormone-sensitive GREB1-positive tumors include, for example, breast cancer and ovarian cancer, whose growth is promoted by estrogen, and prostate cancer, whose growth is promoted by androgen, and these tumors are excluded from the treatment targets of the present invention.

[0044] In the present invention, specific examples of sex hormone-insensitive GREB1-positive tumors to be treated include hepatoblastoma, hepatocellular carcinoma, malignant melanoma, neuroblastoma, and small cell lung cancer.

[0045] Furthermore, the therapeutic agent of the present invention can be used not only for humans but also for mammals such as cows, pigs, dogs, cats, goats, rats, mice, and rabbits, and is preferably used as a pharmaceutical for humans.

[0046] [Administration mode] The administration route of the therapeutic agent of the present invention may be either oral or parenteral, as long as an antitumor effect is obtained, and may be appropriately determined depending on the type of active ingredient used. Specific administration routes of the therapeutic agent of the present invention include parenteral administration such as injection (intravenous injection, subcutaneous injection, intramuscular injection, peritoneal injection, local injection into the affected area, etc.) and suppository administration.

[0047] The dosage of the therapeutic agent of the present invention may be appropriately determined as a therapeutically effective amount depending on the type of active ingredient used, the dosage form, the degree of progression of the sex hormone-insensitive GREB1-positive tumor to be treated, etc. For example, when a nucleic acid molecule is used as the active ingredient, a single dose of the nucleic acid molecule may typically be about 0.1 mg / kg body weight to 100 mg / kg body weight, administered approximately once every 3 to 7 days.

[0048] The therapeutic agent of the present invention may be used alone, or may be used in combination with one or more other agents having antitumor activity and / or radiation therapy.

[0049] [Form of formulation] The therapeutic agent of the present invention is prepared into a formulation suitable for the type of active ingredient and the mode of administration. Examples of the formulation of the antitumor agent of the present invention include liquid formulations such as solutions, suspensions, emulsions, and injections.

[0050] The therapeutic agent of the present invention is formulated by adding pharmaceutically acceptable carriers and additives depending on the dosage form. For example, in the case of a liquid formulation, it can be formulated using physiological saline, a buffer solution, etc.

[0051] Furthermore, when a nucleic acid molecule is used as an active ingredient in the therapeutic agent of the present invention, it is desirable that the nucleic acid molecule be formulated together with a nucleic acid transfer adjuvant to facilitate its transfer into tumor cells. Specific examples of nucleic acid transfer adjuvant include lipofectamine, oligofectamine, RNAifect, liposomes, polyamines, DEAE-dextran, calcium phosphate, and dendrimers.

[0052] 2. Testing methods for hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma The testing method of the present invention is a testing method for predicting the presence or absence of hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma, and is characterized by comprising a step of measuring the expression level of GREB1 in liver tissue, skin tissue, or nerve tissue collected from a subject. According to the present invention, by using the expression level of GREB1 in liver tissue, skin tissue, or nerve tissue collected from a subject as an indicator, it becomes possible to diagnose whether a subject is suffering from hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma.

[0053] When hepatoblastoma or hepatocellular carcinoma is the test subject, the expression level of GREB1 can be measured in liver tissue collected from a subject suspected of having a neoplastic liver disease. Since hepatoblastoma is a malignant tumor that develops in the liver of children, subjects to be tested for hepatoblastoma preferably include children suspected of having a neoplastic liver disease.

[0054] When malignant melanoma is the subject of testing, the expression level of GREB1 can be measured in skin tissue collected from a subject suspected of having a neoplastic skin disease.

[0055] When neuroblastoma is the test subject, the expression level of GREB1 can be measured in neural tissue collected from a subject suspected of having a neoplastic neurological disease. Since neuroblastoma is a malignant tumor that is likely to occur in children, subjects to be tested for neuroblastoma in the testing method of the present invention are preferably children suspected of having a neoplastic neurological disease.

[0056] The expression level of GREB1 in each tumor tissue collected from a subject can be confirmed by tissue immunization of the collected tissue fragment (lesion tissue fragment). Specifically, the collected tumor tissue fragment is immunostained with an anti-GREB1 antibody, and if GREB1 expression is observed in 5% or more of the suspected tumor area, it is determined that GREB1 is expressed. Furthermore, if GREB1 expression is observed in 20% or more of the suspected tumor area, it is determined that the expression level of GREB1 is high, and if GREB1 expression is observed in 50% or more of the suspected tumor area, it is determined that the expression level of GREB1 is particularly high.

[0057] The expression level of GREB1 in each tumor tissue collected from a subject can also be measured by recovering RNA from the tissue fragment (lesion tissue fragment) and performing quantitative PCR. In this case, the expression level of GREB1 can be measured using the tumor and a non-tumor site of the same tissue in the same case as indicators. Specifically, if the amount of GREB1 in a cell lysate from a suspected tumor site is higher than the amount of GREB1 in a cell lysate from a non-tumor site in the same case, it is determined that GREB1 is expressed.

[0058] In the testing method of the present invention, it is presumed that the higher the expression level of GREB1 in each of the tumor tissues, the higher the possibility of the patient being affected with hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma.

[0059] Furthermore, the present invention provides a reagent for detecting GREB1 as a test reagent for carrying out the above-mentioned test method.

[0060] Reagents for detecting GREB1 include, for example, anti-GREB1 antibodies and fragments thereof. The anti-GREB1 antibodies may be labeled with biotin, a fluorescent label, magnetic beads, or the like, as necessary.

[0061] Other examples of reagents for detecting GREB1 include primers that can hybridize with GREB1 cDNA or GREB1 mRNA. When using such primers as reagents for detecting GREB1, the test reagent of the present invention may contain, in addition to the primers, reagents necessary for PCR. [Example]

[0062] The present invention will be described in detail below based on experimental data, but the present invention is not limited thereto.

[0063] Hereinafter, the cells used in the test may be referred to as "cells expressing X," etc., but "cells expressing X" means cells transfected with protein X and artificially transformed to express protein X. For example, "X293T cells expressing GFP-GREB1" means X293T cells transfected with protein GFP-GREB1 and transformed to overexpress GFP-GREB1.

[0064] Furthermore, hereinafter, proteins may be referred to as "AB" (for example, GFP-GREB1), which means that the protein is a protein in which peptide A is fused to peptide B.

[0065] The cells transfected with various proteins used in the following tests were prepared according to known genetic engineering techniques.

[0066] 1. Test materials and methods 1-1. Cells and antibodies HepG2, CHP212, and SKNDZ cells were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). MCF7, HLE, Huh7, HLF, NBTU110, SKMEL28, Mewo, G361, and COLO679 cells were purchased from the Japanese Collection of Research Bioresources (JCRB, Osaka, Japan). Lenti-X TM 293T (X293T) cells were purchased from Takara Bio Inc. (Japan). Huh6 cells were provided by Dr. H. Okuyama (Osaka University, Japan). SNU387, SNU449, BMEL, and HepG2 cells were provided by Dr. T. Kodama (Osaka University, Japan).

[0067] HepG2 cells were grown in Eagle's minimum essential medium (EMEM) containing 10% fetal bovine serum (FBS), non-essential amino acids, and glutamax. HLE, Huh6, X293T, Huh7, SNU387, SNU449, and HLF cells were grown in Dulbecco's modified Eagle's medium (DMEM) containing 10% FBS. MCF7 cells were grown in DMEM containing 10% FBS, non-essential amino acids, and 1 mM sodium pyruvate. BMEL cells were grown in DMEM / F12 containing 10% FBS, 2.5 mM L-glutamine, 0.5 mM sodium pyruvate, 50 ng / ml EGF, 30 ng / ml IGF-II, and 10 μg / ml insulin. SKMEL28, Mewo, and G361 cells were grown in EMEM containing 10% FBS and non-essential amino acids. CHP212 and SKNDZ cells were grown in DMEM containing 10% FBS and non-essential amino acids. OLO679 cells were grown in RPMI 1640 (RPMI) containing 10% FBS and 2 mM L-glutamine.

[0068] Anti-GREB1 antibody (mouse monoclonal), anti-phosphorylated histone H3 (ser10), and anti-acetyl-histone H4 antibodies were purchased from Merck Millipore (Billerica, MA, USA). Anti-HSP90, anti-Cyclin A, anti-Cyclin B, anti-Smad2 / 3 (mouse monoclonal), anti-β-catenin, and anti-clathrin antibodies were purchased from BD Biosciences (San Jose, CA, USA). Anti-GREB1 antibody (rabbit polyclonal), anti-Smad4, anti-p300, and anti-GFP (mouse monoclonal) were purchased from Santa Cruz Santa Cruz Biotechnology (Dallas, TX, USA). Anti-Smad2 / 3 antibodies (rabbit monoclonal), anti-phosphorylated SMAD2 (Ser465 / 467) / SMAD3 (Ser423 / 425) antibodies, anti-cleaved caspase 3 antibodies, anti-PARP1 antibodies, anti-Hystone H3 antibodies, anti-YAP1 antibodies (rabbit monoclonal), anti-Ki67 antibodies (rabbit monoclonal), anti-c-Myc antibodies (rabbit monoclonal), anti-Smad2 / 3 antibodies (rabbit monoclonal for Western blotting), anti-Axin2 antibodies, and anti-MITF antibodies (for Western blotting) were purchased from Cell Signaling Technology (Beverly, MA, USA). Anti-β-tubulin antibodies, anti-β-actin antibodies, anti-phosphorylated β-catenin (pTyr654) antibodies, and anti-MITF antibodies (for immunohistochemistry) were purchased from Sigma-Aldrich (Steinheim, Germany). Anti-GFP antibody (rabbit polyclonal) was purchased from Life Technologies / Thermo Fisher Scientific (Carlsbad, CA, USA). Anti-Smad2 / 3 antibody (rabbit monoclonal) was purchased from Abcam (Cambridge, UK). Anti-FLAG and anti-GAPDH antibodies were purchased from WAKO (Tokyo, Japan). Anti-mouse DLK1 antibody (rabbit monoclonal) was purchased from R&D Systems (Minneapolis, MN, USA).

[0069] 1-2.RNA sequence analysis Libraries were generated from HepG2 cells transfected with control siRNA or β-catenin siRNA using the TruSeq Stranded mRNA Sample Prep kit (Illumina, San Diego, CA). Sequencing was performed on an Illumina HiSeq 2500 platform in 75-base single-end mode. Base calling was performed using CASAVA 1.8.2 software (Illumina). Sequences were mapped to the human reference genome sequence (hg19) using TopHat v2.0.13 in conjunction with Bowtie2 ver. 2.2.3 and SAMtools ver. 0.1.19. Cuffnorm version 2.2.1 was used to calculate fragments per kilobase per million exons (FPKMs).

[0070] Of the 23,284 genes, 8,929 were identified as genes with normalized FPKM values ​​greater than 3.0. Seventy-six genes were downregulated by more than threefold in β-catenin knockdown cells compared to control cells (P < 0.001 [Welch's t-test]). GREB1 binding peaks were downloaded from the ENCODE Transcription Factor Binding Site Profiles resource (http: / / amp.pharm.mssm.edu / Harmonizome / gene#set / TCF7L2#HepG2#hg19#1 / ENCODE+Transcription+Factor+Binding+Site+Profiles) as the TCF7L2#HepG2#hg19#1 geneset (465 genes). Finally, 11 downregulated genes were identified.

[0071] 1-3. Open Source Clinical Data Analysis Clinical data of cancer patients were analyzed and retrieved using The Cancer Genome Atlas (TCGA) datasets available from the websites 'depmap portal (https: / / depmap.org / portal / ),' TCGA SpliceSeq (http: / / projects.insilico.us.com / TCGASpliceSeq), and 'R2: Genomics Analysis and Visualization Platform (http: / / r2.amc.nl)'. All gene expression data, P values, and r values ​​were downloaded.

[0072] 1-4. Patients and cancer tissues Hepatoblastoma tissues (n = 11) obtained from patients with hepatoblastoma who underwent surgical treatment at Osaka University Hospital between January 2008 and March 2015 were used in this study. Patient ages ranged from 0 to 16 years (median age 3 years).

[0073] In addition, neuroblastoma tissues (13 patients) and cutaneous malignant melanoma tissues (50 patients) from patients who underwent surgical treatment at Osaka University Hospital, and hepatocellular carcinoma (HCC) tissues (stages I to IV) from 210 patients who underwent surgical treatment at Kobe University Hospital were also used in this study.

[0074] The resected specimens were macroscopically observed, and the tumor location and size were measured. The specimens were fixed in 10% formalin by volume, and paraffin-embedded blocks were prepared for histological analysis. The specimens used in the study were sectioned to a thickness of 4 μm and stained with hematoxylin and eosin (H&E), immunoperoxidase, immunoalkaline phosphatase, etc., and subjected to each analysis.

[0075] The studies described in Sections 2-1 to 2-8 below were conducted with the approval of the Osaka University School of Medicine Ethics Committee (No. 13552). The studies described in Sections 2-9 to 2-11 below were conducted in accordance with the Declaration of Helsinki and with the approval of the Osaka University School of Medicine Ethics Committee (No. 13455, 19292) and the Kobe University School of Medicine Ethics Committee (No. B190073). Written informed consent was obtained from all patients.

[0076] 1-5. Evaluation of tumor formation by HTVi (hydrodynamic tail vein injection) pT2BH-YAPS127A (20 μg), pT2BH-ΔN90β-catenin-mutLuc3 (20 μg), pT3-EF1a-cMET (20 μg), and GFP2ALuc were diluted in 2.5 ml of saline and injected into the tail vein of 8-9 week-old wild-type C57BL6 / N male mice within 5-7 seconds in the presence of Greb1 shRNA#3 (SEQ ID NO: 8) (20 μg) or control and pLIVE-SB13 (8 μg). Livers were harvested at specified times post-injection (or at the time of disease) to examine various parameters related to tumor formation. For specific methods, see Tao, J. et al. (2014). Gastroenterology 147, 690-701 and Tward, AD et al. (2007). Proc Natl Acad Sci USA 104, 14771-14776. ΔN90β-catenin is a β-catenin mutant in which amino acids 1 to 90 of human β-catenin are deleted, and YAPS127A is a YAP mutant in which serine at position 127 of YAP is replaced with alanine.

[0077] 1-6. Preparation of antisense oligonucleotides against GREB1 Phosphothioated 15-mer antisense oligonucleotides (ASOs) containing AmNA monomers were prepared (synthesized by GeneDesign (Ibaraki, Japan)). The sequences of the synthesized ASOs are shown in Tables 8 and 9. Note that, throughout this specification, "hGREB1-6424-AmNA(15)" will be abbreviated simply as "ASO-6424" or "GREB1 ASO-6424." Other ASOs will be abbreviated in the same manner. HepG2 or COLO679 cells were transfected with 10 nM ASOs using RNAiMAX (Invitrogen, Carlsbad, CA, USA). Experiments were performed 36 to 48 hours after transfection.

[0078] 1-7. In vivo ASO-treated xenograft liver tumorigenesis assay HepG2 cell pellet (1 × 10 7 The cells were suspended in 100 μl of high-concentration Matrigel (Corning) and transplanted into 8-week-old male BALB / cAJcl-nu / nu mice (nude mice; CLEA Japan) under anesthesia by direct injection into the liver. Starting on day 0, ASO (50 μg / body; approximately 2.5 mg / kg) was administered subcutaneously twice weekly. Twenty-seven days after transplantation, the mice were euthanized, and tumors were collected and subjected to histological analysis.

[0079] 1-8. Protein knockdown by siRNA The base sequences of the siRNAs used in the analysis are shown in Table 1.

[0080] [Table 1]

[0081] HepG2, Huh6, Hep3B, JHH7, and SKMEL28 cells were transfected with each siRNA (10 nm) using RNAiMAX (Life Technologies / Thermo Fisher Scientific). Cells were used 48 to 120 hours after transfection.

[0082] 1-9. Immunohistochemical analysis Immunohistochemical staining of tissue specimens was performed using DakoReal TM EnVision TM The analysis was performed using the Detection System (Dako, Carpentaria, CA, USA) and the Warp Red™ Chromogen Kit (Biocare Medical, Concord, CA, USA) according to the manufacturer's recommendations. Specifically, antigen retrieval of tissue specimens was performed using a decloaking chamber (Biocare Medical, Walnut Creek, CA, USA). Endogenous peroxidase activity was blocked with 3% H2O2-methanol for 15 minutes, and sections were then incubated with goat serum for 1 hour to block nonspecific antibody binding. The tissue specimens were then incubated with mouse anti-GREB1 antibody (1:100), mouse anti-β-catenin antibody (1:100), or rabbit anti-YAP1 antibody (1:100) at 4°C for 16 hours, followed by horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG or goat anti-rabbit IgG for 1 hour. Diaminobenzidine (DAB) (Dako) was used as a chromogen to visualize the binding of mouse anti-GREB1, mouse anti-β-catenin, or rabbit anti-YAP1 antibodies to tissue sections. The tissue sections were counterstained with 0.1% (w / v) hematoxylin. The staining areas for β-catenin, GREB1, and YAP were graded into three categories: <5%, 5-30%, and 30-95%. Tumors with a staining area of ​​5% or greater relative to the tumor lesion area (positive staining) were considered positive.

[0083] 1-10. Immunofluorescence staining Cells grown on glass coverslips were fixed in 4% (w / v) paraformaldehyde in PBS at room temperature for 10 minutes and permeabilized in 0.2% (w / v) Triton X-100 and 2 mg / ml BSA in PBS for 10 minutes. Cells grown in 3D cultures were fixed in 4% (w / v) paraformaldehyde in PBS at room temperature for 30 minutes and blocked in 0.5% (w / v) Triton X-100 and 40 mg / ml BSA in PBS for 30 minutes. Fixed and permeabilized cells were incubated with primary antibodies for 3 hours at room temperature or overnight at 4°C, followed by secondary antibody incubation according to the manufacturer's protocol (Molecular Probes, Carlsbad, CA). Samples were analyzed using an LSM880 laser confocal microscope (Carl-Zeiss, Jena, Germany).

[0084] 1-11. Visualization of gene expression using heat maps The expression levels of each gene were normalized to GAPDH mRNA and calculated as a ratio to normal liver. Data were normalized by min-max normalization, and colored heat maps were generated using Excel software (Microsoft, Redmond, WA, USA).

[0085] 1-12. Cell proliferation assay Plate 1.0 x 10 cells 4 Cells were seeded at 100 cells / mL and cultured for 12 hours. The medium was replaced with 10% serum. Cell counts were performed every 48 hours for up to 10 days. Cell counts were performed using Cyquant NF assays (Life Technologies / Thermo Fisher Scientific) according to the manufacturer's recommended protocol.

[0086] 1-13.Quantitative RT-PCR For quantitative RT-PCR, the primers shown in Table 2 were used. [Table 2]

[0087] 1-14. Complex formation and immunoprecipitation HepG2 cells (100 mm diameter dish) were lysed in 400 μl of cell lysis buffer (10 mM Tris-HCl [pH 7.4], 140 mM NaCl, 5 mM EDTA, 1% NP40, 25 mM NaF, 20 mg / ml leupeptin, 20 mg / ml aprotinin, and 10 mM PMSF). The resulting lysates were immunoprecipitated with anti-Smad2 / 3 antibodies, and the immunoprecipitates were probed with the indicated antibodies.

[0088] X293T cells (60 mm dishes) transfected with HA-FLAG-GREB1, GFP-GREB1 mutants, GFP-Smad2 mutants, HA-Smad3, GFP-Smad3, GFP-Smad4, or GFP-Smad7 were lysed in 400 μl of cell lysis buffer to examine the state of the GREB1, Smad3, Smad4, and Smad7 complexes. The resulting lysates were immunoprecipitated with anti-GFP antibodies, and the immunoprecipitates were probed with the indicated antibodies.

[0089] The amino acid sequences of the fusion proteins used in this test are as follows: the amino acid sequence of the HA portion is SEQ ID NO: 85; the amino acid sequence of the FLAG portion is SEQ ID NO: 86; the amino acid sequence of the GFP portion is SEQ ID NO: 87; the GREB1 mutant (mouse GREB1 Δ310-319 (ΔNLS)) is a GREB1 mutant in which positions 310-319 of mouse GREB1 have been deleted, and the amino acid sequence of this portion is SEQ ID NO: 88; the GREB1 mutant (mouse GREB1 1-666(N)) is a GREB1 mutant consisting of the amino acid sequence of positions 1-666 of mouse GREB1, and the amino acid sequence of this portion is SEQ ID NO: 89; the GREB1 mutant (mouse GREB1 667-1333(I)) is a GREB1 mutant consisting of the amino acid sequence of positions 667-1333 of mouse GREB1, and the amino acid sequence of this portion is SEQ ID NO: 90; the GREB1 mutant (mouse GREB1 The human Smad2 mutant (human Smad2 1-265(N)) is a Smad2 mutant consisting of the amino acid sequence of positions 1-265 of human Smad2, and its amino acid sequence is SEQ ID NO: 94. The human Smad2 mutant (human Smad2 266-467(C)) is a Smad2 mutant consisting of the amino acid sequence of positions 266-467 of human Smad2, and its amino acid sequence is SEQ ID NO: 95. The human Smad2 mutant (human Smad2 266-467(C)) is a Smad2 mutant consisting of the amino acid sequence of positions 266-467 of human Smad2, and its amino acid sequence is SEQ ID NO: 96. The amino acid sequence of the Smad4 portion is SEQ ID NO: 97, and the amino acid sequence of the human Smad7 portion is SEQ ID NO: 98. In other tests, each fusion protein was also prepared by linking the above amino acid sequences in combination.

[0090] GST pull-down assay To analyze the binding of recombinant Smad2 / MH1 or Smad2 / MH2 to GREB1, whole cell lysates from HepG2 cells or Huh6 cells expressing GFP-GREB1 were incubated with 20 μg of glutathione S-transferase (GST), GST-Smad2 / MH1, or GST-Smad2 / MH2 bound to glutathione-Sepharose beads for 1 h. After precipitation, the beads were washed three times with cell lysis buffer (10 mM Tris-HCl [pH 7.4], 140 mM NaCl, 5 mM EDTA, 1% NP40, 25 mM NaF, 20 mg / ml leupeptin, 20 mg / ml aprotinin, and 10 mM PMSF), and the resulting precipitates were probed with anti-GREB1 or anti-GFP antibodies.

[0091] The amino acid sequences of the fusion proteins used in this test are as follows: the amino acid sequence of the GST portion is SEQ ID NO: 99; the amino acid sequence of Smad2 / MH1 is SEQ ID NO: 100; and the amino acid sequence of Smad2 / MH2 is SEQ ID NO: 101.

[0092] 1-16. Labeling of cells with ethynyluridine (EU) GFP-Smad3-expressing cells expressing or not expressing HA-FLAG-GREB1 were treated with 1 mM EU for 30 min, then fixed and detected using the Click-iT RNA Alexa Fluor 594 Imaging Kit (Thermo Fisher Scientific, Waltham, MA, USA).

[0093] 1-17. Xenograft tumor formation analysis Five-week-old male BALB / cAnNCrj-nu nude mice (Charles River Laboratory Japan Inc, Osaka, Japan) were anesthetized with medetomidine (0.3 mg / kg body weight) and midazolam (4 mg / kg body weight), and then HepG2 cells (7 × 10 cells) suspended in 100 μl of high-concentration Matrigel (Corning, NY, USA) were introduced into the mice. 6The mice were then sacrificed 28 days after transplantation, and the area containing the transplanted cells was measured and processed for immunohistochemical analysis. All animal experiments in this study were conducted under the approval of the Osaka University Animal Care and Use Committee (No. 2867).

[0094] 1-18. Xenograft subcutaneous tumor formation analysis Five-week-old BALB / cAJcl-nu / nu mice (nude mice; CLEA Japan) were anesthetized with a combination of medetomidine (0.3 mg / kg) and midazolam (4 mg / kg). Hep3B or JHH7 cells (1 × 10 7 Hep3B cells were suspended in 150 μl of high-concentration Matrigel (Corning) and subcutaneously implanted into mice. ASO (50 μg / body; approximately 2.5 mg / kg) was administered subcutaneously twice weekly starting three days after implantation. Mice implanted with Hep3B cells were euthanized 6.5 weeks after implantation. Mice implanted with JHH7 cells were euthanized 2 weeks after implantation. Tumors were collected from euthanized mice and subjected to histological analysis. All animal experiments in this study were conducted under the approval of the Osaka University Animal Care and Use Committee (No. 26-032-048).

[0095] 1-19. Creation of knockout cells The target sequence for human GREB1, 5'-CTTCTCGGTGTTGAAGCCGA-3' (SEQ ID NO: 102), was designed using CRISPR Genome Engineering Resources (http: / / www.genome-engineering.org / crispr / ). Plasmids expressing hCas9 and single-guide RNA (sgRNA) were constructed by ligating the designated oligonucleotides into the BbsI site of pX330 (addgene #42230). Plasmid pX330, carrying the sgRNA sequence targeting GREB1 or MITF and blasticidin resistance, was transfected into HepG2, Hep3B, JHH7, or COLO679 cells using Lipofectamine LTX reagent (Life Technologies / Thermo Fisher Scientific) according to the manufacturer's recommended protocol. GREB1 knockout or MITF knockout cells were selected by culturing in medium containing 5 μg / mL blasticidin S for 2 days. Single colonies were then picked, mechanically dissociated, and replated into individual wells of 24-well plates.

[0096] To generate Smad2 and Smad3 (Smad2 / 3) double knockout cells, a plasmid incorporating hCas9 and dual guide RNAs was designed and synthesized by VectorBuilder Inc. (Guangzhou, China). The target sequences of the gRNAs for human Smad2 and Smad3 were 5'-TATATTGCCGATTATGGCGC-3' (SEQ ID NO: 103) and 5'-GGAATGTCTCCCCGACGCGC-3' (SEQ ID NO: 104), respectively. The pRP[CRISPR] plasmid carrying the dual guide RNA sequences targeting Smad2 / 3 was transfected into HepG2 cells, and single colonies were then picked, mechanically dissociated, and replated into individual wells of a 24-well plate.

[0097] 1-20. Chromatin immunoprecipitation (ChIP) assayConfluent HepG2 cells in a 10 cm dish were stimulated with or without 5 ng / ml TGFβ1 for 3 hours. Subsequently, cells were crosslinked with 1% formaldehyde at room temperature for 10 minutes, and the crosslinking reaction was stopped with 0.125 M glycine. After washing three times with cold PBS, cells were detached and collected. Cells were then pelleted by centrifugation and lysed in sodium dodecyl sulfate (SDS) lysis buffer (50 mM Tris / HCl [pH 8.0], 10 mM EDTA, and 0.5% SDS). DNA was sheared by sonication to a size range of 200–1000 bp. The sheared chromatin supernatant was diluted with ChIP dilution buffer (16.7 mM Tris / HCl [pH 8.0], 167 mM NaCl, 1.2 mM EDTA, and 1.1% Triton X-100) containing protease and phosphatase inhibitors and precleared with salmon sperm DNA / protein A-agarose beads (Millipore, Billerica, MA, USA). The supernatant was then incubated for 12 hours at 4°C with anti-Tcf-4, anti-β-catenin, anti-acetylated histone H4, or negative control IgG (Diagenode, Liege, Belgium). The immune complexes adsorbed to the salmon sperm DNA / protein A-agarose beads were washed once with high-salt buffer (20 mM Tris / HCl [pH 8.0], 500 mM NaCl, 0.1% SDS, 1% Triton X-100, and 2 mM EDTA), once with LiCl buffer (10 mM Tris / HCl [pH 8.0], 0.25 M LiCl, 1 mM EDTA, 1% deoxycholic acid, and 1% Nonidet P-40), and four times with TE buffer (10 mM Tris / HCl (pH 8.0), and 1 mM EDTA).The immune complexes were then eluted from the beads and reversed cross-links by incubation at 65°C for 4 hours in elution buffer (50 mM Tris / HCl [pH 8.0], 10 mM EDTA, 1% SDS). The samples were then treated with RNase A at 37°C for 30 minutes, followed by treatment with proteinase K at 55°C for 1 hour. DNA was then purified by phenol-chloroform extraction and subjected to PCR using the primers listed in Table 3. [Table 3]

[0098] 1-21. Plasmid construction and transfection with lentivirus carrying cDNA Using standard recombinant DNA techniques, we engineered plasmids containing full-length GREB1 or various mutants of GREB1. To generate NSL (nuclear localization sequence)-fused GREB1 mutants, three copies of the NSL sequence from the SV40 T antigen were ligated to the N-terminus of the GFP-GREB1 mutant.

[0099] A lentiviral vector was constructed by subcloning GFP and pEGFPC1-GREB1 into CSII-CMV-MCS-IRES2-Bsd, a kind gift from Dr. H. Miyoshi (RIKEN BioResource Center, Ibaraki, Japan).

[0100] The lentivirus vector was then transfected into X293T cells together with the packaging vectors pCAG-HIV-gp and pCMV-VSV-G-RSV-Rev using FuGENE HD transfection reagent (Roche Applied Science, Basel, Switzerland) to produce lentivirus. 5 × 10 lentivirus vectors were then transfected into each well of a 12-well plate. 4HepG2 cells were cultured in 1000 cells, treated with lentivirus and 10 μg / ml polybrene, centrifuged at 1200 × g for 30 minutes, and then incubated for 24 hours to generate HepG2 cells stably expressing GFP or GFP-GREB1 (GREB1 linked to GFP).

[0101] pT3-EF1a-cMET was purchased from Addgene (Cambridge, MA, USA). The pLIVE-SB13 vector was a kind gift from R. Toru Okamoto (Osaka University). pT2BH-ΔN90β-catenin-Luc was constructed by in-fusion cloning of a CAG promoter, a human CTNNB1 sequence lacking amino acids 1-90, and luciferase into the pT2BH vector (Addgene). pT2BH-YAPS127A was constructed by inserting a FLAG-tagged human YAPS127A fragment between the EcoRI and NotI sites of the pT2BH vector. pT2BH-GFP2ALuc mouse Greb1 shRNA was constructed by inserting a U6 promoter and a mGreb1 shRNA fragment amplified from Mission shRNA (TRCN0000216019) (Sigma-Aldrich) between the PstI and HindIII sites of pT2BH-GFP2ALuc.

[0102] 1-22.Statistical analysis Each test was performed at least three times. Statistical analysis was performed using Excel (Microsoft, Redmond, WA, USA) and GraphPad Prism 7 (GraphPad Software, La Jolla, CA, USA). A P value of less than 0.05 was considered statistically significant.

[0103] 2. Test Results 2-1.GREB1 is a downstream target gene of Wnt / β-catenin signaling in human hepatoblastoma To elucidate the mechanism of hepatoblastoma tumorigenesis, we screened for novel downstream target genes of Wnt / β-catenin signaling using HepG2 hepatoblastoma cells, which harbor a truncated mutation in exons 3 and 4 of the β-catenin gene. RNA sequencing was performed on HepG2 cells transfected with control siRNA or β-catenin siRNA (β-catenin siRNA). Of the 8,929 genes, 76 were selected as candidate genes based on their high expression levels (FPKM ≥ 3) and a ≥3-fold decrease in expression in β-catenin siRNA-transfected cells (β-catenin knockdown cells) compared with control siRNA-transfected cells (see Figure 1a). Furthermore, using the ENCODE Transcription Factor Binding Site Profiles (TCF7L2#HepG2#hg19#1) gene set, we narrowed down the candidate genes based on the presence of TCF7L2 (TCF4) DNA binding sites found by ChIP-seq analysis in HepG2, and selected 11 genes (see Figure 1a and Table 4). The lower panel of Figure 1a shows a heat map of candidate genes whose expression levels were altered in control cells and β-catenin knockdown cells. The genes shown on the left of the lower panel of Figure 1a randomly represent 11 of the 76 altered candidate genes.

[0104] [Table 4]

[0105] Most of the selected candidate genes, including NKD1, LGR5, SP5, ZNRF3, RNF43, Axin2, CCND1, and DKK1, were known targets of Wnt / β-catenin signaling. One of the candidate genes, GREB1, is an estrogen-responsive gene in the estrogen receptor-regulated pathway and is known to be involved in hormone-dependent cancer cell proliferation in breast and prostate cancers. However, whether it is a downstream target gene of Wnt / β-catenin signaling and involved in tumorigenesis in cancers that do not express estrogen receptors has not been elucidated. Therefore, we focused on GREB1 for further analysis.

[0106] Chromatin immunoprecipitation assays revealed that TCF4 and β-catenin form a complex with the TCF4 binding site located at -443 to -448 in the 5'-upstream region of the human GREB1 gene (see Figure 1b). Indeed, β-catenin knockdown in HepG2 cells resulted in decreased GREB1 expression and protein levels (see Figure 1c). On the other hand, real-time PCR analysis of GREB1 mRNA expression after 48 hours of treatment with the estrogen receptor antagonist ICI-182,786 confirmed that GREB1 expression was not affected by ICI-182,786 (see Figure 2a). In bipotential mouse fetal liver (BMEL) cells, treatment with CHIR99021, which activates β-catenin signaling, increased GREB1 expression, as well as Axin2 expression (see Figure 2b). In Huh6 cells, a hepatoblastoma cell line harboring a G34V somatic mutation in the CTNNB1 (β-catenin) gene, β-catenin knockdown reduced GREB1 expression (see Figure 2c). Huh6 cells expressed less GREB1 mRNA than HepG2 cells (see Figure 2d). In MCF7 cells, an estrogen receptor-positive breast cancer cell line, GREB1 expression was dramatically reduced by ICI-182,786 treatment. However, treatment with CHIR99021, which activates Wnt / β-catenin signaling, increased Axin2 expression but not GREB1 expression, either in the presence or absence of ICI-182,786 (see Figure 2e). Furthermore, in the human hepatocellular carcinoma cells HLE, SNU-387, SNU-449, and Huh7, CHIR99021 treatment increased Axin2 expression, but not GREB1 expression (see Figure 2f). These results suggest that GREB1 is a downstream target gene of the Wnt / β-catenin signal specific to hepatoblastoma cells or immature hepatic progenitor cells.

[0107] We analyzed 11 cases of hepatoblastoma. Specifically, hepatoblastoma tissues from these cases were stained with anti-GREB1 antibody and hematoxylin. Results showed that GREB1 expression was observed in the tumor lesions in 10 of the 11 cases (90.9%), but not in non-tumorous areas (Table 5, Figure 1(d)). The percentage of the total tumor area stained by GREB1 immunostaining was classified into three categories: <5% (negative), 5-30% (low), and 30-95% (high). Cases with ≥5% staining were considered GREB1 positive. The region-specific expression pattern of GREB1 in hepatoblastoma tissue tended to positively correlate with β-catenin accumulation in serial sections (Figure 1(e)). Hepatoblastoma tissues were histologically classified into two types: one type composed of solid, non-polarized cells and the other type composed of tubular, polarized cells. Among these, GREB1 was specifically and highly expressed in the former histological type (solid, nonpolarized cells). Furthermore, 11 hepatoblastoma tissues were stained with anti-β-catenin antibody and hematocytosis. The results confirmed that β-catenin expression was elevated in tumor lesions compared with non-tumorous areas (see Figure 2g). The percentage of the area stained by β-catenin immunostaining relative to the total area of ​​tumor lesions was classified into three categories: <5% (negative), 5-30% (low), and 30-95% (high). Cases showing ≥5% staining were considered β-catenin positive.

[0108] [Table 5]

[0109] Furthermore, we analyzed the correlation between GREB1 gene expression and the expression of Wnt / β-catenin signaling target genes in hepatoblastoma using a public dataset of hepatoblastoma patients obtained from R2 genomics and the visualization platform database (http: / / r2.amc.nl). Two parameters, GREB1 gene expression and target gene expression, were available for 50 tumor lesions and 5 non-tumorous regions in hepatoblastoma cases. Analysis revealed that GREB1 mRNA was significantly upregulated in tumor lesions compared with non-tumorous regions (see Figure 1f). Furthermore, a significant positive correlation was observed between the expression levels of Wnt / β-catenin signaling target genes, such as Axin2, DKK1, NKD1, and glutamine synthetase (GS), and GREB1 expression (see Figure 1g). On the other hand, the expression levels of downstream estrogen receptor target genes, such as PRLR and XBP1, did not change significantly (see Figure 2i). These results confirmed that in hepatoblastoma, activation of Wnt / β-catenin signaling, rather than estrogen receptor signaling, increases GREB1 expression.

[0110] According to a public dataset of hepatoblastoma patients, GREB1 expression was not significantly altered in hepatoblastoma cases with mutations or deletions in exon 3 or exon 4 of the CTNNB1 gene compared with cases without mutations (see Figure 2j). Expression of Axin2 and GS, downstream target genes of β-catenin, was also unchanged regardless of the presence or absence of CTNNB1 gene abnormalities. These results demonstrate that GREB1 expression in hepatoblastoma correlates with β-catenin signaling activity but is not necessarily associated with mutations in exon 3 or exon 4 of the CTNNB1 gene. Therefore, it is possible that other CTNNB1 gene mutations or β-catenin signaling activation independent of CTNNB1 gene mutations may be occurring.

[0111] 2-2.GREB1 expression is involved in the proliferation of hepatoblastoma cells As previously reported, β-catenin knockdown reduced cell proliferation in HepG2 cells (Figure 3a). Furthermore, exogenous GREB1 expression partially rescued the β-catenin knockdown phenotype (Figure 3a). To elucidate the role of GREB1 in HepG2 cells, we knocked down GREB1 using two different siRNAs. Lysates from GREB1-knockdown cells were probed with anti-GREB1, anti-Axin2, anti-β-catenin, and anti-HSP90 antibodies. The siRNAs (GREB1 #1 siRNA and GREB1 #2 siRNA) knocked down GREB1 (Figure 3b). However, GREB1 knockdown did not affect the expression of ER signaling target genes PRLR and XBP1 (Figure 3c), nor did it affect the expression of β-catenin or Axin2 (Figure 3b). These results suggest that GREB1 is not a gene that functions upstream of ER and Wnt / β-catenin signaling.

[0112] Mock (GREB1-free vector only) or GREB1-transfected HepG2 cells were transfected with control siRNA or two different siRNAs (GREB1 #1 siRNA and GREB1 #2 siRNA) and cultured in two-dimensional (2D) cultures. The relative cell numbers were measured over time using a Cyquant assay. GREB1 knockdown reduced the proliferation of HepG2 cells in two-dimensional culture, whereas transfection of GREB1-transfected HepG2 cells with GREB1 siRNA (GREB1 #1 siRNA / GREB1 and GREB1 #2 siRNA / GREB1) did not affect HepG2 cell proliferation (see Figure 4a).

[0113] Furthermore, HepG2 cells transfected with control siRNA or siRNA (GREB1 #2 siRNA) were cultured in tertiary Matrigel for 5 days. The cells were stained with phalloidin and the area of ​​the spheres was measured (n = 20). GREB1 knockdown reduced the sphere area by half (Figure 4b). In contrast, GREB1 knockdown increased the percentage of spheres with polarized lumens (Figure 4b). The percentage of spheres with polarized lumens was calculated as the percentage of spheres with an F-actin-positive central microlumen relative to the total number of spheres. These results suggest that GREB1-knockdown HepG2 cells were transformed into a differentiated state with epithelial polarization. These results are consistent with the finding that GREB1 is specifically expressed in solid, non-polarized cells in hepatoblastoma tissue (see Figure 2h). Indeed, knockdown of GREB1 significantly reduced the expression of undifferentiated hepatic progenitor cell marker genes DLK1, AFP, and PEG3 (Fig. 4c). Consistent with these results, a public dataset of hepatoblastoma patients also showed a significant positive correlation between the expression levels of hepatoblastoma marker genes such as DLK1 and TACSTD1 and the expression levels of GREB1 (Fig. 3d).

[0114] Furthermore, in Huh6 cells, which express low levels of GREB1, transfection with GFP or GFP-GREB1 resulted in increased proliferation in both 2D and 3D cultures when GFP-GREB1 was introduced to overexpress GREB1 (see Figure 3e and f). These results demonstrate that GREB1 expression in hepatoblastoma cells contributes to the improvement of proliferation.

[0115] HepG2 cells transfected with control siRNA or GREB1 siRNA (GREB1 #1 siRNA and GREB1 #2 siRNA) were cultured in medium containing 1% FBS for 1 day, and the cell lysates were probed with anti-cyclin A, anti-cyclin B, anti-phosphorylated histone H3, anti-histone H3, anti-GREB1, and anti-HSP90 antibodies. The results confirmed that knockdown of GREB1 downregulated the expression of cell cycle regulators, including cyclin A, cyclin B, and phosphorylated histone H3 (see Figure 4d).

[0116] Furthermore, analysis using public databases revealed that there was a positive correlation between the expression levels of GREB1 and those of MKI67, GMMN, and PCNA in hepatoblastoma cells (see Figure 4e).

[0117] Furthermore, the phenotypes of hepatoblastoma marker expression, cell proliferation, and cell cycle phenotypes induced by GREB1 knockdown in HepG2 cells were similarly confirmed in HepG2 cells in which GREB1 was knocked out using CRISPR / Cas9, and these phenotypes were restored by the expression of exogenous GREB1 (see Figure 3, g to j).

[0118] Furthermore, HepG2 cells were transfected with control siRNA or siRNA (GREB1 #1 siRNA and GREB1 #2 siRNA) and cultured for 2 days in 0.1% FBS-containing medium (with or without the caspase inhibitor Z-VAD). Cell viability was assessed by staining with propidium iodide (PI, viable cells) and Hoechst 33342 (nuclei). The results confirmed that knockdown of GREB1 increased cell death in HepG2 cells when cultured in medium without Z-VAD (see Figure 4f).

[0119] Furthermore, HepG2 cells were transfected with control siRNA or siRNA (GREB1 #1 siRNA and GREB1 #2 siRNA) and cultured in medium containing 0.1% FBS for 2 days. The cell lysates were probed with anti-cleaved caspase 3, anti-PARP1, and anti-HSP90 antibodies. The intracellular levels of anti-cleaved caspase 3 and PARP1, markers of apoptosis, were confirmed to be increased in HepG2 cells in which GREB1 was knocked out (see Figure 4g).

[0120] These results suggest that GREB1 is essential not only for cell proliferation through cell cycle regulation but also for cell survival in hepatoblastoma cells.

[0121] 2-3.GREB1 forms a complex with Smad2 / 3 Consistent with previous reports in breast cancer cells, HA-FLAG-GREB1 was primarily localized in the nucleus in X293T cells (see Figure 5a). A mutant GREB1 (HA-FLAG-ΔNLS-GREB1) lacking the GREB1 nuclear localization sequence (NLS) at amino acids 310–319 was localized in the cytoplasm (see Figure 5a). To elucidate the mechanism by which GREB1 regulates hepatoblastoma cell proliferation independently of ER signaling, we identified candidate proteins that may interact with GREB1 using the Biological General Repository for Interaction Datasets (BioGRID) (https: / / thebiogrid.org / ). As a result, the proteins listed in Table 6 were identified as candidate proteins that may interact with GREB1.

[0122] [Table 6]

[0123] Among these candidate proteins, we focused on Smad4, a central mediator of TGFβ signaling. Smad proteins are classified into three classes based on their functions: Smad3 (co-mediator Smad, or Co-Smad), Smad4 (receptor-regulated Smad, or R-Smad), and Smad7 (inhibitory Smad, or I-Smad). In X293T cells, GFP-Smad3 and GFP-Smad7 were primarily localized in the nucleus, whereas GFP-Smad4 was localized in the cytoplasm (see Figure 6a). Therefore, we assessed the ability of Smad3, Smad4, and Smad7 to form complexes with GREB1. Specifically, we first immunoprecipitated X293T cell lysates expressing HA-FLAG-GREB1 and GFP, GFP-Smad3, GFP-Smad4, or GFP-Smad7 with anti-GFP antibodies. Then, the cell lysates (Input) and the immunoprecipitates (IP) were reacted with anti-HA or anti-GFP antibodies. We confirmed that HA-FLAG-GREB1 formed complexes with Smad3 (Co-Smad) and Smad7 (I-Smad), but not with Smad4 (R-Smad) (Fig. 5b). On the other hand, we confirmed that HA-FLAG-ΔNLS-GREB1 formed complexes with Smad4 (R-Smad), but not with Smad3 (Co-Smad) or Smad7 (I-Smad) (Fig. 5b). These results suggest that GREB1 binds to all Smad family members in a manner dependent on its subcellular localization. Furthermore, we immunoprecipitated HepG2 cell lysates with anti-Smad2 / 3 antibodies, and then reacted the resulting immunoprecipitates (IPs) with anti-Smad2 / 3 and anti-GREB1 antibodies. Although GREB1 bound to endogenous Smad2 / 3 (Fig. 5c), there was little interaction between GREB1 and Smad4 or Smad7 (data not shown). In Huh6 cells expressing GFP-GREB1, the cell lysate was immunoprecipitated with anti-GFP antibody, and the resulting immunoprecipitate (IP) was reacted with anti-Smad2 / 3 antibody and anti-GREB1 antibody. GREB1 was found to bind to endogenous Smad2 / 3 (see Figure 6b).On the other hand, other nuclear proteins, such as β-catenin and c-Myc, did not interact with GFP-Smad3, GFP-Smad4, or GFP-Smad7 (see Figure 6c). Therefore, we further investigated the functional interaction between GREB1 and Smad2 / 3.

[0124] Smad2 is known to have two functional domains, the MH1 domain and the MH2 domain (see Figure 5d). Therefore, we evaluated the binding of Smad2 to GREB1 using full-length Smad2 (Full), a Smad2 mutant (N) lacking the C-terminal region (amino acids 266-467) and consisting only of the C-terminal region including the MH1 domain, and a Smad2 mutant (C) lacking the N-terminal region (amino acids 1-265) and consisting only of the C-terminal region including the MH2 domain (see Figure 5d). Lysates of X293T cells expressing HA-FLAG-mGREB1 and GFP, or GFP-Smad2 (Full, mutant N, or mutant C) were immunoprecipitated with anti-GFP antibodies. Cell lysates (Input) and immunoprecipitates (IP) were then reacted with anti-HA or anti-GFP antibodies. GFP-Smad2 (mutant C) formed a complex with GREB1, whereas GFP-Smad2 (mutant N) did not (see Figure 5d). Furthermore, pull-down assays showed that recombinant GST-Smad2 / MH2 domain bound to endogenous GREB1 and GFP-GREB1 expressed in Huh6 cells under conditions where endogenous Smad4 was bound (see Figure 5e). In contrast, recombinant GST-Smad2 / MH1 domain did not bind under the same conditions (see Figure 5e). A similar tendency was observed in Huh6 cells expressing GFP-GREB1 (see Figure 6d). These results suggest that GREB1 directly binds to Smad2 / 3 in hepatoblastoma cells.

[0125] Furthermore, to analyze which region of GREB1 interacts with Smad2 / 3, we created mutant N(1-666) in which amino acids 667 to 1954 of GREB1 were deleted, mutant M(667-1333) in which amino acids 1 to 666 and 1334 to 1954 of GREB1 were deleted, and mutant C(1334-1954) in which amino acids 1 to 1333 of GREB1 were deleted.Furthermore, we created NLS-GREB1 mutants (NLS / 667-1333 and NLS / 1334-1954) in which three copies of the NLS sequence derived from SV40 T antigen (sequence number 109) were attached to the N-terminus of mutants M and C, which do not have an NLS (see Figure 5f). X293T cells expressing GFP-GREB1 mutants were then stained with anti-GFP antibody and Hoechst 33342 (see Figure 5f). As a result, it was confirmed that the GREB1 mutants (1-666, NLS / 667-1333, and NLS / 1334-1954) were localized in the nucleus.

[0126] Lysates of X293T cells expressing FLAG-Smad3 and GFP, GFP-GREB1, or GFP-GREB1 mutants were immunoprecipitated with anti-GFP antibodies. The resulting cell lysates (Input) and immunoprecipitates (IP) were then reacted with anti-FLAG or anti-GFP antibodies. The results showed that in X293T cells, the GREB1 mutant (NLS / 667-1333) bound to Smad3, whereas the GREB1 mutants (NLS / 1-666 and NLS / 1334-1954) hardly bound to Smad3 (see Figure 5g).

[0127] Furthermore, we prepared a mutant (NLS / 667-1954(ΔN)) in which the amino acids at positions 1 to 666 of GREB1 were deleted and three copies of the NLS sequence (SEQ ID NO: 109) derived from SV40 T antigen were attached to the N-terminus, a mutant (Δ667-1333 / ΔM) in which the amino acids at positions 667 to 1333 of GREB1 were deleted and three copies of the NLS sequence (SEQ ID NO: 109) derived from SV40 T antigen were attached to the N-terminus of GREB1, and a mutant (NLS / 1-1333(ΔC)) in which the amino acids at positions 667 to 1333 of GREB1 were deleted and three copies of the NLS sequence (SEQ ID NO: 109) derived from SV40 T antigen were attached to the N-terminus of GREB1. Lysates of X293T cells expressing these mutants were immunoprecipitated with an anti-GFP antibody. Next, the obtained cell lysates (Input) and immunoprecipitated The immunoprecipitates (IP) were reacted with anti-FLAG or anti-GFP antibodies (see Figure 6e). Wild-type GREB1 was confirmed to bind to Smad3, whereas a GREB1 mutant lacking amino acids 667-1333 of GREB1 did not bind to Smad3 (see Figure 6e). These results demonstrated that the amino acid region 667-1333 of GREB1 plays an important role in binding to Smad3. Furthermore, these results suggest that the amino acid region containing amino acids 667-1333 of GREB1 interacts with the NH2 domain of Smad2 / 3 to form a complex in the nuclei of hepatoblastoma cells.

[0128] 2-4.GREB1 functions as a negative regulator of TGFβ signaling To elucidate the role of GREB1 in TGFβ signaling, we investigated the expression of target genes, the nuclear translocation of Smad2, the complex formation between Smad2 / 3 and Smad4, and the phosphorylation of Smad2 / 3 at the endogenous level.

[0129] Analysis of public datasets from hepatoblastoma patients revealed that the expression levels of β-catenin signaling target genes Axin2 and DKK1 were significantly elevated in tumor lesions compared with non-tumor regions (see Figure 8a). Conversely, the expression levels of TGFβ signaling target genes PAI-1 and GADD45B were significantly decreased in tumor lesions compared with non-tumor regions (see Figure 7a). Furthermore, a significant inverse correlation was observed between the expression levels of GREB1 and the expression levels of TGFβ signaling target genes (PAI-1, p21 / CDKN1A, TSP-1, and CTGF) (see Figure 7b). HepG2 cells expressing GFP (HepG2 / GFP) or GFP-GREB1 (HepG2 / GFP-GREB1) were transfected with control siRNA or GREB1 #2 siRNA. The mRNA levels of PAI-1 and SNAIL2 were measured in these cells by real-time PCR (n = 3). As a result, in HepG2 cells in which GREB1 was knocked down (HepG2 / GFP), the expression levels of PAI-1 and SNAIL2, target genes of TGFβ signaling, were increased, and these phenotypes were restored by the expression of GFP-GREB1 (see Figure 7c).

[0130] We also analyzed the PAI-1 mRNA levels in HepG2 cells (control), HepG2 cells with GREB1 knockout (GREB1 KO), and HepG2 cells with GREB1 expression (GREB1 KO / GREB1) by real-time PCR. The results showed that GREB1 knockout significantly increased the PAI-1 mRNA levels, and exogenous GREB1 expression rescued the GREB1 knockout phenotype (see Figure 8b).

[0131] HepG2 cells were transfected with control siRNA or GREB1 #2 siRNA and cultured in the presence or absence of a TGFβ receptor inhibitor (ALK5 inhibitor), and the levels of PAI-1 and SNAIL2 mRNA in the cells were measured. The results showed that the increase in PAI-1 and SNAIL2 mRNA levels due to GREB1 knockdown was dependent on TGFβ receptor signaling (see Figure 7d).

[0132] HepG2 cells transfected with control siRNA or GREB1 #2 siRNA were cultured for 30 minutes in the presence or absence of 10 ng / mL TGFβ. After the culture, the cells were fixed and immunostained with anti-Smad2 / 3 antibody (see Figure 8c). The cell lysate after the culture was immunoprecipitated with anti-Smad2 / 3 antibody. The resulting cell lysate (Input) and immunoprecipitate (IP) were then reacted with anti-Smad4 antibody or anti-Smad2 / 3 antibody (see Figure 8d). The cell lysate after the culture was then reacted with anti-GREB1 antibody, anti-phosphorylated Smad2 / 3 (pSmad2 / 3) antibody, or anti-Smad2 / 3 antibody (see Figure 8e). As a result, it was found that knockdown of GREB1 did not affect TGFβ-dependent nuclear translocation of Smad2 / 3, complex formation between Smad2 / 3 and Smad4, or phosphorylation of Smad2 / 3 (see Figure 8c-e). These results demonstrate that GREB1 inhibits the function of Smad2 / 3 in the nucleus, thereby inhibiting TGFβ-dependent gene expression.

[0133] Transcriptional activators such as CBP and p300 are known to possess histone acetyltransferase (HAT) activity, which modifies chromatin structure. Furthermore, R-Smads (Smad2 / 3) are known to directly interact with CBP or p300 via the MH2 domain. Therefore, we investigated the effect of GREB1 knockdown on the binding of transcriptional activators to Smad2 / 3.

[0134] Lysates of HepG2 cells transfected with control siRNA or GREB1 #2 siRNA were immunoprecipitated with anti-Smad2 / 3 antibodies. The resulting cell lysates (Input) and immunoprecipitates (IP) were then reacted with anti-p300, anti-GREB1, or anti-Smad2 / 3 antibodies (see Figure 7e). Knockdown of GREB1 in HepG2 cells increased binding of Smad2 / 3 to p300 (see Figure 7e).

[0135] Cell lysates of X293T cells expressing HA-FLAG-GREB1, GFP-Smad2 mutant (C), or GFP were immunoprecipitated with anti-GFP antibody. The resulting cell lysates (Input) and immunoprecipitates (IP) were then reacted with anti-p300 antibody, anti-HA antibody, or anti-GFP antibody. The results showed that the Smad2 / C mutant (C) containing the MH2 domain interacted with p300 in X293T cells, and overexpression of HA-FLAG-GREB1 inhibited the interaction of the Smad2 / C mutant (C) with p300 (see Figure 7f).

[0136] HepG2 cells transfected with control siRNA or GREB1 #2 siRNA were cultured for 30 minutes in the presence or absence of 10 ng / mL TGFβ. Cell lysates were immunoprecipitated with anti-acetylated histone 4 (AcH4) antibody. The exon 2 region of PAI-1 contained in the cell lysate (Input) and immunoprecipitate (IP) was then analyzed by PCR using region-specific primers. This confirmed that GREB1 knockdown increased the amount of acetylated histone 4 at the PAI-1 locus (exon 2) in HepG2 cells (see Figure 7g).

[0137] We overexpressed a constitutively active TGFBR1 mutant (T204D) in HepG2 cells together with GFP, GFP-GREB1, or the GFP-GREB1 mutant (Δ667-1333 / ΔM), and quantified the mRNA expression of SNAIL2 or p15. Expression of GREB1 inhibited the constitutively active TGFBR1-dependent SNAIL2 or p15 gene expression, whereas expression of the GREB1 mutant (Δ667-1333 / ΔM) did not significantly inhibit SNAIL2 or p15 gene expression (see Figure 7h). Furthermore, under the same conditions, overexpression of GREB1 did not affect Axin2 expression (see Figure 7h).

[0138] In addition, we analyzed the levels of AFP mRNA and DLK1 mRNA in HepG2 cells transfected with the TGFBR1 / T204D mutant (a mutant in which the threonine at position 204 of TGFBR1 is replaced by aspartic acid) by real-time PCR. Overexpression of the constitutively active TGFBR1 mutant (T204D) reduced AFP and DLK1 expression in HepG2 cells (see Figure 8f). This result is consistent with previous reports that TGFβ signaling reduces AFP and DLK1 expression in rat fetal hepatocytes or hepatoma cells.

[0139] Lysates from HepG2 cells and HepG2 cells with Smad2 / 3 knockout (HepG2 / Smad2 / 3 KO) were probed with anti-Smad2 / 3 or anti-HSP90 antibodies. Furthermore, HepG2 cells and HepG2 cells with Smad2 / 3 knockout (HepG2 / Smad2 / 3 KO) were transfected with control siRNA or siRNA against GREB1 (GREB1 #2 siRNA), and AFP mRNA levels were analyzed by real-time PCR. The results confirmed that Smad2 / 3 knockout rescued the downregulation of AFP expression caused by GREB1 knockdown, but did not affect the downregulation of DLK1 (see Figure 8, g and h).

[0140] These results suggest that GREB1 prevents Smad2 / 3 from binding to p300, thereby inhibiting the expression of target genes of TGFβ-Smad signaling.

[0141] Furthermore, we analyzed the gene expression levels of TGFβ1, TGFB2, and TGFB3 using RNA sequence data from HepG2 in the mRNA profile dataset of the Cancer Cell Line Encyclopedia (CCLE) (https: / / portals.broadinstitute.org / ccle). We found that HepG2 cells highly expressed TGFβ1 (see Figure 9a).

[0142] In addition, the levels of PAI-1 mRNA, TGFβ1 mRNA, and GREB1 mRNA in HepG2 cells transfected with control siRNA, TGFβ1 siRNA, GREB1 siRNA (GREB1 #2 siRNA), or a combination of TGFβ1 siRNA and GREB1 siRNA (GREB1 #2 siRNA) were analyzed by real-time PCR. As a result, it was found that double knockdown of TGFβ1 and GREB1 significantly suppressed the increase in PAI-1 gene expression caused by GREB1 knockdown (see Figure 9b).

[0143] Lysates of HepG2 cells transfected with control siRNA, TGFβ1 siRNA, GREB1 siRNA (GREB1 #2 siRNA), or a combination of siRNAs for TGFβ1 and GREB1 (GREB1 #2 siRNA) were probed with anti-GREB1, anti-TGFB1, or anti-HSP90 antibodies. The results demonstrated that double knockdown of TGFβ1 and GREB1 efficiently suppressed the protein expression of TGFβ1 and GREB1 (see Figure 9c).

[0144] Furthermore, HepG2 cells transfected with control siRNA, TGFβ1 siRNA, GREB1 siRNA (GREB1 #2 siRNA), or a combination of TGFβ1 siRNA and GREB1 siRNA (GREB1 #2 siRNA) were cultured in two dimensions (in plastic dishes) and counted over time. The results showed that the suppression of cell proliferation due to GREB1 knockdown was reversed by double knockdown of TGFβ1 (see Figure 7i).

[0145] HepG2 cells transfected with control siRNA, TGFβ1 siRNA, GREB1 siRNA (GREB1 #2 siRNA), or a combination of TGFβ1 siRNA and GREB1 siRNA (GREB1 #2 siRNA) were cultured for 4 hours in the absence of TGFB1 at concentrations of 0.01 to 1 ng / ml, and then PAI-1 mRNA levels were analyzed by real-time PCR. Knockdown of GREβ1 and TGFβ1 resulted in TGFβ-dependent enhancement of PAI-1 mRNA expression (see Figure 9d).

[0146] Furthermore, knockdown of GREB1 in HepG2 cells dramatically increased the expression of p15, one of the target genes p15, p21, and p27, which regulate cell proliferation downstream of TGFβ signaling. The increase in p15 expression due to GREB1 knockdown was suppressed by knockout of Smad2 / 3 (see Figure 9e). Furthermore, knockdown of p15 in HepG2 cells rescued the phenotype of cell proliferation suppression and enhanced cell death induced by GREB1 suppression (see Figure 9f). These results suggest that p15 is important for the phenotype of cell proliferation suppression and enhanced cell death induced by GREB1 suppression. Consistent with this result, knockout of Smad2 / 3 in HepG2 cells inhibited the cell proliferation suppression induced by GREB1 knockdown (see Figure 7j).

[0147] Furthermore, HepG2 cells transfected with control siRNA, GREB1 siRNA (GREB1 #2 siRNA), p15 siRNA, or a combination of GREB1 (GREB1 #2 siRNA) and p15 siRNA were cultured in two dimensions (cultured in plastic dishes) and counted over time. The suppression of cell proliferation due to GREB1 knockdown was reversed by p15 knockdown, demonstrating a p15-dependent phenotype (see Figure 9g).

[0148] HepG2 cells transfected with control siRNA, GREB1 siRNA (GREB1 #2 siRNA), p15 siRNA, or a combination of GREB1 (GREB1 #2 siRNA) and p15 siRNA were cultured for 2 days in medium containing 0.1% FBS (with or without the caspase inhibitor Z-VAD). Cell viability was determined by staining with propidium iodide (PI) and Hoechst 33342. The enhanced cell death induced by GREB1 knockdown was reversed by p15 knockdown, demonstrating a p15-dependent phenotype (see Figure 9h).

[0149] These results suggest that GREB1 knockdown enhances sensitivity to TGFβ, suppressing cell proliferation and inducing cell death.

[0150] Furthermore, in breast cancer cells MCF7, treatment with an estrogen receptor antagonist (ICI.182.780) or knockdown of GREB1 suppressed PAI-1 mRNA expression (see Figure 9, i and j). Furthermore, in MCF7 cells, GREB1 bound to endogenous Smad2 / 3 (see Figure 9, k). These results suggest that GREB1 functions as a negative regulator of TGFβ signaling in breast cancer cells as well as hepatoblastoma cells.

[0151] 2-5. Interaction of GREB1 and Smad2 / 3 in a restricted region of the nucleus inhibits transcription HepG2 cells expressing GFP-GREB1 were fixed and stained with anti-GFP antibody and Hoechst33342. GFP-GREB1 was observed as dots in interchromatin compartments, not in chromatin regions (see Figure 10a). Furthermore, HepG2 cells expressing GFP-GREB1 were fixed and stained with anti-GFP, anti-fibrillarin, anti-SC34, anti-PML, or anti-coilin antibodies and Hoechst33342. The dots stained with GREB1 were not colocalized with fibrillarin, SC35, PML, or coilin (see Figure 10b). These results suggest that GREB1 exists independently of nucleoli, nuclear speckles, promyelocytic leukemia (PML) bodies, and Cajal bodies. Furthermore, when HepG2 cells expressing GFP-GREB1 and FLAG-SMAD3 were fixed and stained with anti-GFP antibody, anti-FLAG antibody, and Hoechst33342, it was observed that GFP-GREB1 and FLAG-Smad3 formed a complex and were present in the interchromatin compartment (see Figure 10c).

[0152] HepG2 cells were cultured for 30 minutes in the presence or absence of 10 ng / mL TGFβ. The cells were then fixed and stained with anti-SMAD2 / 3, anti-GREB, and Hoechst 33342 antibodies. The GREB1 fluorescence intensity in the nucleus and cytoplasm was measured, and the results were expressed as the ratio of the nuclear to the cytoplasmic GREB1 fluorescence intensity. The results confirmed that endogenous GREB1 and Smad3 are present in both the cytoplasm and nucleus, and that they accumulate in the nucleus upon TGFβ stimulation (see Figure 11a).

[0153] HepG2 cells were cultured for 30 minutes in the presence or absence of 10 ng / mL TGFβ, and then fixed and stained with anti-GREB1 antibody, anti-phosphorylated SMAD2 / 3 (pSMAD2 / 3) antibody, and Hoechst 33342. The results showed that in TGFβ-treated cells, phosphorylated SMAD2 / 3 accumulated in the nucleus and colocalized with GREB1 (see Figure 11b).

[0154] HepG2 cells were cultured for 30 min in the presence or absence of 10 ng / mL TGFβ. After incubation, the cells were fixed and incubated with mouse anti-GREB1 and rabbit anti-SMAD2 / 3 antibodies. Secondary antibodies (proximity ligation assay (PLA) probes) were then conjugated to these primary antibodies. Smad2 / 3 were found to localize closely to GREB1 at the chromatin-interchromatin interface in a TGFβ-stimulated manner (see Figure 11c). While GREB1 mutants (1-666 and NLS / 667-1333) were present throughout the nucleus but did not form nuclear foci, the GREB1 mutant (NLS / 1334-1954) formed nuclear foci similar to full-length GREB1, suggesting that the C-terminal region of GREB1 plays an important role in localizing GREB1 to specific nuclear regions (see Figure 11d).

[0155] To clarify the functional interaction between the specific localization of GREB1 and Smad2 / 3 in the nucleus, we visualized transcriptional activity by analyzing RNA synthesis using ethynyluridine (EU). Specifically, HepG2 cells expressing GFP-SMAD3 and expressing or not expressing HA-FLAG-GREB1 were first incubated in the presence of EU at a final concentration of 1 mM for 30 min. After incubation, the cells were fixed and stained with anti-GFP antibody, anti-FLAG antibody, and Hoechst 33342. Furthermore, we confirmed that EU-labeled intranuclear regions could be detected when HepG2 cells were incubated with EU for 30 min and then fixed (see Figure 10, d). The results showed that in GFP-SMAD3-expressing HepG2 cells, nascent RNA molecules were observed throughout the nucleoplasm, and some of them colocalized with GFP-Smad3 nuclear foci (see Figure 11, e). In contrast, in HepG2 cells expressing GFP-SMAD3 and HA-FLAG-GREB1, EU labeling did not colocalize with GFP-Smad3 nuclear foci (see Fig. 11e). These results suggest that the transcriptional activity associated with TGFβ-SMAD signaling is selectively suppressed by the interaction between Smad2 / 3 and GREB1 at the boundary between chromatin and interchromatin.

[0156] 2-6.GREB1 is involved in the formation of hepatoblastoma-like tumors in vivo Several mouse liver tumor models of hepatocellular carcinoma and hepatoblastoma have been developed using oncogene genomic manipulation and / or hydrodynamic transfection. It has also been reported that overexpression of constitutively active β-catenin and YAP using hydrodynamic transfection rapidly leads to liver tumors with characteristics of hepatocellular carcinoma and hepatoblastoma. Indeed, immunostaining of hepatoblastoma tissues (n = 11) with anti-β-catenin antibody and hematoxylin revealed that YAP was overexpressed in the cytoplasm or nucleus in tumor cells specifically in 9 of the 11 hepatoblastoma tissues (81.8% positive rate), and all 9 cases were positive for β-catenin and GREB1 (see Figure 12a). Furthermore, it has been reported that the HGF-c-Met pathway is activated in hepatoblastoma, and coexpression of constitutively active forms of β-catenin and c-Met induces liver tumors in mice.

[0157] We analyzed which combination of β-catenin, YAP, and c-Met was effective in hepatoblastoma formation by hydrodynamic transfection. In mice transfected with ΔN90β-catenin, a deletion of amino acids 1-90 of human β-catenin, and YAPS127A, a YAP mutant in which serine 127 is replaced with alanine to prevent serine phosphorylation (BY model), small tumors formed in the liver 6 weeks after transfection, but immunohistochemically, GREB1 and DLK1 were barely expressed (see Figure 12b). In mice transfected with ΔN90β-catenin and c-Met (BM model), immunohistochemically, low levels of GREB1 and DLK1 were expressed in tumors formed in the liver 6 weeks after transfection (see Figure 12b). On the other hand, in mice transfected with a combination of ΔN90β-catenin, YAPS127A, and c-Met (BYM model), large, multiple tumors developed in the liver 6 weeks after transfection, and immunohistochemical analysis revealed high expression of GREB1 and DLK1 (see Figure 12b). Furthermore, when the mRNA levels of GREB1 and TACSTD1 were measured in tumor nodules of the BY, BM, and BYM models by real-time PCR, the mRNA levels were higher in the BYM mice than in the BY and BM models (see Figure 12c). These results suggest that the BYM model is an appropriate model for in vivo functional analysis of GREB1 in hepatoblastoma.

[0158] Next, we analyzed the mechanism by which YAP and c-Met regulate GREB1 expression in hepatoblastoma. Huh6 and HepG2 cells were fixed and stained with anti-YAP antibody and Hoechst33342. We found that YAP was more strongly concentrated in the nucleus in Huh6 cells than in HepG2 cells (Figure 13a). However, knockdown of YAP / TAZ in Huh6 and HepG2 cells did not affect GREB1 expression (Figure 13b). Furthermore, treatment of Huh6 cells with CHIR99021, XMU-MP1, an Mst1 / 2 kinase inhibitor and YAP activator, or a combination of these treatments suppressed GREB1 mRNA expression, as well as Axin2 mRNA expression (Figure 13c). These results suggest that YAP / TAZ is required for hepatoblastoma formation but not for GREB1 expression.

[0159] Furthermore, knockdown of c-Met in HepG2 cells suppressed the expression of Axin2 mRNA and GREB1 mRNA, while increasing the expression of ANKRD1 and Cyr61, downstream target genes of YAP (see Figure 13d). It has been reported that the HGF-c-Met signaling pathway activates β-catenin signaling by promoting the phosphorylation of tyrosine 654 of β-catenin and its nuclear translocation. Indeed, knockdown of c-Met in HepG2 cells attenuated the phosphorylation of tyrosine 654 of β-catenin and reduced the expression of GREB1 and Axin2 (see Figure 13e). These results suggest that c-Met activates β-catenin signaling by inducing tyrosine phosphorylation of β-catenin, leading to increased GREB1 expression.

[0160] We also administered GREB1 shRNA to mice, along with ΔN90β-catenin, YapS127A, and c-Met, and then harvested the livers 7–8 weeks after administration. In mice transfected with ΔN90β-catenin, YAPS127A, and c-Met (BYM mice, control), multiple nodules were observed across the entire liver surface (see Figure 14A). In contrast, in mice transfected with GREB1 shRNA together with ΔN90β-catenin, YAPS127A, and c-Met (BYM GREB1 KD mice), nodules in the liver were suppressed (see Figure 14A). Furthermore, tumor nodules were obtained from the livers of BYM mice (C1 to C7) and untreated mice (NL mice), and the expression levels of ΔN90β-catenin, YAPS127A, and c-Met were measured using real-time PCR.It was confirmed that the BYM mice expressed all of the introduced ΔN90β-catenin, YAPS127A, and c-Met (see Figure 15a).

[0161] Furthermore, total RNA was extracted from the livers of untreated mice (NL), tumor nodules of BYM mice (n = 42), and non-tumor tissues of BYM mice (n = 8), and the amount of GREB1 mRNA was measured by real-time PCR. The results showed that the amount of GREB1 mRNA in each tumor nodule was higher than that in non-tumor tissues and normal liver tissues (see Figure 15b).

[0162] Six tumor nodules were obtained from each BYM mouse (C1–C7) and examined for hepatoblastoma-related gene expression and histological appearance. Heatmap visualization revealed that the amount of GREB1 mRNA varied among individual tumor nodules (see Figure 14b). The tumor nodules were then classified into high (GREB1-high: C1, C4, and C6) and low (GREB1-low: C2, C3, C5, and C7) groups. The GREB1-high group tended to have higher levels of TACSD1, DLK1 (hepatoblastoma marker genes), AFP, GFP3 (undifferentiated hepatoblast marker gene), PEG3, MEG3, BEX1 (imprinted gene), and Axin2 than the GREB1-low group (see Figure 14b). In addition, using tumor nodules obtained from the BYM mice (C1 to C7), the expression levels of REB1 and the mRNA expression levels of hepatoblastoma-related genes (DLK1, TACSTD1, GPC3, MEG3, and Axin2) were measured, and a strong positive correlation was observed between the expression levels of GREB1 and the expression levels of hepatoblastoma-related genes (see Figure 15c, Table 7).

[0163] [Table 7]

[0164] In addition, hematoxylin-eosin staining was performed on liver tissue sections isolated from BYM mice in the GREB1-high (C4) and GREB1-low (C3) groups. Results indicated that the majority of tumors in the GREB1-high group had a high nuclear-to-cytoplasmic ratio and high proliferative capacity (see Figure 14c and Figure 15d). Furthermore, tumors in the GREB1-low group primarily contained large, differentiated cells with clear cytoplasm, uniformly round nuclei, and small nucleoli (see Figure 14c and Figure 15d).

[0165] In addition, liver tissue sections isolated from BYM mice with high (C4) and low (C3) GREB1 expression were stained with anti-GREB1 or anti-DLK1 antibodies and hematoxylin. Immunohistochemical analysis confirmed that GREB1 and DLK1 expression in tumor lesions was higher than in non-tumor regions, and the staining level correlated with the degree of differentiation (see Figure 14d).

[0166] When GREB was knocked down in BYM mice (BYM GREB1 KD mice; K1–K6), the incidence of tumor formation was reduced, with no tumors observed in 4 of 6 mice (see Figure 14a). Furthermore, liver weight and serum AFP levels were dramatically reduced in BYM + GREB1 shRNA mice compared with BYM mice without GREB knockdown (see Figure 15e).

[0167] In addition, total RNA was extracted from tumor nodules of three BYM mice (C1, C4, and C6), four BYM mice (C2, C3, C5, and C7), and two BYM mice (BYM GREB1 KD mice; K2 and K4) that had been administered GREB1 shRNA. The mRNA levels of GREB1, DLK1, and TASCSTD1 were analyzed by real-time PCR. The tumors of the BYM GREB1 KD mice (K2 and K4) showed low GREB1 mRNA expression, while the expression levels of DLK1 and TACSTD1 were comparable to those of the GREB1-low BYM mice (see Figure 14f). Furthermore, liver tissue sections from the BYM mice (C4) and BYM GREB1 KD mice (K2) were stained with hematoxylin and eosin or anti-GREB1 antibody and hematoxylin. As a result, it was confirmed that these tumor cells were well-differentiated hepatoblastoma-like cells with clear cytoplasm (see FIG. 14 g).

[0168] N-cadherin is a target gene of TGFβ signaling and is known to be expressed at higher levels in hepatoblastoma cells than in mesenchymal cells within tumor tissue. Therefore, liver tissue sections from BYM mice (C4) and BYM GREB1 KD mice (K2) were stained with anti-N-cadherin antibody and Hoechst 33342. In BYM mice, N-cadherin expression was lower in tumor lesions than in non-tumor regions. In contrast, in BYM GREB1 KD mice, N-cadherin was upregulated in tumor lesions, with expression levels comparable to those in non-tumor regions (see Figure 14h). These results suggest that TGFβ signaling is activated by downregulation of GREB1 expression.

[0169] These results demonstrate that GREB1 is involved in hepatoblastoma-like histological patterns, marker gene expression, and tumorigenesis in this mouse model.

[0170] 2-7.GREB1 may be a target molecule for hepatoblastoma To investigate whether GREB1 could be a target molecule for hepatoblastoma therapy, we used the CRISPR / Cas9 system to generate HepG2 cells in which GREB1 was knocked out (GREB1 KO cells), HepG2 cells in which GREB1 was knocked out and GGREB1 was expressed (GREB1 KO / GREB1 cells), HepG2 cells in which GREB1 was knocked out and GREB1ΔNLS was introduced (GREB1 KO / GREB1ΔNLS cells), and cells in which GREB1 and Smad2 / 3 were knocked out (GREB1 KO / Smad2 / 3 KO cells).

[0171] Wild-type HepG2 cells (control), GREB1 KO cells, GREB1 KO / GREB1 cells, GREB1 KO / GREB1ΔNLS cells, or GREB1 KO / Smad2 / 3 KO cells (7 × 10 6GREB1 cells) were subcutaneously implanted into 5-week-old male BALB / cAnNCrj-nu nude mice. 28 days after implantation, the mice were sacrificed and xenograft tumors were excised. The appearance and weight of the xenograft tumors were measured. Wild-type HepG2 cells formed subcutaneous xenograft tumors, whereas GREB1 knockout HepG2 cells exhibited reduced tumor size and weight (see Figure 16a). Furthermore, expression of GREB1 rescued the phenotype induced by GREB1 knockout, but expression of GREB1ΔNLS did not (see Figure 16a). Furthermore, knockout of Smad2 / 3 rescued the phenotype induced by GREB1 knockout, confirming that GREB1 inhibits TGFβ signaling-dependent cell proliferation suppression in vivo.

[0172] Lysates from HepG2 cells (control), HepG2 cells with GREB1 knockout (GREB1 KO), or HepG2 cells with GREB1 and Smad2 / 3 knockout combined (GREB1 KO + Smad2 / 3 KO) were probed with anti-GREB1, anti-Smad2 / 3, and anti-HSP90 antibodies. The results revealed that GREB1 or Smad2 / 3 protein expression was lost (see Figure 15e). Furthermore, HepG2 cells (WT), HepG2 cells with GREB1 knockout (GREB1 KO), or HepG2 cells with GREB1 and Smad2 / 3 knockout combined (GREB1 KO / Smad2 / 3 KO) were cultured in two dimensions (in plastic dishes) and cell numbers were measured over time. As a result, the suppression of cell proliferation caused by GREB1 knockout was rescued by knockout of Smad2 / 3, demonstrating that this is a Smad2 / 3-dependent phenotype (see Figure 15f). These results also confirmed that GREB1 can inhibit TGFβ signaling-dependent suppression of cell proliferation in vivo.

[0173] 2-8. Antisense oligonucleotides against GREB1 reduce proliferation and tumorigenesis of hepatoblastoma cells To verify the effect of antisense oligonucleotides against human GREB1 (GREB1 ASOs) on tumor growth, we designed the target sequence (15 nucleotides) of human GREB1 based on the secondary structure of GREB1 mRNA. Specifically, we first eliminated potentially cytotoxic GREB1 ASOs from thousands of candidate GREB1 ASOs and selected 20 GREB1 ASO sequences using high-dimensional structure prediction. Based on these sequences, we designed and synthesized various modified GREB1 ASOs (Table 8). [Table 8]

[0174] Next, HepG2 cells were transfected with either the control ASO or each of the GREB1 ASOs and cultured in 10% FBS-containing medium for 2 days. The cell lysates were probed with anti-GREB1 and anti-HSP90 antibodies. Among the 20 GREB1 ASOs, ASO-6434, ASO-6921, ASO-6968, and ASO-7724 were found to strongly suppress GREB1 expression in HepG2 cells without cytotoxicity (see Figure 17a).

[0175] In addition, HepG2 cells expressing GFP or GFP-GREB1 were transfected with a control ASO or each of the GREB1 ASOs and cultured in three-dimensional Matrigel for 4 days. After culture, the cells were stained with phalloidin and Hoechst 33342, and the sphere area was calculated (n=50). Results showed that ASO-6434, ASO-6921, ASO-6968, and ASO-7724 were able to inhibit the sphere-forming activity of HepG2 cells. However, the inhibitory effects of ASO-6921, ASO-6968, and ASO-7724 on sphere-forming activity were lost in HepG2 cells expressing ASO-resistant GREB1 (HepG2 cells expressing GFP-GREB1) (see Figure 16b). These results demonstrated that at least ASO-6921, ASO-6968, and ASO-7724 inhibited sphere formation in HepG2 cells through an on-target effect (not an off-target effect).

[0176] On day 0, HepG2 cells (1.0 × 10 7Matrigel containing GREB1 ASO-6921 (n = 5), or GREB1 ASO-7724 (n = 6) was implanted into the livers of nude mice. Starting on day 3, 50 μg of control ASO (n = 9), GREB1 ASO-6921 (n = 5), or GREB1 ASO-7724 (n = 6) was administered subcutaneously twice weekly. On day 29 after implantation, mice were euthanized and tumors were observed. Both GREB1 ASO-6921 and GREB1 ASO-7724 suppressed tumor formation by HepG2 cells and reduced tumor weight compared with the control ASO (see Figure 16c). The mRNA levels of GREB1 and PAI-1 in each liver tumor were analyzed by real-time PCR. Furthermore, sections of each liver tumor were stained with anti-Ki-67 antibody and hematoxylin to determine the percentage of Ki-67-positive cells relative to the total number of hematoxylin-positive cells. We found that GREB1 ASOs-6921 and -7724 inhibited GREB1 expression in liver tumors and reduced the number of Ki-67-positive cells (see Figure 16(d)). Furthermore, sections of each liver tumor were stained with anti-cleaved caspase 3 antibody and hematoxylin to determine the ratio of cleaved caspase 3-positive cells to the total number of hematoxylin-positive cells. We found that GREB1 ASOs-6921 and -7724 induced tumor cell apoptosis (see Figure 16(e)). Furthermore, we analyzed the mRNA levels of GREB1 and PAI-1 in each liver tumor by real-time PCR. We found that PAI-1 gene expression in hepatoblastoma tumors tended to increase with GREB1 ASOs-6921 and -7724 (see Figure 16(f)). This suggests that activation of TGFβ signaling suppresses tumor cell proliferation and survival. In addition, sections of non-tumorous livers were stained with anti-cleaved caspase 3 antibody and hematoxylin. Results confirmed that GREB1 ASO-6921 and -7724 did not induce histological damage or cell death in non-tumorous liver regions (see Figure 17b). These results demonstrate that GREB1-specific ASOs may be a novel therapeutic agent for hepatoblastoma.

[0177] Furthermore, we constructed mGREB1 ASO-5715 and a control ASO (see Table 9) targeting mouse GREB1 and analyzed their effects on liver tumor formation in the BYM model. ΔN90β-catenin, YAPS127A, and c-Met (BYM) were transfected into nude mice, and starting 3 days after transfection, 50 μg of the control ASO or the mouse GREB1-targeting ASO (mGREB1 ASO-5715) were subcutaneously administered twice weekly. Six to seven weeks after transfection, the appearance of liver tumors was observed, and the liver weight was analyzed as a percentage of total body weight. The results showed that mGREB1 ASO-5715 administration tended to suppress liver tumor formation (see Figure 17c). Furthermore, real-time PCR analysis of GREB1 mRNA levels in each liver tumor confirmed that GREB1 expression was suppressed in tumor tissue in BMY mice treated with mGREB1 ASO-5715 (Figure 17d).

[0178] [Table 9]

[0179] 2-9. GREB1 is overexpressed in neuroblastoma We analyzed GREB1 mRNA expression in various human cancer cell lines using the dataset from The Cancer Cell Line Encyclopedia 1 (CCLE). GREB1 was highly expressed in breast cancer cell lines, as previously reported, as well as skin cancer (melanoma) cell lines, neuroblastoma cell lines, and some hepatocellular carcinoma cell lines (see Figure 18A). Analysis of GREB1 expression at the protein level in neuroblastoma cell lines SKNDZ, CHP212, and NBTU110 revealed expression levels comparable to those in the hepatoblastoma cell line HepG2, which also expresses GREB1, and higher than those in the hepatocellular carcinoma cell lines Hep3B and JHH7, which also express GREB1 (see Figure 18B). We also immunohistochemically examined the expression of GREB1 and β-catenin in 13 neuroblastoma tissues. As previously described, GREB1 is expressed as a downstream target gene of β-catenin in hepatoblastoma, a pediatric liver cancer. GREB1 was detected in the nuclei of tumor cells in four cases (30.7%) of neuroblastoma, and β-catenin was detected in the cytoplasm or nucleus in 11 cases (84.6%) (see Figure 18C). No significant correlation was observed between the expression of GREB1 and β-catenin (see Figure 18D). These results suggest that GREB1 is overexpressed in neuroblastoma cells in a β-catenin-independent manner.

[0180] We investigated the effect of GREB1 on cell proliferation in neuroblastoma cells. When GREB1 expression was suppressed in CHP212 cells using antisense oligonucleotides (ASOs), cell proliferation was suppressed in plate culture (see Figure 18E). These results demonstrated that GREB1 is important for cell proliferation in neuroblastoma.

[0181] 2-10. GREB1 is overexpressed in some hepatocellular carcinomas and is involved in cell proliferation and tumorigenesis As mentioned above, GREB1 is expressed as a downstream target gene of β-catenin in hepatoblastoma, a type of pediatric liver cancer. The CCLE dataset also suggested that GREB1 is highly expressed in some adult hepatocellular carcinomas (see Figure 18A). Therefore, we immunohistochemically examined the expression of GREB1 and β-catenin in 210 hepatocellular carcinoma tissues. Results included cases with high nuclear expression of GREB1 alone, cases with high nuclear or cytoplasmic expression of β-catenin alone, cases with high expression of both GREB1 and β-catenin, and cases with no staining for either GREB1 or β-catenin (see Figure 19A). GREB1 was detected in the nuclei of tumor cells in 73 hepatocellular carcinoma cases (34.7%), and β-catenin was detected in the cytoplasm or nucleus in 93 cases (44.2%) (see Figure 19B). Furthermore, a significant positive correlation between GREB1 and β-catenin expression (Phi correlation coefficient: 0.41, P value <0.0001) was observed (see Figure 19B). Furthermore, we analyzed GREB1 expression by Western blotting in Hep3B, JHH7, and Huh7 cells, which had high GREB1 expression from the CCLE dataset, and in HLE and HLF cells, which had low GREB1 expression. The results showed that GREB1 expression in Hep3B, JHH7, and Huh7 hepatocellular carcinoma cells was slightly lower than in the GREB1-positive breast cancer cell line MCF7 and the hepatoblastoma cell line HepG2, but higher than in the weakly GREB1-expressing hepatoblastoma cell line Huh6. On the other hand, GREB1 expression was not observed in HLE and HLF cells (see Figure 19C). We investigated the β-catenin dependence of GREB1 in hepatocellular carcinoma cell lines with high GREB1 expression. When β-catenin expression was suppressed in Hep3B, JHH7, and Huh7 cells using siRNA, GREB1 expression was also suppressed in all cells (see Figure 19D). These results indicate that GREB1 is overexpressed in a β-catenin-dependent manner in some hepatocellular carcinoma cells.

[0182] Next, we examined the effect of GREB1 on cell proliferation in hepatocellular carcinoma (HCC) cells. Silencing of GREB1 expression in Hep3B and JHH7 cells with siRNA suppressed cell proliferation in plate culture (see Figure 19E). Furthermore, to clarify the role of GREB1 in HCC tumorigenesis in vivo, we generated GREB1-knockout Hep3B and JHH7 cells and performed xenograft subcutaneous tumor formation assays. The GREB1-knockout Hep3B and JHH7 cells significantly reduced tumor weight 6.5 weeks after transplantation (Hep3B cells) and 2 weeks after transplantation (JHH7 cells) compared with control cells (see Figure 19F and G). These results demonstrate that GREB1 plays an important role in HCC cell proliferation and tumorigenesis in vivo.

[0183] 2-11. GREB1 is overexpressed in cutaneous melanoma and is involved in cell proliferation The CCLE dataset suggested that GREB1 is frequently expressed in cutaneous melanoma cells (see Figure 18A). Therefore, we analyzed GREB1 expression by Western blotting in melanoma cell lines Mewo, SKMEL28, and G361 cells, which showed high GREB1 expression in the CCLE dataset. The results showed that GREB1 was detected at approximately 216 kDa (wild type) in hepatoblastoma HepG2 cells, whereas it was detected at approximately 100 kDa in melanoma cells (see Figure 20A). To clarify the expression of GREB1 transcript variants, we investigated the exon expression of GREB1 in breast cancer and cutaneous melanoma using TCGA SpliceSeq (http: / / projects.insilico.us.com / TCGASpliceSeq) based on the TCGA dataset. Four isoforms (IS) (transcript variants) of GREB1 are registered in the Universal Protein Resource (UniProt). IS1, also known as GREB1a, is the full-length form of 1949 amino acids (216 kDa); IS2, also known as GREB1b, is a 457 amino acid (49 kDa) transcript with a C-terminal deletion of amino acids 458-1949; IS3, also known as GREB1c, is a 409 amino acid (43 kDa) transcript with a C-terminal deletion of amino acids 410-1949; and IS4 is a 947 amino acid (107 kDa) transcript with an N-terminal deletion of amino acids 1-1002 (see Figure 20B). GREB1 consists of 38 exons, and expression was observed throughout the protein-coding exons (exons 4-38) in breast cancer tissues. In contrast, GREB1 expression was observed from exon 24 onward in cutaneous melanoma, with IS4 specifically expressed (see Figure 20B). On TCGA, the only cancers expressing IS4 were skin and intraocular melanoma (data not shown).

[0184] To clarify the regulatory mechanism of GREB1 expression in cutaneous melanoma, we used the TCGA dataset to identify a group of genes whose expression is positively correlated with GREB1 in melanoma. Among these genes, the top 10 genes with the highest correlation coefficients included MITF and eight of its known downstream target genes (see Figure 20C). MITF is a transcription factor that regulates the expression of melanin biosynthetic enzyme genes and regulates the differentiation of melanocytes and retinal pigment epithelial cells (RPE). Immunohistochemical analysis of GREB1 and MITF expression in human melanoma tissues revealed that GREB1 and MITF were coexpressed in the same tumor region in serial specimens (see Figure 20D). These results suggested that GREB1 may be a target gene of MITF in melanoma. Therefore, we generated MITF knockout cells in the GREB1-high-expressing melanoma cell line COLO679, and found that GREB1 expression was suppressed compared to control cells (see Figure 20E).

[0185] We investigated the effect of GREB1 on cell proliferation in cutaneous melanoma cells. Silencing of GREB1 expression in SKMEL28 and COLO679 cells using siRNA (SKMEL28 cells) or antisense oligonucleotide (ASO) (COLO679 cells) suppressed cell proliferation in both cell cultures (see Figure 20, F and G). These results demonstrate that GREB1 IS4, specifically expressed in cutaneous melanoma, is important for melanoma cell proliferation.

Claims

1. A therapeutic agent for a GREB1-positive tumor that is not sex hormone sensitive, comprising: It contains a substance that suppresses the expression of GREB1 as an active ingredient, the substance is at least one nucleic acid drug selected from the group consisting of siRNA, shRNA, dsRNA, antisense nucleic acid, and ribozyme against GREB1; The above therapeutic agent, wherein the tumor is hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma.

2. A method for predicting the presence or absence of a GREB1-positive tumor that is not sex hormone sensitive, comprising: measuring the expression level of GREB1 in liver tissue, skin tissue, or nerve tissue collected from a subject using at least one selected from the group consisting of an anti-GREB1 antibody or a fragment thereof, and a primer capable of hybridizing with GREB1 mRNA or GREB1 cDNA; the tumor is hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma; The above-mentioned testing method, in which the measurement is performed by immunostaining, and if GREB1 expression is observed in 5% or more of the area suspected of being a tumor, this indicates a high possibility of being affected by the tumor; or, alternatively, the measurement is performed by quantitative PCR, and if the amount of GREB1 in a cell lysate from a site suspected of being a tumor is greater than the amount of GREB1 in a cell lysate from a non-tumor site of the same case, this indicates a high possibility of being affected by the tumor.

3. A test agent for use in the test method according to claim 2, The above test agent, comprising as a detection reagent for GREB1 at least one member selected from the group consisting of an anti-GREB1 antibody or a fragment thereof, and a primer capable of hybridizing with GREB1 mRNA or GREB1 cDNA.

4. Use of a substance that suppresses GREB1 expression for the manufacture of a therapeutic agent for a GREB1-positive tumor that is not sex hormone sensitive, the substance is at least one nucleic acid drug selected from the group consisting of siRNA, shRNA, dsRNA, antisense nucleic acid, and ribozyme against GREB1; The above use, wherein the tumor is hepatoblastoma, hepatocellular carcinoma, malignant melanoma, or neuroblastoma.

Images