Search

KIT and PDGFRA play a major role in the oncogenic process in gastrointestinal stroma tumors (GIST) and small molecules have been employed with great success to target the KIT and PDGFRA pathways in this cancer. However, approximately 10% of patients with GIST are resistant to current targeted drug therapy. There is a need to explore other potential targets. Although p53 alterations frequently occur in most cancers, studies regarding p53 in GIST have been limited. The CDKN2A/MDM2/p53 axis regulates cell cycle progression and DNA damage responses, which in turn control tumor growth. This axis is the major event required for transformation from low- to high-risk GIST. Generally, p53 mutation is infrequent in GIST, but p53 overexpression has been reported to be associated with high-risk GIST and unfavorable prognosis, implying that p53 should play a critical role in GIST. Also, Wee1 regulates the cell cycle and the antitumor activity of Wee1 inhibition was reported to be p53 mutant dependent. In addition, Wee1 was reported to have potential activity in GIST through the regulation of KIT protein and this mechanism may be dependent on p53 status. In this article, we review previous reports regarding the role of p53 in GIST and propose targeting the p53 pathway as a novel additional treatment strategy for GIST.

Additional file 2: Figure S1. Expression of TMED9 transcripts in paired adjacent (GSE76311, probe: TC05001018.hg.1; a and unpaired normal and tumor tissues (GSE102079, probe: 208757_at; b derived from patients with hepatocellular carcinoma (HCC). Figure S2. TMED9 overexpression promotes cell growth and migration of HCC36 cells. a HCC36 cells were infected with a lentivirus carrying control vector or HA-TMED9. After 72 h, the expression of TMED9 was determined by Western blotting. b, c Cell-proliferative b and migratory c abilities of HCC36 cells expressing control vector or HA-TMED9. Values are presented as the mean ± SD from three independent experiments. *p

Table S3. The summary of the cancer-related somatic mutations and CCND1 CNV gain and CDKN2A CNV loss of the 5 NPC-PDX tumors. (PDF 314 kb)

Table S7. Immnunohistochemical (IHC) staining of cyclin D1 in 139 NPC tissues from year 2002 to 2016. (PDF 415 kb)

Figure S3. C666.1 cells and PDX-C666.1 xenograft drug screening. Drug sensitivity tests in (A) C666.1 cells and (B-D) PDX-C666.1 xenograft. The changes in PDX-C666.1 (B) tumor volume, (C) tumor weight (g), and (D) mice body weight are indicated. Abbreviation, GSK, GSK126; DEC, decitabine; GEM, gemcitabine; PAL, palbociclib. (E) Flow cytometry analysis of C666.1 cells in the presence of PAL (0, 0.1, 0.5 and 1Â ÎźM). (PDF 485 kb)

Supplementary Materials and Methods. (DOCX 16 kb)

Figure S4. Drug screening in PDX-LN (NPC02F12). PDX-LN (A) tumor volume; (B) tumor weight; and (C) mice body weight change in the presence of DMSO (control), DEC (reduced dose) and PAL. (PDF 395 kb)

Figure S5. Heatmap of RNA-seq data of PDX-B and with 5 drug treatments. RNA expression profiles of PDX-B (DMSO, control) and treated with 1. GEM, 2. GSK126, 3. DEC, 4. PAL, 5. GEM+PAL. The color scale (red, yellow and blue) indicates the expression level (log10 RPKM) from high to low. (PDF 375 kb)

Table S9. PDX tumor growth during drug treatment (A) PDX-Bone tumor volume (mm3) and tumor weight (g), and (B) PDX-Bone mice body weight (g) in drug screening; (C) PDX-LN tumor volume (mm3) and tumor weight (g), and (D) PDX-LN mice body weight (g) in drug screening. (PDF 241 kb)

Table S4. RNA-seq mapping statistics of PDX-Bone in the presence of five drug treatments. (PDF 407 kb)

Table S5. Correlation between CNV detected by WES and Q-PCR. (PDF 298 kb)

Table S8. Clinical characteristics of metastatic NPC patients with FFPE tissue cyclin D1 immnunohistochemical staining (2002â 2016). (PDF 364 kb)

Table S6. CNV of CCND1, p16 and RAD52 in the 11 NPC patientsâ plasma with (A) high and (B) low EBV copy number. (PDF 783 kb)

Figure S2. CNV profile comparison between CP-409 FFPE-LN and PDX-LN. CNV profiles of NPC (A) FFPE-LN and (B) PDX-LN based on ultra-deep sequencing of CP-409. Observed copy number for each evaluated position is shown on the y-axis as a log 2 scale. Genes associated with or without copy number alteration are indicated in different colors or in grey, respectively. Correlation plots with Pearsonâ s correlation coefficient, r, is indicated. (PDF 467 kb)

Figure S1. Single nucleotide variations (SNV) of five metastatic NPC-PDX tumors. Somatic mutations (including non-synonymous missense and splice site mutations) of the five NPC-PDX tumors identified from sequencing data are listed (upper panel). Each bar represents the number of base substitutions. For the 282 SNVs, the percentage of each base substitution is indicated (lower panel). (PDF 402 kb)

Figure S6. Differentially expressed genes of PDX-Bone exposed to five different drugs. Volcano plots (left panel) generated by analysis of differential gene expression based on RNA sequencing data on PDX-Bone with five drug treatments: (A) GEM, (B) GSK126, (C) DEC, (D) PAL, and (E) GEM + PAL. The x-axis represents the log2 base fold change of PDX-B treated with different drugs and y-axis represents the p-value (â log10) for differential gene expression. Other genes that passed quality control are presented as gray dots. Up-regulated (red dots) and downregulated (green dots) genes are indicated. Enriched KEGG cellular pathways (right panel) are presented as circles according to scores from enrichment p-value (â log10) (y-axis) and topology analysis (pathway enrichment factor, x-axis). The circle color scale (red to green) indicates the significance of the pathway. The size of the circle represents the number of genes involved in the pathway. (F) The KEGG cell cycle pathway. Green boxes represent downregulated genes in the cell cycle pathway of PDX-B treated with PAL. (PDF 961 kb)

Table S1. Characteristics of total 17 metastatic NPC patients and the final 5 NPC-PDX lines (grey). (PDF 401 kb)

Figure S7. Correlation between CNV of cellular genes and low EBV copy number in NPC plasma. (A) The CNV of CCND1, CDKN2A and RAD52 in 24 NPC plasma with low EBV DNA load (

Mitotic index has been found to be an important prognostic factor in primary gastrointestinal stromal tumor (GIST). By using gene sets related to cell cycle (Proc Natl Acad Sci U S A. 2008; 22;105(3):955-60), we can stratify a group of patients with advanced GIST (J Clin Oncol 2008;26:4100-4108.) into two risk groups with distinct survival. So it is interesting to explore whether genes involved in regulation of cell cycle could be used as a target in treating GIST. Aurora kinase A (AURKA) and B (AURKB) were two important genes in cell cycle regulation. In the same cohort of patients, expression levels of both AURKA and AURKB were found higher in high risk group patients than those in low risk group. By using MTT assay, GIST882, an imatinib-sensitive GIST cell line, was relatively resistant to both AURKA inhibitor MLN8237 and AURKB inhibitor ZM447439 than GIST 48IM, an imatinib-resistant GIST cell line. MLN8237 could induce significant cell cycle arrest in both GIST cell lines, but apoptosis could only be seen in GIST48IM. On the other hand, after ZM447439 treatment, cell cycle inhibition could only be found in GIST48IM, and none of them had apoptosis. Instead, ZM447439 could induce autophagy in both cell lines. We then used immunohistochemistry (IHC) to examine 76 primary GIST patients. High expression of AURKA but not AURKB was more often seen in non-gastric GIST. In univariate analysis, patients with tumors having strong AURKA IHC staining, as well as large tumor size, high mitotic index and NIH high risk group, had significant lower survival than comparative groups. However, in the multivariate analysis, only high mitotic count and high AURKA expression were independent adverse prognostic factors. In conclusion, cell cycle-related genes could be used in predicting prognosis in advanced GIST, and cell cycle regulated genes such as AURKA and AURKB could be used as treatment targets. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 5364. doi:10.1158/1538-7445.AM2011-5364