Turkish Journal of Biology Turk J Biol (2014) 38: 898-905 http://journals.tubitak.gov.tr/biology/ © TÜBİTAK Research Article doi:10.3906/biy-1405-73

Loss of heterozygosity in ING3 and ING5 in breast cancer

1, 1 1 1 2 Esra GÜNDÜZ *, Gökhan NAS , Muradiye ACAR , Eyyüp ÜÇTEPE , Mikdat BOZER , 1 3 1,4 Murat ÖZNUR , Reyhan BAYRAK , Mehmet GÜNDÜZ 1 Department of Medical Genetics, Faculty of Medicine, Turgut Özal University, Ankara, Turkey 2 Department of General Surgery, Faculty of Medicine, Turgut Özal University, Ankara, Turkey 3 Department of Pathology, Faculty of Medicine, Turgut Özal University, Ankara, Turkey 4 Department of Otolaryngology, Faculty of Medicine, Turgut Özal University, Ankara, Turkey

Received: 28.05.2014 Accepted: 27.07.2014 Published Online: 24.11.2014 Printed: 22.12.2014

Abstract: The tumor suppressor genes (TSGs) ING3 and ING5, members of the inhibitor of growth family, are effective in inhibition of cell growth and induction of apoptosis. However, in many cancer types, one of the alleles of a TSG is lost through carcinogenesis, while the remaining allele is usually inactivated through a process called loss of heterozygosity (LOH). Previous studies in head and neck cancer revealed that allelic loss and reduced expression is a common pattern of ING gene family members. Fifty paraffin-embedded breast cancer tissues were analyzed by polymerase chain reaction and denatured-polyacrylamide gel electrophoresis for LOH status. The allelic deletion frequency of ING3 and ING5 were detected as 14% and 17% in breast cancer patients, respectively. No significant relationship was detected between ING3 LOH status and clinicopathological variables. Our data also suggest that both ING3 and ING5 LOH statuses have no significant effect in overall survival and disease-free survival of breast cancer patients. These results provide a rational explanation and relative contribution for the complexity of tumor formation, whereby allelic loss of ING3 and ING5 genes is not a major factor for breast cancer but is rather a part of a larger complex mechanism.

Key words: Clinicopathological factors, early diagnosis, survival analysis, tumor suppressor genes

1. Introduction tumor suppressor genes (TSGs), is considered one of the Breast cancer is the most deadly form of cancer after lung most important mechanisms in tumorigenesis. Hence, cancer, and after skin cancers it is the most common cancer localization and identification of TSGs are of great interest among women in the world. About 1 in 8 (12%) women in for a better understanding of cancer and the development the United States will have breast cancer in their lifetime. of molecular diagnostics as well as new drugs. One of the According to the American Cancer Society, 232,340 new important steps for the identification of a TSG is loss of women had an invasive breast cancer diagnosis and about heterozygosity (LOH) analysis. LOH is commonly seen 40,000 women died from breast cancer in the United States in cancer, where it denotes the lack of a functional TSG in in 2013. It is the cause of death for 1 out of every 36 women the lost region. On the other hand, the other allele serves (about 3%) (American Cancer Society, 2013). as a functional gene, so this loss does not cause illness. Breast cancer mortality has been decreasing since However, a point mutation, nondisjunction during mitosis, 1989, and this decline has been seen especially in women or deletion of a segment can inactivate the younger than 50. It is known that earlier detection through remaining copy of the TSG, leaving no TSG to protect the screening and increased awareness about the disease body, and then the individual goes on to develop cancer has led to these declines, as well as enhanced treatment (Knudson, 1971). (Andersson and Janzon, 1997). Early diagnosis of breast The first member of the inhibitor of growth (ING) cancer before symptoms emerge is the most effective form family was discovered by Riabowol’s group in 1996 by using of treatment. Thus, understanding the genetic background subtractive hybridization (Garkavtsev et al., 1996). Our of breast cancer is essential for early diagnosis and finding group identified its genomic structure in 2000 (Gündüz et new biomarkers. al., 2000). We showed that ING1 encodes a 33-kDa Breast cancer is the end point of multiple genetic events (p33ING1b). In that study, ING1 was also shown to be a in the breast tissue. One of these events, inactivation of tumor suppressor gene. The loss ofING1 gene expression * Correspondence: [email protected] 898 GÜNDÜZ et al. / Turk J Biol has been shown in numerous cancer types and ING1 decreased in 44% of human primary breast cancers and gene knockout mice have been found to be cancer-prone. 100% of established breast cancer cell lines. Decreased Subsequently, we and other groups identified 4 additional ING1 expression has been identified in many solid and human ING genes (ING2–ING5) (Gündüz et al., 2000; blood tumors (Gündüz et al., 2009). Additionally, ING2, Gündüz et al., 2002; Shiseki et al., 2003; Gong et al., 2005; ING3, and ING4 gene expression is decreased in human Soliman and Riabowol, 2007; Gündüz et al., 2008; Unoki melanoma cancer (Aguissa-Toure et al., 2011). et al., 2008; Coles and Jones, 2009). Among these genes, The role of ING3 in regulation of the cell cycle and ING3, encoding a 46.8-kDa protein (p47ING3) with a apoptosis has been determined. Ectopic expression of C-terminal plant homeodomain (PHD) finger motif, ING3 in RKO cells, a colon cancer cell line, resulted in was subsequently identified through a computational decreased colony formation, possibly by reducing the domain search in 2003 (Gündüz et al., 2000). It has been number of cells in S phase (Nagashima et al., 2003; Shiseki mapped on at the 7q31 and encodes 2 et al., 2003). variants (Figure 1). The other members of the ING family, Although we have limited information about ING5 including ING5, were identified through a computational function, very recent transfection experiments have search in the same year (Shiseki et al., demonstrated that ING5 reduces colony-forming efficiency, 2003). ING5 has been mapped on at locus inhibits S-phase progression, and induces apoptosis in a 2q37.3 and encodes 1 variant (Figure 2). -dependent manner. ING5 can also induce expression ING genes are TSGs. They play roles in a wide of the cyclin-dependent kinase inhibitor p21 (Shiseki et al., variety of cellular processes. They are involved in cell 2003), a p53-target gene. These results suggest that ING5 cycle checkpoints and cell cycle progression as well as has a role in regulation of cell growth and p53 activity, but induction of apoptosis. ING1 expression is significantly further studies of ING5 function are needed.

Figure 1. ING3 gene is composed of 13 exons and has 2 transcripts. Domains are as indicated above.

Figure 2. ING5 gene is composed of 8 exons and has 1 transcript. Domains are as indicated above.

899 GÜNDÜZ et al. / Turk J Biol

To our knowledge, there has been no prior study was loaded into denatured-polyacrylamide gel. Next, examining the LOH status of ING3 and ING5 in breast 2–4 µL of the PCR mixture was mixed with 2X loading cancer. In this study, we examine for the first time the LOH dye (95% formamide, 20 mM EDTA, 0.05% bromphenol status of ING3 and ING5 in breast cancer tissues. We also blue, and 0.05% xylene cyanol, ddH2O), heat denatured, investigate the relation between the genes’ LOH statuses chilled on ice, and then electrophoresed through an 8% and breast cancer patient survival. polyacrylamide gel containing 8 M urea. DNA fragments were visualized by the silver staining method (Bassam and 2. Materials and methods Gresshoff, 2007). 2.1. Tissue samples LOH was determined if one of the heterozygous alleles Formalin-fixed paraffin-embedded archival matched showed at least 50% reduced intensity in tumor DNA as tumor and macroscopically normal samples of 50 patients compared with the corresponding normal DNA (Figure (median age: 51.9 years; range: 27–78 years) diagnosed 3). This decision was made by direct visualization in cases with invasive ductal breast cancer between 2006 and 2011 with clear deletion as compared to normal allele. were obtained from the Department of Pathology, Faculty 2.4. Statistical analysis of Medicine, Turgut Özal University. All tissues were frozen The Mann–Whitney U test as a nonparametric method in liquid nitrogen immediately after surgery and stored at and Pearson’s chi-square test were used to evaluate the –80 °C until the extraction of DNA and RNA. All tumors correlation between ING3 and ING5 genes and the were confirmed as invasive ductal breast carcinoma in the clinicopathological characteristics of the patients (Table). Department of Pathology. Survival curves were calculated according to the Kaplan– These samples were collected after acquisition of Meier method. For comparison of survival between LOH- informed consent from each patient and approval of the negative and LOH-positive ING genes, the log-rank test study by the Institutional Human Ethics Committee of was used. Overall survival in months was calculated from Turgut Özal University. the day after surgery to the last follow-up examination or Tumor histology was determined according to the death. The survival periods of the patients who were still 2003 criteria of the World Health Organization. Tumors alive were noted along with the date of the most recent were graded according to the Bloom–Richardson grading follow-up appointment. The duration of disease-free scale. Each tumor and normal breast tissue sample was survival (DFS) was determined from the day after surgery histopathologically confirmed for the presence or absence to the initial recurrence of the surgically resected cancer, of cancer cells. Tumor samples containing more than 70% as evaluated by clinical examination. SPSS 15 for Windows tumor cells were selected. ER, PR, and HER2 statuses were (SPSS, Chicago, IL, USA) was used for all statistical determined by immunohistochemistry. manipulations and all results with a P-value of less than 0.05 were considered to be statistically significant. 2.2. Extraction of genomic DNA Tumor and normal tissues were deparaffinized and 3. Results rehydrated in a decreasing alcohol series prior to DNA We included 50 breast cancer samples in the study. We were extraction by use of the Pure Link Genomic DNA Mini unable to carry out PCR for the ING5 gene in some samples. Kit (Invitrogen, USA). The extracted genomic DNA was Our results showed that 14% of the patients presented LOH finally eluted in 100 µL of Tris-HCl buffer (10 mM, pH of ING3, 58% of the patients were normal for ING3, and 9.0). 28% of the patients’ results were not informative (Figures 2.3. Polymerase chain reaction and denatured- 4 and 5). Likewise, 17% of the patients presented LOH of polyacrylamide gel electrophoresis the ING5 gene, 62% of the patients had no deletion at the Specific primers were designed for ING3 and ING5 gene ING5 gene locus, and 21% of the patients’ results were not regions for allelic deletion. Polymerase chain reaction informative (Figures 4 and 5). Lymphovascular invasion (PCR) was carried out in 25 µL of total reaction volume was identified in 35% of the patients. Nipple invasion was with 12.5 µL of master mixture (QIAGEN), 0.5 µL (20 identified in 70% of the patients, while ER was identified pmol) of each primer, 2 µL of genomic DNA, and 9.5 µL of in 60.5% of the patients and PR in 60.5% of the patients. dH2O. Initial denaturation at 94 °C for 5 min was followed Fifty-eight percent of the patients were smokers and 2.3% by 35 cycles of a denaturation step at 94 °C for 30 s, an of patients consumed alcohol. annealing step at 54 °C (ING3) or 60 °C (ING5) for 30 s, First, we investigated the relationship between and an extension step at 72 °C for 1 min. A final extension LOH status of ING3 and ING5 genes and most clinical step at 72 °C for 5 min was added. Amplicons were run markers. No significant relationship was detected between in agarose gel electrophoresis and an appropriate amount ING3 LOH status and tumor grade (P = 0.967), absence of PCR amplicon for both normal and tumor samples or presence of lymph-node metastasis (P = 0.185),

900 GÜNDÜZ et al. / Turk J Biol

Figure 3. Primary LOH data from 6 representative breast cancer samples. N, Normal DNA; T, tumor DNA. Case 5 shows LOH of ING5 and case 10s show LOH of ING3; case 6 and case 11 are not informative; case 4 and case 13 show retention of heterozygosity.

Table. Clinicopathological characteristics of patients.

Criteria n % Criteria n % Criteria n % Age (mean) 52.6 Smoking 25 58 Alcohol usage 1 2.3 Tumor status Node status T1 9 21 N0 12 28 ING3 LOH- 21 58 T2 25 58 N1 13 30 ING3 LOH+ 5 14 T3 5 11 N2 11 25.6 ING5 LOH- 30 62 T4 1 2.3 N3 5 11.6 ING5 LOH+ 8 17 Lymphovascular invasion 15 35 ER 26 60.5 Chemotherapy 15 40 Nipple invasion 30 70 PR 26 60.5 Radiotherapy 14 32.6 lymphovascular invasion (P = 0.782), nipple invasion (P = 0.213), age (P = 0.166), smoking (P = 1.000), alcohol = 0.199), age (P = 0.845), smoking (P = 1.000), alcohol consumption (P = 0.564), and application of chemotherapy consumption (P = 0.739), and application of chemotherapy (P = 0.112) or radiotherapy (P = 0.619). (P = 0.201) or radiotherapy (P = 0.626). 3.1. ING3 and ING5 genes and survival analysis Likewise, no significant relationship was detected Both overall survival (OS) and DFS were determined. All between LOH status of ING5 and tumor grade (P = 0.711), patients were enrolled in a follow-up program. The follow- absence or presence of lymph-node metastasis (P = 0.107), up time was between 22 and 90 months. Valid follow-up lymphovascular invasion (P = 0.100), nipple invasion (P data were available for 42 (84%) of 50 breast cancer patients.

901 GÜNDÜZ et al. / Turk J Biol

Figure 4. Percentage of ING3 and ING5 LOH samples with respect to normal and noninformative samples.

Figure 5. LOH analysis of ING3 and ING5 genes in breast cancer patients. Schematic representation of LOH distribution. Case numbers are shown across the top. Genes are shown to the left.

Correlations between ING3 and ING5 LOH statuses and the for the log-rank test = 0.125; Figure 7b). These data suggest patients’ OS and DFS were analyzed using the univariate that neither ING3 nor ING5 LOH status has significant Kaplan–Meier method. The mean overall survival in the effect on OS and DFS of breast cancer patients. ING3 gene LOH-negative group was 54 months (range: 22–90 months) and that in the LOH-positive group was 40 4. Discussion months (range: 28–71 months) (P-value for log-rank test = Although the exact functions of the ING genes have not 0.824). The log-rank test showed that patients in the ING3 been fully clarified, the gene products mainly have roles LOH-positive group did not have significantly shorter OS in transcriptional regulation, apoptosis, the cell cycle, than those in the ING3 LOH-negative group (P = 0.824; angiogenesis, and DNA repair through p53-dependent Figure 6a). The mean DFS in the ING3 gene LOH-negative and -independent pathways, and they form complexes group was 49 (range: 22–90) months and that in the ING3 with histone acetyltransferases and histone deacetylases gene LOH-positive group was 40 (range: 21–51) months (Campos et al., 2004). (P-value for the log-rank test = 0.767; Figure 6b). The mean Loss of heterozygosity in ING3 was found in OS in the ING5 gene LOH-negative group was 61 months ameloblastoma, renal cell carcinoma, and head and neck and that in the ING5 gene LOH-positive group was 51.5 squamous cell carcinoma (HNSCC; Shridhar et al., 1997; months (P-value for log-rank test = 0.102; Figure 7a).The Borkosky et al., 2010). Decreased ING3 mRNA expression mean DFS in the LOH-negative group was 54 months and was also found in HNSCC and melanoma (Wang and Li, that in the LOH-positive group was 51.5 months (P-value 2006; Gündüz et al., 2008).

902 GÜNDÜZ et al. / Turk J Biol

Figure 6. Overall and disease-free survival in the groups of breast cancer patients with LOH-negative and LOH-positive ING3 gene. Kaplan–Meier survival curves for the total number of cases (n = 26) were stratified by ING3 LOH states. The cases were divided into LOH-negative and-LOH positive groups. (a) The mean overall survival in the LOH-negative group was 54 months and that in the LOH- positive group was 40 months (P-value for log-rank test = 0.824). (b) The mean disease-free survival in the LOH-negative group was 49 months and that in the LOH-positive group was 40 months (P-value for the log-rank test = 0.767).

Figure 7. Overall and disease-free survival in the groups of breast cancer patients with LOH-negative and LOH-positive ING5 gene. Kaplan–Meier survival curves for the total number of cases (n = 37) were stratified by ING5 LOH states. The cases were divided into LOH-negative and LOH-positive groups. (a) The mean overall survival in the LOH-negative group was 61 months and that in the LOH- positive group was 51.5 months (P-value for log-rank test = 0.102). (b) The mean disease-free survival in the LOH-negative group was 54 months and that in the LOH-positive group was 51.5 months (P-value for the log-rank test = 0.125).

903 GÜNDÜZ et al. / Turk J Biol

Our previous study on head and neck cancer showed of ING5 protein was observed in carcinoma in comparison that about half of 71 tumor samples demonstrated with adjacent mucosa in 14 cases (77.8%; P < 0.05), and downregulation of ING3 compared to their matched 71.4% (10/14) of carcinoma cases exhibited upregulation normal counterparts. Even if most clinicopathological of ING5 mRNA. Nuclear ING5 expression was negatively variables were not significantly related to ING3 correlated with tumor size, depth of invasion, degree of downregulation or p53 mutation status, a significant dedifferentiation, and cancer staging (P < 0.05). In contrast, relationship was determined in terms of OS between the cytoplasmic ING5 expression was positively correlated cases with low and normal to high ING3 expression. At 5 with depth of invasion, lymphatic invasion, and cancer years of follow-up, approximately 60% of the patients with staging (P < 0.05). It was suggested that aberrant ING5 normal to high ING3 expression had survived, whereas expression may have a role in pathogenesis, growth, and this rate was 35% in the patients with low ING3 expression. invasion of colorectal carcinomas and could be considered Multivariate analysis also showed downregulation of ING3 as a remarkable marker to estimate aggressiveness of as an independent prognostic factor for poor OS. These colorectal carcinomas (Zheng et al., 2011). results reveal that ING3 would function as a potential Increased expression of ING5 mRNA was demonstrated tumor suppressor molecule and that low levels of ING3 in gastric carcinoma in comparison with paired mucosa (P may point to an aggressive nature of head and neck cancer < 0.05). Lower expression of nuclear ING5 was detected (Gündüz et al., 2008). in gastric dysplasia and carcinoma as compared to that in Loss of heterozygosity in ING5 was found in HNSCC nonneoplastic mucosa (P < 0.05). Gastric nonneoplastic and oral cancer. Decreased ING5 mRNA expression was mucosa and metastatic carcinoma showed more expression also identified in ovarian cancer and HNSCC (Cengiz et of cytoplasmic ING5 than did gastric carcinoma and al., 2007, 2010). Our group determined that loss of the dysplasia (P < 0.05) (Xing et al., 2011). ING5 chromosome locus was related to oral cancer. In our Survival analysis showed that nuclear ING5 was study, a high rate of LOH in oral cancer was demonstrated. associated with favorable prognosis of gastric carcinoma Decreased expression of ING5 mRNA was also revealed patients (P < 0.05). It was proposed that aberrant ING5 in 61% of oral squamous cell carcinomas as compared to expression may play a role in pathogenesis, growth, and the matched normal samples. Because of tumor-specific invasion of gastric carcinomas and could be considered as mutation and downregulation of ING5 mRNA, it was a striking marker to estimate aggressiveness and prognosis suggested as a TSG in oral squamous cell carcinoma of gastric carcinoma (Xing et al., 2011). (Cengiz et al., 2010). It has also been shown that ING5 can affect sensitivity A previous study suggested that a decline in nuclear to some drugs. Mendes-Pereira et al. (2012) showed that ING5 localization and cytoplasmic translocation are silencing ING5 causes sensitivity to tamoxifen. involved in tumorigenesis and tumor differentiation in HNSCC. Nuclear ING5 may regulate the transactivation This is the first study to analyze LOH status of ING3 of target genes and may accelerate apoptosis and cell cycle and ING5 and examine its possible correlation with arrest by interacting with the p300 and p21 . ING5 clinicopathological factors. In the current study, we may function as a TSG or oncogene tightly linked with p53 analyzed the LOH status of the ING3 and ING5 genes; status, and may have an important role in the prognosis of afterwards, we compared it with the clinicopathological HNSCC patients (Li et al., 2010). characteristics of breast cancer patients. No significant Zhang et al. showed that overexpression of ING5 sup- relationship was detected between LOH statuses of the pressed cell proliferation only in the presence of INCA1, ING3 and ING5 genes and clinical markers including age, while ING5 had no effect in INCA1 ( - / - ) murine em- smoking and alcohol consumption, absence or presence of bryonic fibroblasts (MEFs). ING5 overexpression also trig- lymph-node metastasis, nipple invasion, and application gered a delay in S-phase progression, which required the of chemotherapy or radiotherapy. We did not find a inhibitor of cyclin A1 (INCA1). Furthermore, ING5 over- relationship between LOH statuses of ING3 and ING5 expression accelerated Fas-induced apoptosis in INCA1 ( genes and OS or DFS; other genetic and nongenetic factors + / + ) MEFs, while Fas antibody did not induce apoptosis could be considered to contribute to the factors related to in INCA1 ( - / - ) MEFs. Taken together, these results sug- OS and DFS, such as local recurrences, metastasis, and gest that ING5 is a growth suppressor with reduced expres- aggressive phenotype. sion in AML whose functions depend on its interaction In conclusion, our current study has confirmed that with INCA1 (Zhang et al., 2011). patients with LOH in both ING3 and ING5 have shorter Another study showed that among 18 frozen samples OS and DFS, but this is not significant statistically. There of colorectal carcinoma, significantly increased expression is also no significant relationship between the LOH

904 GÜNDÜZ et al. / Turk J Biol status of ING3 and ING5 genes and any clinical features. Acknowledgment Comparative analysis of other genes, including various This work was supported by the Scientific Research Fund ING family members, in subsequent studies would give us of Turgut Özal University and Fatih University under the more valuable and clinically applicable results. project number P53010913_2.

References

Aguissa-Toure AH, Wong RP, Li G (2011). The ING family tumor Gündüz M, Ouchida M, Fukushima K, Hanafusa H, Etani T, Nishioka S, suppressors: from structure to function. Cell Mol Life Sci 68: Nishizaki K, Shimizu K (2000). Genomic structure of the human 45–54. ING1 gene and tumor-specific mutations detected in head and American Cancer Society. Cancer Facts and Figures 2013. Atlanta, neck squamous cell carcinomas. Cancer Res 60: 3143–3146. GA, USA: American Cancer Society, 2013. Gündüz M, Ouchida M, Fukushima K, Ito S, Jitsumori Y, Nakashima T, Nagai N, Nishizaki K, Shimizu K (2002). Allelic loss and reduced Andersson I, Janzon L (1997). Reduced breast cancer mortality expression of the ING3, a candidate tumor suppressor gene at 7q31, in women under age 50: updated results from the Malmo in human head and neck cancers. Oncogene 21: 4462–4470. Mammographic Screening Program. J Natl Cancer Inst Monogr 63–67. Knudson AG Jr (1971). Mutation and cancer: statistical study of retinoblastoma. P Natl Acad Sci USA 68: 820–823. Bassam BJ, Gresshoff PM (2007). Silver staining DNA in Li X, Nishida T, Noguchi A, Zheng Y, Takahashi H, Yang X, Masuda S, polyacrylamide gels. Nat Protoc 2: 2649–2654. Takano Y (2010). Decreased nuclear expression and increased Borkosky SS, Gündüz M, Beder L, Tsujigiwa H, Tamamura R, Gündüz cytoplasmic expression of ING5 may be linked to tumorigenesis E, Katase N, Rodriguez AP, Sasaki A, Nagai N et al. (2010). and progression in human head and neck squamous cell carcinoma. Allelic loss of the ING gene family loci is a frequent event in J Cancer Res Clin Oncol 136: 1573–1583. ameloblastoma. Oncol Res 18: 509–518. Mendes-Pereira AM, Sims D, Dexter T, Fenwick K, Assiotis I, Kozarewa I, Campos EI, Chin MY, Kuo WH, Li G (2004). Biological functions of Mitsopoulos C, Hakas J, Zvelebil M, Lord CJ et al. (2012). Genome- the ING family tumor suppressors. Cellular and Molecular Life wide functional screen identifies a compendium of genes affecting Sciences 61: 2597–2613. sensitivity to tamoxifen. P Natl Acad Sci USA 109: 2730–2735. Nagashima M, Shiseki M, Pedeux RM, Okamura S, Kitahama-Shiseki M, Cengiz B, Gündüz E, Gündüz M, Beder LB, Tamamura R, Bagci Miura K, Yokota J, Harris CC (2003). A novel PHD-finger motif C, Yamanaka N, Shimizu K, Nagatsuka H (2010). Tumor- protein, p47ING3, modulates p53-mediated transcription, cell specific mutation and downregulation of ING5 detected in oral cycle control, and apoptosis. Oncogene 22: 343–350. squamous cell carcinoma. Int J Cancer 127: 2088–2094. Shiseki M, Nagashima M, Pedeux RM, Kitahama-Shiseki M, Miura K, Cengiz B, Gündüz M, Nagatsuka H, Beder L, Gündüz E, Tamamura Okamura S, Onogi H, Higashimoto Y, Appella E, Yokota J et al. R, Mahmut N, Fukushima K, Ali MA, Naomoto Y et al. (2007). (2003). p29ING4 and p28ING5 bind to p53 and p300, and enhance Fine deletion mapping of chromosome 2q21-37 shows three p53 activity. Cancer Res 63: 2373–2378. preferentially deleted regions in oral cancer. Oral Oncol 43: Shridhar V, Sun QC, Miller OJ, Kalemkerian GP, Petros J, Smith DI (1997). 241–247. Loss of heterozygosity on the long arm of human chromosome 7 in Coles AH, Jones SN (2009). The ING gene family in the regulation of sporadic renal cell carcinomas. Oncogene 15: 2727–2733. cell growth and tumorigenesis. J Cell Physiol 218: 45–57. Soliman MA, Riabowol K (2007). After a decade of study-ING, a PHD Garkavtsev I, Kazarov A, Gudkov A, Riabowol K (1996). Suppression for a versatile family of proteins. Trends Biochem Sci 32: 509–519. of the novel growth inhibitor p33ING1 promotes neoplastic Unoki M, Kumamoto K, Robles AI, Shen JC, Zheng ZM, Harris CC transformation. Nat Genet 14: 415–420. (2008). A novel ING2 isoform, ING2b, synergizes with ING2a to prevent cell cycle arrest and apoptosis. FEBS Lett 582: 3868–3874. Gong W, Suzuki K, Russell M, Riabowol K (2005). Function of the ING family of PHD proteins in cancer. Int J Biochem Cell Biol Wang Y, Li G (2006). ING3 promotes UV-induced apoptosis via Fas/ caspase-8 pathway in melanoma cells. J Biol Chem 281: 11887– 37:1054–1065. 11893. Gündüz M, Beder LB, Gündüz E, Nagatsuka H, Fukushima K, Xing YN, Yang X, Xu XY, Zheng Y, Xu HM, Takano Y, Zheng HC (2011). Pehlivan D, Cetin E, Yamanaka N, Nishizaki K, Shimizu K et The altered expression of ING5 protein is involved in gastric al. (2008). Downregulation of ING3 mRNA expression predicts carcinogenesis and subsequent progression. Hum Pathol 42: 25–35. poor prognosis in head and neck cancer. Cancer Sci 99: 531– Zhang F, Baumer N, Rode M, Ji P, Zhang T, Berdel WE, Muller-Tidow 538. C (2011). The inhibitor of growth protein 5 (ING5) depends on Gündüz M, Demircan K, Gündüz E, Katase N, Tamamura R, INCA1 as a co-factor for its antiproliferative effects. PLoS One 6: Nagatsuka, H (2009). Potential usage of ING family members e21505. in cancer diagnostics and molecular therapy. Curr Drug Targets Zheng HC, Xia P, Xu XY, Takahashi H, Takano Y (2011). The nuclear to 10: 465–476. cytoplasmic shift of ING5 protein during colorectal carcinogenesis Gündüz M, Gündüz E, Rivera RS, Nagatsuka H (2008). The inhibitor with their distinct links to pathologic behaviors of carcinomas. of growth (ING) gene family: potential role in cancer therapy. Hum Pathol 42: 424–433. Curr Cancer Drug Targets 8: 275–284.

905