DDA1, a novel oncogene, promotes lung cancer progression through regulation of cell cycle

Abstract Lung cancer is globally widespread and associated with high morbidity and mortality. DDA1 (DET1 and DDB1 associated 1) was first discovered and registered in the GenBank database by our colleagues. DDA1, an evolutionarily conserved gene, might have significant functions. Recent reports have demonstrated that DDA1 is linked to the ubiquitin–proteasome pathway and facilitates the degradation of target proteins. However, the function of DDA1 in lung cancer was previously unknown. This study aimed to investigate whether DDA1 contributes to tumorigenesis and progression of lung cancer. We found that the expression of DDA1 in normal lung cells and tissue was significantly lower than that in lung cancer and was associated with poor prognosis. DDA1 overexpression promoted proliferation of lung tumour cells and facilitated cell cycle progression in vitro and subcutaneous xenograft tumour progression in vivo. Mechanistically, this was associated with the regulation of S phase and cyclins including cyclin D1/D3/E1. These results indicate that DDA1 promotes lung cancer progression, potentially through promoting cyclins and cell cycle progression. Therefore, DDA1 may be a potential novel target for lung cancer treatment, and a biomarker for tumour prognosis.

Introduction DDA1 was first discovered by our colleagues through the serological analysis of recombinant cDNA expression libraries (SEREX) technique as a gene encoding a 1086-bp cDNA and a 309-bp open reading frame [1]. DDA1 encodes an 11-kD protein with 102 amino acid residues. DDA1 orthologs share 82-92% identity and include those of Arabidopsis, invertebrates and vertebrates [2,3]. We thus suggested that DDA1 could have a very important biological function across these species. DDA1, together with DET1 and DDB1, forms the DDD complex, which is recruited to specific UBE2E (E2 ubiquitin ligase) enzymes such as UBE2E1, UBE2E2 and UBE2E3 to form DDD-E2 complexes [3]. Components of the DDD-E2 complexes provide a platform for interaction with Cullin4A (Cul4A) and betatransducing (also called WD40) repeat proteins, which indicated that this complex might be involved in ubiquitination and subsequent proteasomal degradation of target proteins [2,4]. Moreover, DDA1 was shown to be a core subunit of multiple Cul4-based E3 ligases (CRLs) and could regulate CRL4s [5]. In addition, DDA1 was also shown to interact with oncoproteins such as EIF3S10, PSAP and ACTN4 [6]. These results indicate that DDA1 might be involved in tumour formation and progression.
Lung cancer is the leading cause of cancer-related mortality in the United States, accounting for more deaths than breast, prostate, colon and pancreatic cancers combined [7]. This disease is often diagnosed at a more advanced stage and has poor prognosis. The combination of platinum-based drugs and third-generation anticancer drugs or molecular targeted therapies has been used for its clinical treatment [8]. However, drug resistance leads to cancer recurrence. Therefore, the development of more effective molecular targets is important to improve current therapies for lung cancer treatment. Whether DDA1 has prognostic or therapeutic value in patients with lung cancer has not been previously assessed.
Loss of growth control is a hallmark of all cancers [9], including lung cancer. Major regulatory events leading to cell proliferation occur in the G1 phase, including the uncontrolled expression of cyclins. Cyclin E1, one of the most important cyclins, specifically drives the G1/S-phase transition during this process. Cyclin E1 binds to and activates cyclin-dependent kinase (CDK) 2, leading to phosphorylation of downstream substrates that control the initiation of DNA replication and other S-phase events [10]. Up-regulation of cyclin E1 has been found in a variety of human cancers including breast, ovarian, colorectal and lung [10][11][12][13][14]. Furthermore, D-type cyclins are also important in cancer, as they represent the ultimate downstream targets of many oncogenic pathways [15,16]. D-cyclins, binding and activating CDK4 and CDK6, phosphorylate the retinoblastoma tumour suppressor protein (pRB), and pRB-like p107 and p130 proteins, leading to activation of E2F transcription factors [17]. E2Fs then activate several downstream target genes that are required for the entry of cells into the DNA synthesis (S) phase [17]. Amplification of individual cyclin D genes and overexpression of their encoded proteins have been documented in a large proportion of human cancers [18].
These results motivated us to investigate whether DDA1 participates in lung cancer tumorigenesis and tumour progression through the regulation of cyclins and cell cycle progression and thus could serve as a prognostic biomarker for lung cancer. Here, we demonstrated that DDA1 levels were increased in both lung adenocarcinoma and lung squamous cell carcinoma and that high expression of DDA1 was associated with poor prognosis. DDA1 promoted lung cancer cell proliferation, increased cell cycle progression in vitro through G1/S transition and S-phase acceleration and regulation of cyclins. In addition, inhibition of DDA1 was shown to suppress tumorigenesis in vivo in a subcutaneous xenograft mouse model. Taken together, these results indicate that DDA1 promotes the progression of lung cancer by regulating the cell cycle, especially S phase, and cyclins such as cyclin D1/D3. DDA1 could be a powerful indicator of tumour prognosis in patients with lung cancer.

Tissue microarrays and immunohistochemistry (IHC)
Tissue microarrays containing FFPE (formalin-fixed paraffin-embedded) samples of lung cancer, adjacent tissue and normal lung tissues were purchased from US Biomax, Inc. (Rockville, MD, USA, LC10012, n = 100; T047, n = 18). Tissue microarrays with survival data were purchased from Shanghai Outdo Biotech CO. LTD. (Shanghai, China, HLug-Ade150Sur-02, n = 150; HLug-Squ150Sur-02, n = 150). The institutional review board approved the use of de-identified samples; informed consent was obtained from all patients. A total of 418 tissues were analysed for DDA1 expression by IHC according to the manufacturer's recommendations (Vector Lab Inc., Burlingame, CA, USA). IHC scores were calculated as previously described [19].
Quantitative PCR (qPCR), western blotting and immunofluorescence qPCR, Western blotting and immunofluorescence were performed as described previously [20]. For 5-bromo-2 0 -deoxyuridine (BrdU) staining, Cells were probed by BrdU incorporation for 30 min., and then, cells were fixed and treated with 1.5 M HCl for 30 min. at room temperature and washed before blocking. Primers used for qPCR are summarized in Table S2. Antibodies are provided in supplemental materials.

In vitro cell growth and colony formation assay
For cell growth assays, transfected cells were seeded at 2 9 10 3 cells per well and six wells for each group in 96-well plates. A Cell Counting Kit-8 (Dojindo, Shanghai, China) was used, and absorbance was measures at 450 nm for each well at different time-points using a microplate reader (Thermo fisher scientific, Waltham, MA, USA). For colony formation assays, transfected cells were plated at 500-1000 cells per well and three wells for each group into six-well plates and cultured for approximately 14 days, followed by crystal violet staining.
Flow cytometry assay for cell cycle Cells were collected and fixed with 70% ethanol. The cell pellet was then treated with 50 lg/ml propidium iodide (PI, Sigma-Aldrich, St. Louis, MO, USA) containing 0.1 mg/ml RNase A (Sigma-Aldrich) and 0.1% Triton X (Sigma-Aldrich), which was followed by flow cytometric analysis (BD Biosciences, San Jose, CA, USA). Data were analysed using FlowJo 9.1 software (Tree Star Inc., Ashland, OR, USA).

Cell synchronization and cell cycle analyses
For thymidine and nocodazole blocking and release experiments, cells were treated with 2 mM thymidine (Sigma-Aldrich) for 24 hrs and released for 9 hrs, followed by 200 ng/ml nocodazole (Sigma-Aldrich) for 12-18 hrs. For cell cycle analyses, processing was performed as described previously herein.

Xenograft tumours in nude mice
Lung cancer cells A549/LV-mock, A549/LV-DDA, H1299/LV-shMock and H1299/LV-shDDA1 (1 9 10 7 cells in 100 ll of RPMI-1640) were injected subcutaneously into the flanks of female BLAB/c nude mice (5 week old, Vital River Laboratories (VRL), China). Tumour size was determined by collecting length and width with a sliding caliper every 3 days, and calculating the tumour volume (mm 3 ) as length 9 (width) 2 9 0.52. When mice were killed, tumours from each animal were collected, weighed and used for histopathological studies and Western blotting. Procedures involving animals conformed to the guidelines of the Institutional Animal Care and Use Committee of West China Hospital, Sichuan University.

Statistical analysis
A chi-square test or Fisher's exact test was used to compare the differences in categorical variables. A Student's t-test or an one-way analysis of variance was used to analyse differences in continuous variables. A two-way ANOVA analysis, Huynh-Feldt correction and Tukey's range test were used to analyse tumour volumes. Kaplan-Meier analysis with log-rank tests was used to evaluate overall survival. SPSS 22.0 statistical software (SPSS Inc., Chicago, IL, USA) was used to analyse all data. P < 0.05 was regarded as statistically significant.

DDA1 expression profile in mouse tissue and lung cancer lines
As a novel gene, the DDA1 expression profile had not been documented in mouse tissues and organs, previously. We analysed DDA1 protein and mRNA by Western blot and RT-PCR, respectively, in 14 types of organs or tissues of C57BL/6j mouse under normal physiological conditions. We found that DDA1 expression was relatively lower in the lung and heart, while it was relatively higher in the spleen and thymus ( Fig. 1A and B). We further wondered whether DDA1 was dysregulated in lung cancer cell lines compared to normal lung cells. Lung fibroblast MRC-5 cells were used as a normal control, and 15 different lung cancer cell lines were assessed, and we found that both protein and mRNA were higher in cancer cell lines than in normal cells ( Fig. 1C and D). These results indicate that the overexpression of DDA1 in lung cancer cells is common. Thus, DDA1 could be a molecular marker of lung cancer. Whole-cell lysates were used to validate DDA1 and GAPDH (as a loading control) expression by Western blot analysis. (B) H1299 and H292 cells were transfected and then treated and analysed as in (A). (C) A549 and H441 cells were plated in 96-well plates, and CCK8 measurements were performed every day as indicated (n = 6, ***P < 0.001, Student's t-test). (D) H1299 and H292 cells were plated in 96-well plate, and CCK8 measurements were performed every day as indicated (n = 6, ***P < 0.001, Student's t-test). (E, G) A549 and H441 cells were subjected to colony formation assays in six-well plates and cultured for 14 days followed by crystal violet staining. Colony number was calculated and analysed (n = 3, **P < 0.01, ***P < 0.001, Student's t-test). (F, H) H1299 and H292 cells were subjected to colony formation assays as in (E, G) (n = 3, **P < 0.01, ***P < 0.001, Student's t-test). (I, J) Tumour volume (n = 5, *P < 0.05, twoway ANOVA) and end-stage tumour weight (n = 5, *P < 0.05, Student's t-test) after injection of A549 cells stably transduced with lentivirus mock (Mock) or lentivirus DDA1 (DDA1) into nu/nu mice (n = 5). (K, L) Tumour volume (n = 5, ***P < 0.001, two-way ANOVA) and end-stage tumour weight (n = 5, **P < 0.01, Student's t-test) after injection of H1299 cells stably transduced with lentivirus shmock (shMock) or lentivirus shDDA1 (shDDA1) into nu/nu mice.

DDA1 promotes cell proliferation and colony formation in vitro
To explore the functional role of the DDA1 in regulating lung cancer development in vitro, DDA1 was overexpressed in A549 and H441 cells, which express endogenous DDA1 at a low level, through transfection with a plasmid encoding DDA1 ( Fig. 2A) . 2F and H). Taken together, DDA1, in lung cancer cells, determined the cell proliferative and colony formation abilities in vitro.

DDA1 affects cell cycle progression
To determine how DDA1 affects cancer cell prognosis and cell survival in vitro, we assessed the cell proliferation rate by CFDA-SE (5-(and-6)-carboxyfluorescein diacetate, succinimidyl ester) staining and flow cytometry. After staining with CFDA-SE for 4 days, the fluorescence intensity of A549 cells overexpressing DDA1 was lower than that of control cells, indicating that DDA1 increased the frequency of cell division (Fig. 3A). In contrast, inhibition of DDA1 resulted in a higher proportion of cells exhibiting high fluorescence intensity, which suggested that H1299 cells grow slowly with ablation of DDA1 (Fig. 3B). As cell division is a direct reflection of cell cycle, we then assessed the proportion of cells in different cell cycle phases through PI staining and flow cytometry. We found that DDA1 overexpression resulted in a decreased proportion of G0/G1-phase (Mock: 55.60 AE 0.69%

DDA1 promotes G1/S-phase cell cycle transition
To examine the effect of DDA1 on different phases of cell cycle progression, we first assessed the cell cycle following thymidine and nocodazole synchronization. After synchronized to G2/M by thymidine and nocodazole, cells were transferred to normal cell culture medium for different durations and the G1/S transition to S-phase ratio was assessed. We found that up-regulation of DDA1 resulted in reducing the G1/S transition duration and an increased proportion of cells in S phase ( (Figs 3F and S3B). Furthermore, to examine the S-phase kinetics in lung cancer cells, we monitored BrdU incorporation in asynchronized cells. Consistent with previous results, overexpression   Fig. 3G and H), respectively. These results suggest that DDA1 is responsible for G1/S-phase transition and cell proliferation in lung cancer cells.

DDA1 affects G2/M phase following G1/S transition
The median number of BrdU-positive cells was altered after DDA1 up-regulation or knockdown. As a result, the ratio of cells in the mitotic phase was affected, consistent with the fact that S-phase cells yield M-phase cells and that DDA1 altered the proportion of S-phase cells. By measuring phospho-histone H3 in these cells, we observed increased (Mock: 2.08 AE 0.07% versus DDA1: 3.50 AE 0.08%) or decreased (shMock: 2.31 AE 0.07% versus shDDA1: 1.41 AE 0.08%) mitotic entry with overexpression or silencing of DDA1 (Fig. 4A-D), respectively, in agreement with our previous results. These results are consistent with the hypothesis that lung cancer cells enter mitosis despite replication stress and that delayed mitotic entry is primarily a consequence of reduced replication.

DDA1 promotes cell proliferation through regulating cyclins
To explore the molecular mechanism of how DDA1 affects lung cancer, we suggested that DDA1 regulates the cyclins, which are the key regulators of cell cycle progression and proliferation. Overexpression or knockdown of DDA1 in lung cancer cells induced increase or decrease, respectively, in the expression of cyclin D1, cyclin D3, cyclin E1 and cyclin B1; increased expression of these cyclins is indicative of proliferation, G1 to S-phase transition and subsequently M-phase progression (Fig. 5A). In addition, similar results were observed in DDA1-deficient (78.80 AE 6.99% versus 9.50 AE 1.85%) xenografts from nude mice (Fig. 5B) and cyclin B1 (32.10 AE 3.12% versus 9.40 AE 1.96%) levels than control tissues (Fig. 5C and D). These results indicate that DDA1  High expression of DDA1 correlates with lung cancer and poor prognosis DDA1 promoted lung cancer tumorigenesis in vitro and in vivo; however, the relationship between DDA1 expression and human lung cancer was still unclear. To assess the expression of DDA1 in lung cancers, we performed immunohistochemistry (IHC) of DDA1 expression using a commercially available tissue microarray (TMA) containing 118 cases of lung cancer, adjacent tissue and normal tissue. Negative and positive staining of DDA1 in the epithelial cells was observed in the normal lung tissue, whereas DDA1 expression was high in lung cancer and the adjacent region, in most cases (Figs 6A and S4). DDA1 was expressed in the cytoplasm and nucleus, and its expression in cancer epithelial and stromal regions varied. To quantitate the protein expression of DDA1, we scored the staining intensity and proportion using five grades and found that DDA1 was significantly higher in lung cancer tissue and adjacent tissue than in normal tissue (**P < 0.01) (Figs 6B and S5). Next, we proceeded to determine whether overexpression of DDA1 is 1539 associated with prognosis. In both lung adenocarcinoma and squamous cell carcinoma, five-year survival rates were much higher with low expression of DDA1 than with high expression, as assessed by the Kaplan-Meier method (*P < 0.05) (Fig. 6C).
The overexpression of DDA1 in lung cancer, based on positivity by IHC, was significantly associated with tumour size (P = 0.002) and AJCC stage (P = 0.016) ( Table 1). However, the expression of DDA1 had no significant relationship with other clinical characteristics including gender, age, tumour type, positive LN number and histological grade.
Univariate and multivariate Cox regression analyses were performed to evaluate the association between DDA1 and clinical factors, with prognostic hazards. Based on univariate Cox analysis, DDA1 (P = 0.003), tumour size (P = 0.012), positive LN number (P = 0.022) and AJCC stage (P < 0.001) were found to have an adverse impact on overall survival in patients with lung cancer ( Table 2). Based on multivariate analysis, DDA1 (P = 0.005), tumour size (P = 0.012), positive LN number (P < 0.001) and AJCC stage (P < 0.001) were significant predictors of overall survival (Table 2). These results indicate that DDA1 is up-regulated and positively correlated with poor prognosis of patients with lung cancer.

Discussion
Lung cancer is the leading cause of cancer-related mortality. An understanding of the mechanisms through which tumours initiate, Fig. 6 DDA1 is overexpressed in lung cancer tissue, and DDA1 expression correlates with shorter survival in lung cancer patients. (A) Analysis of tissue microarrays (TMAs). Immunohistochemical probing for DDA1 using TMAs containing 118 lung cancer, adjacent and normal tissue specimens (scale bar = 50 lm). (B) Statistical analysis was performed using GraphPad Prism 6 software to compare the relative levels of DDA1 among normal, adjacent and lung cancer tissues (nonparametric test, n = 118, P < 0.0001; Dunn's multiple comparison test, **P < 0.01). Black circle and triangles indicate the specimen, and the lines indicate the mean AE S.E.M. (C) Kaplan-Meier curve showing overall survival in patients (percentage) with lung cancer, stratified by DDA1 expression (high and low expression). DDA1 expression was determined by immunohistochemical staining and scored as described in the methods. The log-rank test was used to compare differences between groups (n = 75; long-rank test). progress, metastasize or acquire resistance to targeted therapies is critical for the identification of additional pathways that can be targeted and for predicting which patients will respond to specific therapies to maximize clinical benefits [21]. EGFR, KRAS and EML4-ALK are well-understood oncogenic drivers of lung cancer [22][23][24], and patients with lung tumours that are dependent on these oncogenes are treated with EGFR inhibitors, KRAS pathway inhibitors (antroquinonol or AZD6244) and ALK inhibitors (crizotinib), as well as newer second-generation inhibitors [25,26]. As responses to these agents are generally short-lived, an understanding of other mechanisms leading to lung cancer is essential for the design of therapeutic strategies to improve efficacy. We found that enhanced DDA1 expression is common in human lung cancer. Overexpression of DDA1 was observed in most lung cancer samples and all lung cancer cell lines investigated. Our results also indicated that DDA1 is up-regulated and positively correlated with poor prognosis in patients with lung cancer. DDA1 could thus be a molecular marker of lung cancer.
Loss of growth control is a hallmark of cancer [9]. Cell cycle and cell growth are governed by two distinct levels of regulation. One is the transcriptional control of genes [27], and the other is the degradation of proteins involved in chromosomal DNA replication and segregation [28]. Therefore, cell cycle phases, particularly G1 and G2, are not defined by the content of genomic DNA alone, but also by the status of the above two systems [29]. In the present study, DDA1 was shown to be responsible for lung cancer cell G1/S transition and cell proliferation through the regulation of cyclin D1, cyclin D3, cyclin E1 and cyclin B1 expression. The proliferation of mammalian cells is driven by cyclins, which are key components of the cell cycle machinery that bind to, activate and provide substrate specificity for CDKs, thereby driving cell cycle progression [30]. Cyclin D1 and cyclin D3 are dysregulated in a variety of human cancers, and the inhibition of these proteins represents a highly selective anticancer strategy that specifically targets cancer cells without significantly affecting normal tissues [15,18,[31][32][33]. Cyclin E1 overexpression correlates with poor overall survival in several tumour types including breast, ovarian, lung and pancreatic cancer [34,35]. Attenuation of the PI3K/ PKCiota/cyclin E1 pathway is a target in ovarian cancer [36]. Deregulation of cyclin E1 causes human mammary epithelial cells to enter mitosis with short unreplicated genomic segments at a small number of specific loci, leading to anaphase anomalies and ultimately deletions [37]. The fundamental role of cyclin B1/Cdk1 is in the regulation of cytoplasmic and nuclear events required for G2/M transitions. Cyclin B1/Cdk1 might also function in co-ordinating mitochondrial bioenergetics for cell cycle G2/M progression [38]. Recently, Aregger et al. [39] found that RNMT is phosphorylated and activated by CDK1-cyclin B1, resulting in elevated cap methyltransferase activity at the beginning of the G1 phase. DDA1 could regulate these cyclins leading to manipulation the G1/S transition and cell proliferation.
DDA1, together with DET1 and DDB1, might be involved in ubiquitination and subsequent proteasomal degradation of target proteins [2,4]. Moreover, DDA1 was demonstrated to be a core subunit of multiple CRLs and might regulate CRL4s [5].
CRL4s are involved in the regulation of the DNA damage response, histone modification and nucleosome assembly [40][41][42]. Knockdown of Cul4A results in chromatin dysfunctions in yeast and mammalian cells, and Cul4A is overexpressed in many cancer types [43]. CRL4s are also reported to have important functions in the cell cycle, especially in S-phase-related protein degradation. CRL4s have been shown to target several key factors including Cdt1, p21, p27, E2f1 and Chk1 for degradation, to maintain proper S-phase and S-G2 progression [41,[43][44][45]. DDB1, as the linker protein for the Cul4 E3 ubiquitin ligase, regulates proteins that are essential for nucleotide excision repair, cell cycle progression, DNA replication and cell growth [46]. Future studies are required to elucidate the molecular mechanism through which DDA1 regulates cyclin expression and cell cycle progression. In the study, we found that both protein and mRNA were higher in cancer cell lines than in normal cells, while DDA1 mRNA was not positively associated with protein level and the mechanism was still unclear. The cellular proteome is a complex microcosm of structural and regulatory networks that requires continuous surveillance and modification to meet the dynamic needs of the cell [47]. Genetic alterations, including chromosome alteration, oncogene activation, miRNA dysregulation, mRNA decay, autophagy and ubiquitin-proteasome system [47,48], can affect protein biogenesis and degradation systems, which often results in proteome imbalance. mRNA half-lives differ significantly between various transcripts in all eukaryotic organisms investigated so far [49,50]. Interestingly, decay rates for some mRNAs seem to be conserved between different species to some extent [51]. Notably, mRNAs encoding housekeeping proteins tend to have considerably longer half-lives than those encoding regulatory proteins [52]. Connexin 31.1 was newly reported to be down-regulated in NSCLC cell line through both ubiquitin-proteasome system (UPS) and autophagy [53]. So, further investigations on the regulatory mechanism should be needed in the future.
In summary, in this study, we found that DDA1 is commonly upregulated in lung cancer tissue and cell lines and that higher expression level of DDA1 is associated with poor prognosis in patients with lung cancer. The results in vitro show that the DDA1 enhances proliferation in tumour cells and promotes S-phase entry through the regulation of cell cycle-related proteins. In vivo studies determined that overexpression of DDA1 can promote the tumour growth, while inhibition of DDA1 gene expression can significantly inhibit the tumour growth in subcutaneous xenograft model. DDA1 as an oncogenic factor could be a target of future therapeutics or could be applied as a predictive marker of lung cancer. Further evaluation of DDA1 as a target for lung cancer therapy and the molecular mechanism is required using additional preclinical animal models and other approaches. In addition, the predictive value of DDA1 should be expanded to larger patient cohort studies. Figure S1 Interference effect of shRNA targeting DDA1. H1299 cells was transfected and cultured for 48 hrs followed by western blot analysis of the whole cell lysates Figure S2 (A) A549 cells were transfected and cultured for 2 days followed by PI staining and flow cytometry for cell cycle analysis. The ratio of G0/G1 phase was decreased and the proportion of cells in S and G2/M phases was significantly increased after overexpression of DDA1. (B) H1299 cells were transfected and cultured for 2 days followed by PI staining and flow cytometry for cell cycle. The ratio of G0/G1 phase was increased and proportion of cells in S and G2/M phases was significantly decreased after inhibition of DDA1 Figure S3 (A) A549 cells were transfected and cultured for 24 hrs followed by synchronization to G2/M phase by thymidine and nocodazole. The cells were released from blocking for indicated times and analyzed by PI staining and flow cytometry. The proportion of S-phase cells was significantly increased after 6 hrs (B) H1299 cells were transfected and treated as in (A). Then cells were released from blocking for indicated times and analyzed by PI staining and flow cytometry. The percentage of S-phase cells was decreased significantly after 10 hrs Figure S4 DDA1 is overexpressed in lung cancer tissue. 8 pairs of tumor (T) and normal (N) tissue of lung cancer patients were assessed by western blot and DDA1 level in all these tumor tissues was higher than that of normal tissues Figure S5 Representative IHC score of TMA tissue sections Table S1 shRNA sequence of DDA1