ALAD Expression and Prognostic Value in Multiple Human Cancers: A Bioinformatics Analysis

Background: δ-aminolevulinic acid dehydratase (ALAD) is a kind of metalloenzyme, which can catalyze δ-aminolevulinic acid to synthesize bilirubin. However, the expression level, prognostic role, biological function and mechanism of action of ALAD in a variety of human tumors are still not clear. Methods: GEPIA Online (http://gepia.cancer-pku.cn) website was used for mRNA analysis of ALAD in multiple human solid tumors. NCBI/GEO database and UALCAN website (http://ualcan.path.uab.edu) were used to analyze the ALAD mRNA expression in breast cancer and lung cancer. Immunohistochemical (IHC) analysis from HPA database was used to detect the protein expression of ALAD in different human tumors. Kaplan-Meier plotter database (http://kmplot.com/analy sis/) was used to evaluate the prognostic value of ALAD expression in different human tumors. Gene Ontology (GO) analysis, Kyoto Encyclopedia of Gene and Genome (KEGG) pathway analysis and Gene Set Enrichment Analysis (GSEA) analysis were used for ALAD enrichment analysis and signal pathway analysis. Results: Analyzed by GEPIA, UALCAN online website and GEO analysis in multiple human tumors, the results showed that ALAD mRNA expression is signicantly down-regulated and positively correlated with the overall survival (OS) time of different cancer patients. Besides, the expression of ALAD protein is always none or low in multiple solid tumors. In addition, GO and KEGG enrichment analysis suggested that ALAD may play an anti-tumor role in many directions, and GSEA further found that the overexpression of ALAD is negatively correlated with cell cycle signal pathway and PI3K/AKT/mTOR signal pathway. Conclusions: In conclusion, our study demonstrated for the ﬁ rst time that ALAD is downregulated in multiple human cancers and ALAD low expression is associated with worse overall survival of different cancer patients. Furthermore, we showed that ALAD low expression is associated with tumor cell cycle process and PI3K/AKT/mTOR signaling pathway. ALAD may be a valuable prognostic biomarker and therapeutic ALAD. The results showed that the overall survival time is longer in ALAD high expression group than in ALAD low expression group in BRCA (HR = 0.79; P = 0.029), ESCA (HR = 0.37; P = 0.0058), KIRC (HR = 0.42; P = 5.5e-07), KIRP (HR = 0.42; P = 0.0041), LIHC (HR = 0.51; P = 8.8e-05), LUAD (HR = 0.67; P = 6e-04), LUSC (HR = 0.45; P = 0.029), SARC (HR = 0.56; P = 0.0035), THYM (HR = 0.12; P = 0.016) and UCEC (HR = 0.53;P = 0.003; Fig. 5), and statistical analysis suggested that the p value is less than 0.05 in all the above ten tumors. These results indicate that ALAD may act as a tumor suppressor and may be a valuable biomarker for predicting prognosis of patients with BRCA, ESCA, KIRC, KIRP, LIHC, LUAD, LUSC, SARC, THYM and UCEC. the differentially expressed genes (DEGs) related to ALAD were mainly concentrated in the negative regulation of RNA polymerase II promoter transcription, cell proliferation and protein binding, while KEGG analysis showed that these DEGs were mainly concentrated in cancer pathways and Wnt signaling pathway. This suggests that ALAD has a wide range of biological functions. However, what is the regulation mechanism of ALAD in these tumors? Aiming to solve this problem, we conducted GSEA analysis to screen the signal pathways closely correlated with the regulation of ALAD. The results showed that cell cycle (HALLMARK_G2M_CHECKPOINT) regulation was negatively correlated with ALAD expression in breast cancer and lung cancer, suggesting that ALAD may inhibit the tumor progression by inhibiting the cell cycle process. Next, we used the same method to further study the mechanism of action of ALAD, and the results showed that ALAD expression was negatively correlated with PI3K/AKT/mTOR signaling pathway (HALLMARK_PI3K_AKT_MTOR_ SIGNALING) which can inhibit tumor growth, proliferation, migration and invasion. The combination of the two indicates that ALAD may inhibit tumor cells growth, proliferation, migration and invasion by inhibiting the cell cycle process and the PI3K/AKT/mTOR pathway. Collectively, we believe that ALAD can be used as a molecular target in the diagnosis and treatment of cancers.


Background
With the continuous deepening and development of human genomics, pharmacogenomics and tumor biology, the diagnosis and treatment of cancer has gradually stepped into the era of molecular level. Nowadays, the difference between tumor and normal cells can be found through its special "molecular characteristics" in the early stage of malignant tumor [1], including oncogene, tumor suppressor, gene mutation and chromatin modi cation. Based on this, tumor molecular targeting therapy becomes increasingly important. It takes speci c molecules of tumors as the therapeutic target, and speci c drugs aimed at these speci c molecules are used to kill tumor cells and destroy their growth environment with little or no damage to normal tissues and cells as much as possible. In recent years, great progress has been made in the eld of tumor targeted therapy, such as in the identi cation of new tumor antigens [2,3], in the mechanism studies [4][5][6], and in the optimization of antibody structure [7][8][9]. At present, the Page 3/17 molecular targeted therapy has showed de nite effects in clinical practice: targeted therapy combined with radiotherapy can strongly inhibit the expression of COX-2 and VEGF in bone metastasis of lung cancer [10]; HRG is a new target for ER positive breast cancer [11]; sorafenib, a multi-target kinase inhibitors, is the preferred choice for advanced liver cancer [12]; and imatinib, a receptor tyrosine kinase inhibitor, is effective for chronic myeloid leukemia [13]. The precondition of employing tumor targeted therapy is to accurately evaluate the expression of tumor molecular targets that can be used as therapeutic targets and biological markers of the disease and predict the biological characteristics of the disease.
δ-aminolevulinic acid dehydratase (ALAD), a homomeric metallomeric enzyme that catalyzes 5aminolevulinic acid to produce bilirubin, is located on human chromosome 9q34 [14]. The active site of ALAD has the common characteristics of both metal aldolase and Schiff base aldolase, so ALAD represents an intriguing combination of both classes of enzyme. Generally, ALAD is believed to be closely related to lead. On the one hand, lead ions can replace the zinc bound to the enzyme and effectively inhibit ALAD [15], and besides, lead exposure increases the abnormal methylation of ALAD [16]. On the other hand, the activity of ALAD can be used as a biomarker of environmental lead exposure and toxicity in prokaryotes and eukaryotes [17]. More importantly, it is reported that ALAD has been shown to be associated with the risk of a variety of cancers. For example, common genetic mutations in ALAD may change the overall risk of renal cell carcinoma [18]; genetic variation of ALAD may change the association between lead and prostate cancer [19]; single nucleotide polymorphisms in the ALAD gene affect the association between lead exposure and brain tumor risk [20]. Besides, it has also been reported that ALAD-proteasome complexes may have a clinical impact on improving anticancer therapy depended on HDAC inhibitors [21]. Based on the above reports, it is an attractive research direction to study whether ALAD can be used as a molecular targets in a series of human solid tumors.
In this study, TCGA database and GEPIA analysis showed that ALAD mRNA expression was signi cantly down-regulated in BRCA, ESCA, KIRC, KIRP, LUAD, LUSC, THYM and UCEC cancer tissues. Besides, GEO database and UALCAN website indicated that the mRNA expression of ALAD in BRCA, LUAD and LUSC cancer tissues was signi cantly reduced compared with adjacent normal tissues. In addition, HPA database suggested that ALAD protein was not expressed or low expressed in different tumors. At the same time, KM plotter analysis showed that enhanced ALAD expression was associated with a more favorable outcome of overall survival time in patients with BRCA, ESCA, KIRC, KRIP, LIHC, LUAD, LUSC, SARC, THYM and UCEC. These ndings suggest that ALAD may be a potential biomarker for predicting the prognosis of many human cancers. Further, GO enrichment analysis showed that ALAD had signi cant effects on the negative regulation of transcription from RNA polymerase II promoter and positive regulation of cell proliferation, and KEGG pathway enrichment analysis suggested that ALAD mainly involved in cancer pathways. Mechanistically, GSEA analysis indicated that upregulation of ALAD might inhibit cell cycle and PI3K/AKT/mTOR signaling pathway. Taken together, all the above ndings suggest that ALAD may be a potential biomarker for predicting the prognosis of a variety of human cancers and play a tumor inhibitory role in human breast cancer and lung cancer.

TCGA Dataset Analysis
Gene Expression Pro ling Interactive Analysis (GEPIA) online (http://gepia.cancer-pku.cn/) is a newly developed interactive web server for analyzing the RNA sequencing expression data of 9736 tumors and 8587 normal samples from the TCGA and the GTEx projects [22]. In this study, GEPIA database, being presented in Expression DIY-box plots (in the 'Boxplot' tab) and containing ten dataset selections which are matched normal TCGA data for analysis, was used to analyze expression levels of ALAD and lncRNAs in them. Genes with |log2FC| > 1 and p-value < 0.05 were considered as statistically signi cant. TCGA Study Abbreviations are shown in Table 1.

Gene Expression Omnibus Expression analysis
The Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) is an international public repository for high-throughput microarray and next-generation sequence functional genomic data sets submitted by the research community and all data are freely available for download in a variety of formats [23]. Four independent breast cancer and lung cancer gene expression pro les (GSE5364, GSE134359, GSE33532, and GSE56044) were downloaded from GEO database and these datasets were UALCAN Online Expression analysis UALCAN (http://ualcan.path.uab.edu) is used to construct an algorithm based on the TCGA level 3 RNAseq database and all data are divided into 31 types according to the type of cancer. It can be used to analyze relative transcriptional expression of potential genes of interest between tumor and normal samples and association of the transcriptional expression with relative clinicopathologic parameters [24]. In this study, UALCAN was used to analyze the mRNA expression of ALAD in BRCA, LUAD and LUSC dataset from TCGA, including total expression, sex, age, race, tumor stage, lymph node metastasis and TP53 gene mutation.

Immunohistochemical analysis
The expression of ALAD protein was analyzed by the HPA database (https://www.proteinatlas.org/). ALAD was entered into the database to obtain the ALAD expression level of a single tumor of each cancer type (scar strip = 200 μm). All IHC images have been manually annotated by a certi ed pathologist [25,26].

Prognostic analysis
The Kaplan-Meier plotter (http://kmplot.com/analysis/) is an online tool applied to assess the effect of 54,675 genes on survival using 10,461 cancer samples (5,143 breast, 1,816 ovarian, 2,437 lung, and 1,065 gastric cancer) [27]. Online KM plotter database based on TCGA database is used to analyze the relationship between a speci c gene in tumor and patients' survival time of death. The results are shown by 95% con dence interval and risk-to-risk ratio.

Enrichment analysis
The LinkedOmics database (http://www.linkedomics.org/login.php) was used to look for the differentially expressed genes (DEGs) related to ALAD. We analyzed methylation data by search dataset and target dataset using Spearman test in TCGA-BRCA dataset, TCGA-LUAD dataset and TCGA-LUSC dataset, to work out the correlation coe cient between DEGs and ALAD (Without subset database). Then, the selected genes were analyzed by the Database for Annotation, Visualization, and Integrated Discovery (DAVID, https://david.ncifcrf.gov/home.jsp) for Gene Ontology (GO) analysis and Kyoto Encyclopedia of Gene and Genome (KEGG) pathway enrichment analysis to get the results. Based on the enrichment theory of go and KEGG pathways, we encode each essential / nonessential gene into a vector, where each component represents the relationship between the gene and a go term or KEGG pathway [28]. Eventually, those results were visualized by bioinformatics online tool (http://www.bioinformatics.com.cn).

Correlation analysis
Gene Set Enrichment Analysis (GSEA) is routinely used to analyze and interpret coordinate pathway-level changes in transcriptomics experiments [29]. Besides, GSEA is a computational method that assesses whether an a priori de ned set of genes shows statistically signi cant, concordant differences between two biological states [30]. GSEA was used to analyze the relationship between ALAD expression and cell cycle signal pathway and PI3K/AKT signaling pathway in patients with breast cancer and lung cancer.

Statistical analysis
GraphPad PRISM 5 software (GraphPad Software, San Diego, California, USA) was used for all statistical analyses. Differences between two groups were analyzed using the two-tailed unpaired Student's t-test; P<0.05 was considered statistically signi cant.

ALAD mRNA Expression Is Down-regulated in Ten Solid Tumor Types
To determine the potential role of ALAD in a variety of human cancers, we rst veri ed the expression level of ALAD mRNA using the TCGA database and GEPIA analysis. The results showed that ALAD mRNA expression was signi cantly down-regulated in breast invasive carcinoma (BRCA), esophageal cancer (ESCA), renal clear cell carcinoma (KIRC), renal papillary cell carcinoma (KIRP), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), thymoma (THYM) and endometrial carcinoma (UCEC). In liver hepatocellular carcinoma (LIHC) and sarcoma (SARC), although the results were not statistically signi cant, the average expression level of ALAD mRNA in cancer tissues was lower than that in normal tissues (Fig. 1).
ALAD mRNA Expression Is Low in BRCA, LUAD and LUSC By downloading the microarray results related to breast cancer and lung cancer from the NCBI/GEO public database, we found ALAD mRNA expression was obviously down-regulated in cancer tissues compared to adjacent normal tissues in the GSE5364 (ANT = 13; T = 83), the GES134359 (ANT = 12; T = 74), the GSE33532 (ANT = 20; T = 80) and the GSE56044 (ANT = 12; T = 124) (Fig. 2). Moreover, by analysis in UALCAN website, we found the mRNA expression of ALAD was higher in the normal samples (BRCA group n = 114; LUAD group n = 59; LUSC group n = 52) than in the cancer samples (BRCA group n = 1097; LUAD group n = 515; LUSC group n = 503; Fig. 3A, 3H, 3O). Further, subgroup analysis showed that ALAD was also signi cantly down-regulated in different subgroups of BRCA, LUAD, and LUSC, including gender, age, and race subgroups (Fig. 3B-D, 3I-K, and 3P-R). Regarding to cancer stages, whether in BRAC, LUAD or in LUSC, we found that ALAD mRNA expression in stage 1-4 was yet lower than that in normal tissues (Fig. 3E, 3L, and 3S). In view of lymph node metastasis -N staging, ALAD mRNA expression was also signi cantly lower in cancer tissues with N0-N3 stages than that in normal tissues (Fig. 3F, 3M, and 3T). In addition, ALAD mRNA expression was signi cantly decreased in BRCA and LUAD tumors with TP53 mutation than that without TP53 mutation ( Fig. 3G and 3N). Through all the above analysis, we suggest that ALAD may play a tumor inhibitory role in the progression of breast cancer and lung cancer.

Alad Enrichment Analysis Of Brca, Luad And Lusc
From LinkedOmics website and based on the Spearman test, we screened the differentially expressed genes (DEGs) changed with ALAD expression level in BRCA, LUAD and LUSC (Fig. 6A, 6D, 6G). The rst 50 positively related genes and the rst 50 negatively related genes in different tumors were shown in the form of heat map (Fig. 6B-C, 6E-F, 6H-I). Moreover, 561 DEGs were screened by Spearman test more than 0.2 in the BRCA dataset; 1401 DEGs were screened by Spearman test more than 0.4 in LUAD dataset; 1910 DEGs were screened by Spearman test more than 0.4 in LUSC dataset. Among all the above DEGs, 542 DEGs overlap each other (Fig. 7A). GO and KEGG enrichment analysis on the DAVID website obtains the relevant data about Biological Process (BP), Cell Composition (CC), Molecular Function (MF) and KEGG pathway. The results are visualized by bioinformatics online tools. For the BP, the negative regulation of transcription from RNA polymerase II promoter, positive regulation of cell proliferation and cell proliferation were the top three relevant pathways (Fig. 7B). For the CC, the potential target genes were predominantly enriched in the cytoplasm, cytoso and membrane (Fig. 7C). For the MF, the mainly signi cantly involved items were protein binding (Fig. 7D). Furthermore, KEGG analysis showed that cancer pathways, proteoglycans in cancer, Wnt signaling pathways and pathways regulating pluripotency of stem cells were the main pathways related to ALAD regulation (Fig. 7E). Collectively, the above results show that ALAD may be involved in many regulation processes in tumors and may be a hopeful therapy target. More studies are needed to further investigate its functions and regulation mechanism.

ALAD Expression Is Negatively Correlated with Cell Cycle and PI3K/AKT/mTOR Signaling Pathway
To more deeply explore the regulation mechanism of ALAD, we conducted GSEA analysis and the results revealed that ALAD expression negatively correlated with the cell cycle characteristics (HALLMARK_G2M_CHECKPOINT) in the publicly available GEO breast cancer database (NCBI/GEO/GES5364, n = 341) and lung cancer database (NCBI/GEO/GSE33532, n = 100) (Fig. 8A). To understand the mechanism of ALAD-mediated cell cycle in breast cancer and lung cancer, we used the same GEO databases to study the relationship between ALAD expression levels and the cancer signaling pathways which closely related to cell cycle regulation, and we found that ALAD expression was negatively correlated with the PI3K/AKT/mTOR pathway (HALLMARK_PI3K_AKT_MTOR_SIGNALING) (Fig. 8B). The above results suggest that ALAD may inhibit the cell cycle process and the PI3K/AKT/mTOR signaling pathway and consequently suppress the cancer cells growth and proliferation. Discussion ALAD, involved in heme biosynthesis process, is an endogenous inhibitor of 26S proteasome and a target of cancer therapy [31,32]. It has been reported that the gene level and protein level of ALAD are low in breast cancer cell lines, and ALAD plays a tumor inhibitory role in the progression of breast cancer [33]. In addition, Ge et al. point out that ALAD can inhibit the EMT process induced by TGF-β and reduce the invasive ability of cancer cells, suggesting that ALAD may be a potential tumor suppressor.
In this study, we selected ten human solid tumors in TCGA database to compare the expression of ALAD mRNA in cancer tissues and normal tissues. The results showed that ALAD mRNA was signi cantly lower in BRCA, ESCA, KIRC, KIRP, LUAD, LUSC, THYM and UCEC compared with normal tissues, suggesting that ALAD mRNA is under-expressed in most solid tumors. Secondly, to explore the anti-function roles of ALAD, we selected the breast cancer and lung cancer for further data mining. The results from NCBI/GEO public database showed that the cancer groups had a signi cantly lower level of ALAD expression than the control groups in GSE5364, GSE134359, GSE33532 and GSE56044 datasets. Besides, the UALCAN database showed that ALAD mRNA expression levels inversely correlated with sex, age, race, tumor stage, regional lymph node metastasis and TP53 mutation. IHC results from HPA database showed that ALAD protein was not expressed or under-expressed in most human solid tumors. In addition, the patients with high ALAD had a longer overall survival time than the ones with low ALAD, and the p values of the ten Kaplan-Meier survival curves were all less than 0.05, suggesting that high expression of ALAD indicates a better prognosis. Taken together, these results support that ALAD may be a tumor suppressor gene in human solid tumors and an important biomarker for predicting the survival of cancer patients.
To more systematically explain the biological functions and molecular mechanism of ALAD and nd signal pathways involved in the regulation of ALAD, we conducted enrichment analysis. GO analysis showed that the differentially expressed genes (DEGs) related to ALAD were mainly concentrated in the negative regulation of RNA polymerase II promoter transcription, cell proliferation and protein binding, while KEGG analysis showed that these DEGs were mainly concentrated in cancer pathways and Wnt signaling pathway. This suggests that ALAD has a wide range of biological functions. However, what is the regulation mechanism of ALAD in these tumors? Aiming to solve this problem, we conducted GSEA analysis to screen the signal pathways closely correlated with the regulation of ALAD. The results showed that cell cycle (HALLMARK_G2M_CHECKPOINT) regulation was negatively correlated with ALAD expression in breast cancer and lung cancer, suggesting that ALAD may inhibit the tumor progression by inhibiting the cell cycle process. Next, we used the same method to further study the mechanism of action of ALAD, and the results showed that ALAD expression was negatively correlated with PI3K/AKT/mTOR signaling pathway (HALLMARK_PI3K_AKT_MTOR_ SIGNALING) which can inhibit tumor growth, proliferation, migration and invasion. The combination of the two indicates that ALAD may inhibit tumor cells growth, proliferation, migration and invasion by inhibiting the cell cycle process and the PI3K/AKT/mTOR pathway. Collectively, we believe that ALAD can be used as a molecular target in the diagnosis and treatment of cancers.

Conclusion
After in-depth data mining of ALAD, we found that the expression of ALAD in cancer tissues was downregulated compared with normal tissues, and ALAD high expression is associated with better overall survival of cancer patients. The mechanism may be that ALAD can regulate the cell cycle process and PI3K/AKT/mTOR signal pathway. In conclusion, ALAD may be a valuable prognosis biomarker and hopeful therapeutic target for multiple human tumors.