The kynurenine pathway (KP) is a major route for the metabolism of an essential amino acid, tryptophan which results in the production of biologically active molecules, kynurenines. Most of the intermediates of the kynurenine pathway are neuroactive and known to play roles in the regulation of N-methyl-D-aspartate (NMDA) receptor function and free radical production. A central compound of the pathway is Kynurenine (KYN) which can be metabolized in two separate ways: to Kynurenic acid (KYNA), or to 3-hydroxykynurenine (3-OH-KYN) and quinolinic acid (QUIN), the precursors of NAD. Most of the importance has been given to KYNA due to its broad spectrum antagonistic properties (Ganong and Cotman, 1986; Stone and Connick, 1985)
Epilepsy is characterized by recurrent spontaneous seizures due to hyperexcitability and hypersynchrony of brain neurons. Although there are limited clinical evidence to support the deregulation of kynurenine pathway in epilepsy, but the understanding of the proinflammatory cytokine signaling regulated kynurenine pathway and neuroinflammation in the recurrence of epileptic seizure activity may strengthen this possibility. Further based on the pre-clinical data (Gleeson et al., 2010; Lehrmann et al., 2008), it may be predicted that in some forms of epilepsy, over activation of microglial branch with respect to the astrocytic branch of the KP leads to accumulation of QUIN and 3-HK in the CNS. It has been proposed that adjunctive treatment with KMO inhibitor along with anti-convulsants can improve the treatment outcome, as inhibition of KMO increase the KYNA production and decrease the 3-HK and QUIN production in the CNS (Campbell et al., 2014).
More than 20 years ago various neurophysiological experiments revealed the neuroinhibitory properties of KYNA. (Perkins and Stone, 1982) discovered that KYNA is a broad spectrum antagonists of glutamate receptor which can inhibit three types of ionotropic receptors- N-methyl-D-aspartate (NMDA), kainate and a-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptors. KYNA act as a competitive antagonist at the glycine site on NMDA receptors (Kessler et al., 1989). KYNA also block the ?7- nicotinic acetylcholine receptor and this inhibition is noncompetitive in nature (Hilmas et al., 2001). In addition to this in past few years some other receptor targets have been identified. For example, KYNA is also reported to interact with GPR35 (Wang et al., 2006) and arylhydrocarbon receptors (DiNatale et al., 2010). However, further studies are required to elaborate their role in neuronal disorders.
Various studies have shown that KYNA alters the progression of kindled seizures. (Thompson et al., 1988) have shown that pretreatment with administration of intracerebroventricular KYNA prior to administration of electrical kindling stimulus in rats significantly reduced the rate of kindling. A study by Szyndler and co-workers demonstrated that during the electrically induced kindling of seizures, the seizures development is associated with significant alterations in the levels of KYNA and its precursor, tryptophan (TRP). The levels of KYNA and ratio of KYNA/TRP (theoretical index of kynurenine pathway activity) was found to be increased in the amygdale and hippocampus of kindled animals whereas concentration of tryptophan in the prefrontal cortex and hippocampus was found to be decreased (Szyndler et al., 2012). Carpenedo et al. 1994 demonstrated that the administration of inhibitors of kynurenine hydroxylase and kynureninase, evoked a significant increase in brain levels of KYNA which protected the rats from electroshock induced seizures or DBA/2 mice from audiogenic seizures (Carpenedo et al., 1994).
An enhanced liberation of KYNA was reported in the hippocampus of electrically kindled animals and in the extracellular fluid (ECF) elevation of KYNA levels was not found to be associated with electrical stimulation per se but with the progression of kindling development to the stage of generalized tonic or clonic convulsions (Wu et al., 1995). Whereas in another study, it has been found that in nucleus accumbens kindling induced a lasting increase in kynurenate although no significant changes were seen in hippocampus, striatum, cerebral cortex, olfactory bulb, thalamus, tectum, pons/medulla, cerebellum, or plasma (Loscher et al., 1996). Nemeth et al (2004) found that PTZ induced seizures was inhibited by KYNC, and it also protected the animals from the death induced by repeated PTZ injections (Nemeth et al., 2004). Maciejak et al. (2009) also reported a significant decrease in tissue concentrations of KYNA in some brain structures (the caudate entorhinal cortex, piriform cortex, putamen, amygdala and hippocampus) after PTZ-induced kindling (Maciejak et al., 2009). Scharfman et al ( 1999) have demonstrated that KYNA converted in situ from KYN was potent enough to block the epileptiform discharges induced in Mg2 -free medium and KYN applied at 200 uM prevents spontaneous activity in the hippocampal CA3 region (Scharfman et al., 1999). However, a study carried out by Rozsa et al., 2008 on hippocampal slices from young animals showed that KYN pretreatment was effective on neural hyperecxitability even at low concentration of 16uM. This study also found that the KYNA produced by conversion of KYN?KYNA was sufficient to prevent the neuroexcitatory effect of PTZ. This study thus supports the hypothesis that the various neuronal disorders which are affected by neuronal hyperexcitation, the kynurenine pathway might be a valuable drug target for these disorders (Rozsa et al., 2008).
Szyndler et al 2012 have shown an increase in the glutamate/GABA ratio and a decrease in GABA levels in the amygdale of kindled animals. Further in the amygdala of kindled rats, between the levels of KYNA and glutamate, negative correlation was found (Szyndler et al., 2012). This finding suggested that the disrupted balance between the excitatory and primary inhibitory brain systems may change the local KYNA levels as an adaptive reaction. Further onset of seizures can modify the function of the GABA system, and in some studies in response to kindling the stimulation as well as the inhibition of local GABAergic activity has been found (Kamphuis et al., 1990; Kaura et al., 1995). Reduction in the extracellular concentration of GABA has been found in PTZ-induced kindling in the rat hippocampus (Szyndler et al., 2008). These studies indicated that the imbalance between KYNA synaptic transmission and glutamatergic, GABAergic characterizes hyperexcitable, epileptic and local neuronal circuits.
Mechanisms responsible for the increase level of KYNA were not clearly understood. During the kindling of seizers, few studies have indicated the association of hypertrophied astrocystes in brain regions with increase activity of KAT which is main enzyme for synthesis of KYNA synthesis (Wu et al., 1995). Thus, the astrocytic changes and process of kindling was found to be accompanied by KYNA elevation (Du et al., 1993). In contrary to this no significant changes were found in KAT activity in the amygdala of electrically kindled animals (Wu et al., 1995). The lasting elevation of KYNA levels appeared to be different in fully kindled animals from the increase which occurs immediately following generalized convulsions. Wu and Schwarcz 1996 have suggested that it may occur due to desensitization of presynaptic glial amino acid receptors which regulate local KYNA release, in response to repeated release of glutamate (Wu and Schwarcz, 1996).
Kynurenine, 3-hydroxykynurenine and other metabolites of kynurenine pathway
Some other metamobolities of kynurenine pathway metabolites, including 3-hydroxyanthranilic acid (3-HANA), kynurenine, 3-hydroxykynurenine (3-HK), and anthranilic acid , have failed to show direct effects on neuronal activity (Stone, 1993). Recently kynurenine has been described as an endogenous ligand of the human aryl hydrocarbon receptor (Opitz et al., 2011), and the activation of metabotropic glutamate receptors has been shown by xanthurenic (Copeland et al., 2013; Schwarcz et al., 2012). 3-hydroxyanthranilic acid has been described as neurotoxin and known to play a role in immunoregulation (Chen and Guillemin, 2009; Lopez et al., 2008). In the brain various kynurenine pathway metabolites may participate in pro- and anti-oxidative processes(Giles et al., 2003). Goldstein et al ,2000 have shown that 3HK and 3HAA produce hydrogen peroxide and superoxide in a copper-dependent manner(Goldstein et al., 2000). This study has suggested that in the oxidative damage of proteins (such as alpha-crystallin), both these metabolites may act as cofactors by interacting with redox-active metals. Thus these findings may have important implications in the understanding of cataractogenesis and other degenerative conditions, where kynurenine pathway is activated (Goldstein et al., 2000).. However, these metabolites also have antioxidant properties, scavenging peroxyl radicals more effectively than equimolar concentrations of ascorbic acid or Trolox analogue of vitamin E) (Christen et al., 1990). Further Like quinolinic acid, intracerebral injection of 3-hydroxyanthranilic acid in rats also leads to a decrease in activity of choline acetyltransferase, but QUIN were found as more potent (Jhamandas et al., 1990).
The first evidence that matabolites of kynurenine pathway can influence the brain function was provided by Lapin (Lapin, 1978), who after an intracerebroventricular QUIN injection observed convulsions in mice. Quinolinic acid is a selective agonist at the neuronal NMDA subtype of glutamate receptors (Stone and Perkins, 1981). Inspite of its low cerebral content and low receptor affinity, its potency as an excitotoxin is due to its selective interaction with NMDAR2 receptor subunit (de Carvalho et al., 1996) . For example QUIN has 10-fold higher affinity for NR2B subunit of the NMDA receptor, which predominates in the forebrain as compared to hindbrain specific NR2C subunit (Schwarcz et al., 2012). Several mechanisms have been found to be associated with neurotoxic properties of QUIN. In addition to agonist of NMDA receptor, it has been found to induce the lipid peroxidation (Rios and Santamaria, 1991), and produce reactive oxygen species (ROS) (Rodriguez-Martinez et al., 2000) (Rodriguez-Martinez et al, 2000; Santamaria et al., 2001) which may accounts for its neurotoxic properties. Its role in lipid peroxidation was found to be modulatd by its intetraction with Fe2 ions to form Quin Fe2 (Stipek et al., 1997). The production of ROS by QUIN was found as secondary to the generation and auto-oxidation of Fe2 – QUIN complexes and it can be inhibited by iron chelation (Platenik et al., 2001). Binding of QUIN to the NMDA receptor is also controlled by endogenous iron (St’astny et al., 1999).
All the studies discussed above suggest that alterations in levels of certain metabolites of KYN pathway (mostly KYNA, QUIN and 3-OH-KYN), either alone or in combination affect the pathology of some brain disorders. In most of the neurodegenerative disorders, these changes were appeared as secondary to the basic pathological process, and in some conditions might be the mediators of CNS pathology (Schwarcz and Pellicciari, 2002). Astrocytes now known to lack KYN-3- hydroxylase, thus favor the synthesis of KYNA, on the other hand microglial cells consist little KYN-aminotransferase (KAT) and thus likely generate the intermediate metabolites of the QUIN branch of the KA pathway (Guillemin et al., 2000). It has been found that astrocytes when present alone act as neuroprotectant by elevating the synthesis of KYNA and diminishing the QUIN production, whereas it become neurotoxic in the presence of macrophages and/or microglia, by the production of large amount of Kynurenine which can be metabolized to form the neurotoxin QUIN by neighboring or infiltrating monocytic cells (Guillemin et al., 2001).
Combine role of cytokine and kynurenine in Epilepsy
Lehrmann et al have shown that inoculation of hamster neurotrophic measles virus in mice increases the microglial activation and brain levels of QUIN and 3-HK. These changes produced the subclinical seizure activity, behavioral seizures and neurodegeneration (Lehrmann et al., 2008). It has been found that administration of kainic acid in rat, induced an inflammatory response in the hippocampus which was characterized by activation of microglia (elevated expression of CD11b), increased expression of pro-inflammatory cytokines such as IFN- ? and IL-1 ? and induction of cytokine-inducible enzymes IDO, iNOS and KMO (Gleeson et al., 2010). Further proinflammatory cytokines IL-1?, TNF-?, and IFN-? have been found as inducers of IDO and IFN-? also induced the expression of KMO (Mandi and Vecsei, 2012). Both the enzymes IDO and KMO, increased the production of 3-HK and QUIN. Role of QUIN as neurotoxin or excitotoxin has been described in various studies. Thus it appears that there is close interaction between the cytokine, kynurenine pathway and nervous system. Further studies are required to examine the interaction between the kynureniney and cytokine signaling pathway in case of epilepsy.
Therapeutic applications of kynurenines
Based on kynurenine pathway modulation, different approaches has been suggested for the development of therapeutic agents. One approach is to use kynurenic acid analogues as antagonists at glutamate receptors. Other is to inhibit the enzymes activities which are responsible for quinolinic acid synthesis. For example inhibition of enzyme kynurenine hydroxylase leads to decrease endogenous quinolinic acid levels and an increase kynurenic acid levels (Stone, 2001). It has been proposed that by maintaining the balance between these metabolites, i.e towards the neuroprotectant and away from the excitotoxin could have neuroprotective and anticonvulsant properties in stroke (Pellicciari et al., 1994; Varasi, 1996).
Antagonists of kunurenic acid
Several studies have used the kynurenate structure to target the glycine-2 receptor site as a prelude for the development of therapeutic agents (Manallack, 1990; Stone, 2001). Various authors have examined the diffrenet portions of kynurenat to explore the different approaches of their modifications(Bigge, 1993; Carling et al., 1993; Leeson et al., 1992; Leeson et al., 1993). These modifications have been reviewed by Stone 2001 (Stone, 2001). For example the halogen atoms substitution, yielded the potent analogue 5,7-dichlorokynurenic (Baron et al., 1990). This formula has been retained in many of the analogues. Compound 4-carboxymethylamino-5,7-dichloroquinoline-2-carboxylicacid(MDL 100,748) and 3-(4,6-dichloro-2-carboxyindol-3-yl)-propionic acid (MDL 29,951) was found as potentanticonvulsantsafter their i.c.v. administration to audiogenic seizure-susceptible DBA/2J mice (Baron et al., 1992; Harrison et al., 1990). A kynurenate analogues with a 3-phenyl substituent resulted in lipid soluble compounds which retained potent activity at the glycine-2 site(McQuaid et al., 1992). For example MDL 104,653 (3-phenyl-4-hydroxy-7-chloro-quinolin-2(1 H)-one), has been found as protective against sound-induced clonic seizuresin DBA/2 mice following intracerebroventricular, intraperitoneal or oral administration. In rats, fully amygdala-kindled motor seizures were found to be significantly reduced and the duration of the after-discharge was significantly shortened after the i.p. administration (Chapman et al., 1995). The enlargement of nitrogenous ring of KYNC into a 7-membered ring has produced benzazepinedione compounds which were found to reduce seizures in DBA/2 mice(Jackson, 1995 ).
Prodrugs of KYNC
Moore et al., 1993 have used agents which act as prodrugs to deliver kynurenic acid into the brain. For example, L-4-chlorokynurenine and 4,6-dichlorokynurenine when transported into brain, converted into 7-chlorokynurenic acid and 5,7-dichlorokynurenic acid respectively (Hokari et al., 1996; Moore, 1993). Different esters have been developed by linking 7-chlorokynurenic acid to D-galactose or D-glucose (Bonina et al., 2000). One such example is 7-chlorokynurenic acid-glucopyranos-3ylester which has been found to be rapidly metabolized in the brain into 7-chlorokunurenic acid and suppressed the NMDA induced seizures (Battaglia et al., 2000).
Modulators of kynurenic acid concentrations
The kynurenine 3-hydroxylase and kynureninase, the degradative enzymes of L-KYN are recognized as important targets for kynurenergic drug development. It has been proposed that hampering with these enzymes, which act at a branching point of the KP can selectively alter the QUIN/KYNA ratio and thus repair the chemical impairments in the brain. Connick et al., 1992; Russi et al., 1992 have shown that administration of Nicotinylalanine with L-kynurenine and probenecid increase the level of kynurenic acid in brain and prevent the induction of seizures (Connick et al., 1992; Russi et al., 1992). Miranda et al. (1997, 1999) have shown the protective potential of nicotinylalanine for the nigrostriatal neurons against damage caused by the local injection of NMDA or quinolinic acid (Miranda et al., 1997; Miranda et al., 1999). Cerebral kynurenic acid levels were not found to be increased more than 3.3-fold. Scharfman and Ofer, 1997 have found that epileptiform bursting activity generated by the prefusion of brain slices with the precursor L-kynurenine was inhibited by the small increase of kynurenic acid (Scharfman and Ofer, 1997) .
Meta-nitrobenzoylalanine and the related compound ortho-methoxybenzoylalanine were found as potent inhibitors of kynurenine-3-hydroxylase and kynureninase respectively (Natalini and Moroni, 1995; Pellicciari et al., 1994). Both these compounds were able to increase the levels of kynurenate in hippocampal extracellular spaces. This l effect was found to be associated with a protection from audiogenic convulsions in DBA/2 mice and decrease in locomotion in rats (Chiarugi et al., 1995). The 3,4- dichlorobenzoylalanine, also a kynurenine-3-hydroxylase inhibitor was found as more effective than meta-nitrobenzoylalanine (Speciale et al., 1996). S-aryl-L-cysteine S,S-dioxides have been found as kynureninase inhibitors (Dua, 1993). S-(2- minophenyl)-L-cysteine-S,S-dioxide has been found as particularly strong inhibitor. The 5-methyl derivative was 3 times more efficient against human kynureninase and found to reduce the stimulation of quinolinic acid synthesis induced by interferon-? in human macrophages (Drysdale and Reinhard, 1998). The 4-aryl-2-hydroxy-4-oxobut-2- enoic acids and esters were found as the most potent kynurenine-3-hydroxylase inhibitors with nanmolar potencies (Drysdale et al., 2000).
Modulators of quinolinic acid concentrations
Another approach to prevent the synthesis of quinolinic acid is to inhibit 3-hydroxyanthranilic acid oxygenase. Compound 4-halo-3-hydroxyanthranilic acids have been found as inhibitor of this enzyme and thus lead to reduction in quinolinic acid formation (Walsh et al., 1991; Walsh et al., 1994). This compound, as well as norharmane and 6-chlorotryptophan were found to be effective in several cell lines including peripheral monocytes and attenuate quinolinic acid synthesis (Saito et al., 1994). Another compound 4,6-dibromo-3-hydroxyanthranilic acid (NCR-631) has been found to inhibit 3-hydroxyanthranilic acid oxygenase and minimize the loss of hippocampal cells produced by anoxia, bacterial lipopolysaccharide or injurious cytokines such as interleukin- 1 ? (Luthman et al., 1998).
Significant DEGs in Bladder Cancer
Differences analysis of gene expression in bladder cancer with microarrays
Totally 619 common DEGs were screened by using 2 mocroarrays
PGR, MAFG, CDC6 and MCMs may play key roles in BC
The core histones were also considered to have major functions in BC.
Purpose: We aim to identify significant differentially expressed genes (DEGs) and analyze the modification of gene expression in bladder cancer (BC) by using bioinformatics analysis.
Methods: GSE24152 and GSE42089 microarray datasets were downloaded from the Gene Expression Omnibus database for further analysis. GSE24152 gene expression data included 17 samples (10 tumor cells from bladder and 7 normal tissues from bladder) while GSE42089 included 18 samples (10 tissues from urothelial cell carcinoma and 8 tissues from normal bladder). Differentially expressed genes (DEGs) were screened, followed by gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) analysis. Furthermore, protein-protein interaction (PPI) network and sub-networks were constructed for the identification of key genes and main pathways.
Results: Totally 1325 DEGs in GSE24152 and 647 DEGs in GSE42089 were screened, in which 619 common DEGs were identified. The DEGs were mainly enriched in pathways and GO terms associated with mitotic and chromosome assembly, such as nucleosome assembly, spindle checkpoint and DNA replication. In the interaction network, HNF4, PGR, MAFG and CDC6 were identified as key genes in BC. Besides, the histones were also considered to be significant factors in BC.
Conclusion: the DEGs, including HNF4, PGR, MAFG and CDC6, and core histones family were closely related to the development of bladder cancer via pathways associated with mitotic and chromosome assembly.
Key words: bladder cancer; differentially expressed genes; interaction network; clustering analysis.
Bladder cancer (BC) is a heterogeneous disease with a variable disease history and at present is the ninth most common tumor world wide . In 2009, 70980 new cancer cases of the urinary bladder were diagnosed in the United States and 14330 patients died from bladder cancer . The most common type of BC recapitulates the normal histology of the urothelium cell carcinoma and the 5-year survival rate in America is approximately 77% . Depending on the depth of of invasion, BC can be classified as 5 forms, including papillary (pTa), lamina propria invasion (pT1), muscle invasion (pT2), invasion to peri-vesical fat (pT3), and locally advanced (pT4) . Surgery is the standard therapy and the use of radiotherapy is considered as an alternative, especially in less fit patients .
In recent years, numerous researches have identified risk factors or related genes for the development of BC  . Shen et al have analyzed the differentially expressed genes and interacting pathways in BC by bioinformatics analysis and genes such as activator protein 1, nuclear factor of activated T-cells were identified to be significant in BC. Zhou et al analyzed the gene expression in human BC samples by using microarray GSE42089 and a set of genes leading to mitotic spindle checkpoint dysfunction were identified to be key genes in BC .
In the present study, significant differentially expressed genes (DEGs) in bladder tumor cells were identified and two microarray profiles, GSE24152 and GSE42089, were used for the screening of significant differentially expressed genes (DEGs), followed by gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) analysis. Furthermore, protein-protein interaction (PPI) network and sub-networks were constructed for the identification of key genes and main pathways. By using the bioinformatics methods above, we aim to identify significant differentially expressed genes (DEGs) and analyze the modification of gene expression in bladder cancer (BC) by using bioinformatics analysis.
Materials and methods
Two gene expression profiles, GSE24152 and GSE42089 were downloaded from Gene Expression Omnibus (GEO) database in National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov/geo/) based on the platform of GPL6791 and GPL9828 in Affymetrix GeneChip Human Genome U133 Plus 2.0 Array, respectively. The microarray GSE24152 was based on 17 samples including10 tumor cells from bladder and 7 normal tissues from bladder while the microarray GSE42089 was based on 18 samples including 10 tissues from urothelial cell carcinoma and 8 tissues from normal bladder.
Data preprocessing and DEGs analysis
Robust multiple average (RMA) algorithm in affy package  was used for the normalization of microarray data and boxplots were generated. The microarray data were divided into two groups, a bladder carcinoma set and a normal set. By using Limma package , the probe-level data of two sets were converted into expression measures and two groups of DEGs were obtained from two microarrays. Venn diagram was generated using VennDiagram package  to screen common DEGs for the further analysis. A combination of FDR 0.5 was used as the threshold. Heat maps were generated by heatmap.2 function in ggplot 2  to display the relative expression differences of DEGs. What`s more, cor.test function was used for evaluating the changing trends of two DEGs groups.
Gene ontology (GO) and pathway enrichment analysis
GO analysis has become a widely used approach for the studies of large-scale genomic or transcriptomic data in function . Kyoto encyclopedia of genes and genomes (KEGG) is a widely used collection of online database which deals with genomes, enzymatic pathways, and biological chemicals.  In this study, the functions and pathways of the screened DEGs were analyzed using the DAVID  from the GO and KEGG pathway database with the p-value < 0.01, respectively.
Interaction network and sub-network construction
Cytoscape  is an open source bioinformatics software platform, which is used for the visualization of molecular interaction networks and integrating with gene expression profiles and other state data. In this study, Biosgenet in Cytoscape  was used to predict and visualize the interactions of selected DEGs and proteins assistant with BIND database  with p-value < 0.05. Besides, sub-networks were constructed and clustering analysis was performed on DEGs using ClusterOne in cytoscape  with p-value < 0.01.
As shown in Figure 1, the obscuring variations in raw expression data were normalized after preprocessed. A total of 1325 genes were differentially expressed in GSE24152 and 637 genes were differentially expressed in GSE42089. The volcano plots of both two microarrays were shown in Figure 2. Totally 619 common DEGs were identified using Venn diagram which was shown in Figure 3, including 313 up-regulated genes and 306 down-regulated genes. The heat maps of the 619 common DEGs expression in GSE24152 and GSE42089 were shown in Figure 3, in which significant expression levels were observed. Person r value was 0.998.
GO and KEGG enrichment analysis of DEGs
GO and KEGG enrichment were performed with p-value < 0.01 for the functional and pathway analysis of DEGs and totally 74 GO terms and 4 KEGG pathways were obtained. The main GO terms and KEGG pathways of 619 common DEGs were listed in Table 1. We can see from the results that the DEGs were mainly enriched in the pathways related to chromosome and cell cycle.
Interaction network and sub-network construction
The network on all DEGs was constructed (Figure 3) and thereby clustering analysis was performed on the network. Totally 6 sub-networks were obtained, which were shown in Figure 4 and Table 2. Progesterone receptor (PGR) was a key gene in cluster 1 and the proteins enriched in cluster 1 were all histone proteins. In cluster 2, v-maf avian musculoaponeurotic fibrosarcoma oncogene homolog G (MAFG) was identified as a key gene, which was also detected in cluster 5. Cell division cycle 6 (CDC6) was a key gene in both cluster 3 and cluster 4. minichromosome maintenance complex component (MCM) family members, including MCM2, MCM4, MCM7 and MCM10, were also key genes in cluster 3 and cluster 4. Nucleosome assembly and sequence-specific DNA binding were significant GO terms of sub-networks cluster 1 and cluster 2, respectively. No GO terms were obtained in clusters 3-6.
Bladder cancer is a common malignancy, which requires a high degree of surveillance because of the frequency recurrences and poor clinical outcome of invasive disease . Bioinformatics analysis on the gene level of bladder cancer cells provides a new insight for the research of this disease. In the present study, by using the expression profile microarray GSE24152 and GSE42089, the significant DEGs in BC were identified. In the interaction network, PGR, MAFG, CDC6 and MCMs were identified as key genes in BC. Besides, the histones were also considered to have major functions in BC. According to the GO term analysis and pathways enrichment, the main GO terms or pathways were associated with cell cycle and chromosome assembly, such as nucleosome assembly, spindle checkpoint and DNA replication.
PGR encodes a member of the steroid receptor superfamily, which mediates the physiological effects of progesterone . In the present study, PGR was an important gene in cluster, which was regulated by a set of transcriptions, revealing the significance of PGR in BC. There is a fact that men are more frequently affected than women, which indicates the hormone as a regulation factor  and Miyamoto et al clarified the sex hormone androgen receptor (AR), were involved in BC . PGR may also has functions in BC, as bothAR and PGR are determining sex hormone in gonadal . A significant pathway detected was progesterone-mediated oocyte maturation, which may be the pathway PGR functions in BC.
Histones are the main structural proteins that are associated with DNA in eukaryotic cells and can be divided into two groups including core histones and the nucleosomal histone . Core histones are some of the most conserved proteins in eukaryotes play key roles in organizing DNA folding . The altered patterns of modifications on histone in various human cancers have been studied in recent years . Schneider et al show that global histone modification levels are lower in BC than normal urinary tissue . Besides, the conserved histone H2A variant has been reported to be over-expressed in BC cells and contributes to cancer-related transcription pathways . In the present study, a set of core histones were clustered in cluster 1 and enriched in the process of nucleosome assembly. Considering the function of histones in the mitotic, we concluded that the nucleosome and chromatin assembly were modified in BC.
MAF encodes for the nuclear transcriptional regulating proteins, which are characterized by a basic region and leucine zipper structure have crucial roles in a variety of cellular processes . MAFG is a small MAF protein member of the family and encode slightly more than the DNA binding and dimerization motif . MAFG is able to partially co-localize with FBJ murine osteosarcoma viral oncogene homolog (FOS) in the nucleus and form heterodimers with FOS . FOS is a member that the transcription factor activator protein 1 (AP-1) is composed of. AP-1 family members are immediate early genes induced by a variety of stress signals and control the stress response including cell proliferation, apoptosis and tumoregenesis . According to our data, the expression of MAFG in BC was up-regulated, which may increase the tumorigenesis via the process mentioned above.
CDC6 is an essential regulator of DNA replication in eukaryotic cells with the function of assembly of prereplicative complexes at origins of replication during the G1 phase of the cell division cycle . MCMs encode highly conserved proteins that presumed to act as an enzymatically active helicase . MCMs drive the formation of prereplicative complexes (PRCs), which is the first key event during the G1 phase during cell cycle . Both MCM and CDC6 are key proteins in the mechanism of DNA replication licensing and have related functions during the cell cycle . CDC6 is responsible for the loading of MCM proteins onto origins of replication and in the absence of CDC6, MCM could not associated with the chromatin. The increased expression of CDC6 and MCM has been seen in dysplastic cells and as a consequence, CDC6 and MCM are considered as specific biomarkers of proliferating cells . Recent studies have unveiled the proto-oncogenic activity of CDC6, with the overexpression interfering with the expression of some tumor suppressor genes and may promote the DNA hyperreplication and induce a senescence response similar to that caused by oncogene activation. Besides, some members of MCM family, such as MCM7, were also suggested to be overexpressed and amplified in a variety of human malignancies . In the present study, the expression of CDC6 was detected to be up-regulated, revealing that in BC cells, the process of DNA replication was aberrant,
In conclusion, our data reveals that the genes, PGR, MAFG, CDC6 and MCMs, and a set of histones are important factors in BC and play key roles in the processes related to mitotic, such as nucleosome assembly, spindle checkpoint and DNA replication. However, the small size of the microarray sample and no experimental variation was the limitation of this study. Thus further study on larger size sample and experiment should be performed to confirm our conclusion.