Physiol. Genomics Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Genomics 24: 144-153, 2006. First published November 29, 2005; doi:10.1152/physiolgenomics.00258.2005 Free Article
1094-8341/06 $8.00
This Article
Free upon publication Free Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figure and Tables
Right arrowFree Article All Versions of this Article:
24/2/144    most recent
00258.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Huang, Y.
Right arrow Articles by Sutter, T. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Huang, Y.
Right arrow Articles by Sutter, T. R.
Received 17 October 2005; accepted in final form 23 November 2005.
Physiological Genomics 24:144-153 (2006)
1094-8341/05 $8.00 © 2006 American Physiological Society

Identification of novel transcriptional networks in response to treatment with the anticarcinogen 3H-1,2-dithiole-3-thione

Yong Huang1, Jian Yan1, Ronald Lubet2, Thomas W. Kensler3 and Thomas R. Sutter1

1 W. Harry Feinstone Center for Genomic Research, University of Memphis, Memphis, Tennessee
2 Division of Cancer Prevention, National Cancer Institute, National Institutes of Health, Rockville, Maryland
3 Department of Environmental Health Sciences, Johns Hopkins University Bloomberg School of Public Health, Baltimore, Maryland


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
3H-1,2-dithiole-3-thione (D3T), an inducer of antioxidant and phase 2 genes, is known to enhance the detoxification of environmental carcinogens, prevent neoplasia, and elicit other protective effects. However, a comprehensive view of the regulatory pathways induced by this compound has not yet been elaborated. Fischer F344 rats were gavaged daily for 5 days with vehicle or D3T (0.3 mmol/kg). The global changes of gene expression in liver were measured with Affymetrix RG-U34A chips. With the use of functional class scoring, a semi-supervised method exploring both the expression pattern and the functional annotation of the genes, the Gene Ontology classes were ranked according to the significance of the impact of D3T treatment. Two unexpected functional classes were identified for the D3T treatment, cytosolic ribosome constituents with 90% of those genes increased, and cholesterol biosynthesis with 91% of the genes repressed. In another novel approach, the differentially expressed genes were evaluated by the Ingenuity computational pathway analysis tool to identify specific regulatory networks and canonical pathways responsive to D3T treatment. In addition to the known glutathione metabolism pathway (P = 0.0011), several other significant pathways were also revealed, including antigen presentation (P = 0.000476), androgen/estrogen biosynthesis (P = 0.000551), fatty acid (P = 0.000216), and tryptophan metabolism (P = 0.000331) pathways. These findings showed a profound impact of D3T on lipid metabolism and anti-inflammatory/immune-suppressive response, indicating a broader cytoprotective effect of this compound than previously expected.

functional cluster scoring; Gene Ontology; Ingenuity Pathways Knowledge Database; canonical pathway; Nrf2


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
DITHIOLETHIONES ARE A CLASS of effective cancer chemopreventive agents (33), of which the unsubstituted parent compound, 3H-1,2-dithiole-3-thione (D3T), is the most potent. This agent blocks carcinogenesis through induction of phase 2 enzymes that enhance the detoxification of xenobiotic compounds and thereby protect against mutation and initiation of neoplasia (22, 26). However, the benefits of inducing a broad adaptive response are not limited to cancer chemoprevention. Because oxidative and electrophilic stress has been implicated in numerous pathophysiological conditions including cancer, inflammation, cardiovascular diseases and aging, agents that enhance cellular antioxidative capacity may exert profound cytoprotective effects (16). Understanding the molecular pathways affected by inducers of phase 2 and antioxidative enzymes is of crucial importance for developing better prophylactic strategies for disease prevention. Toward this goal, a comprehensive view of both beneficial and toxic effects elicited by D3T is required. However, it is difficult to concurrently characterize the multiple molecular mechanisms of biological response with traditional methods.

Microarray simultaneously monitors the expression of a large number of genes covering various biological machineries. With progress in systems biology, numerous knowledge databases have been developed to host the constellation of genes, their functions, and relationships. The global influence of molecular events can be illustrated by combining the gene expression profiles detected by microarray and the gene functional profiles annotated by such knowledge databases. Previous microarray studies performed using either D3T-treated rats or mice deficient in the transcription factor Nrf2, a target of D3T (41), have produced lists of differentially expressed genes (28, 44) but have failed to establish how these genes may form regulatory networks. A systematic examination of D3T-mediated functions and pathways with statistical analysis has not yet been available. Moreover, these studies have ignored the genes that do not pass the randomly or empirically determined criteria for gene selection.

In the present study, we have coupled novel approaches to identify functions and pathways regulated by D3T, focusing not only on the genes selected by strict criteria but also on all of the genes that are expressed in liver. The functional profiles of all expressed genes were ranked according to the response to D3T treatment using a Gene Ontology (GO) clustering tool, functional class scoring (FCS) (37). We also adopted a computational tool, Ingenuity pathway analysis, to identify regulatory networks of differentially expressed genes and the corresponding canonical pathways that govern the response to D3T treatment. A significance P value was calculated for each function and pathway. Interestingly, the functional profiles detected by FCS using all expressed genes and the canonical pathways revealed by Ingenuity pathway analysis using differentially expressed genes both demonstrate that the known D3T-mediated functions (e.g., transferase and oxidoreductase activities) and pathways (e.g., glutathione metabolism) are significantly regulated by D3T (low P value). These known D3T-mediated functions served as benchmarks to assess the relative significance of other functions and pathways. The FCS and Ingenuity pathway analysis also uncovered unique novel functional classes, gene networks, and pathways that have not been previously related to D3T, indicating that these two approaches can complement one another in the exploration of the functional profile of microarray data. Additional experiments can be proposed based on the new findings to investigate the broad protective and potential toxic effects of D3T and related compounds.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
Animal treatment and microarray analysis.
Male Fischer F344 rats were obtained from Harlan at 6 wk of age. General procedures for animal care and housing were in accordance with the Guide for the Care and Use of Laboratory Animals, National Research Council, 1996. The animals were singly housed in polycarbonate cages in a temperature (22 ± 1°C)- and humidity (30–70%)-controlled room with a 14:10-h light-dark cycle, respectively. Rat AIN 76A chow (Harlan Teklad) and tap water were provided ad libitum. Eight animals were randomly assigned into two groups (n = 4) and treated by gavage with 100 µl of either vehicle (saturated sucrose) or D3T (0.3 mmol/kg body wt) once every other day for 5 days. Animals were killed 24 h after the third dose. Liver tissues were harvested and snap frozen. The protocol for this study was approved by the University of Memphis Animal Care Committee.

Total RNA was isolated from the frozen tissue using Stat-60 (Tel-Test). The mRNA levels were measured using Affymetrix RG-U34A chips (containing 8,799 probe sets) according to standard GeneChip Expression assay protocol. After hybridization, the chips were scanned using a GeneChip Scanner. The eight chips were normalized by Affymetrix GeneChip Operating Softwear (GCOS) v1.2 to determine the probe set intensities and "Present" (P) or "Absent" (A) calls. The microarray data have been submitted to the National Center for Biotechnology Information (NCBI)'s Gene Expression Omnibus repository (series accession no. GSE3173; or sample IDs GSM71311, GSM71368–GSM71374).

RT-PCR analysis of gene expression.
The D3T-regulated genes selected by microarray were further validated by quantitative, real-time, two-step RT-PCR (qPCR) to evaluate potential false positives. qPCR was performed with SYBR Green PCR kit (Bio-Rad). The amplification curve was generated using iCycler iQ Real-Time Detection system (Bio-Rad). Primers were designed by use of DNASTAR (DNASTAR). Fold changes of expressions relative to vehicle control animals were determined after normalization to the ß-actin gene. Amplification specificity was confirmed by melting-curve analysis. The gene symbols and sequence of forward (F) and reverse (R) primers are as follows: ß-actin, F-TCACCCACACTGTGCCCATCTATGA/R-GAGGAAGAGGATGCGGCAGTGG; Afar, F-CGCAGCGGCTGCAAAGTAAA/R-AGTGCCGTGGTCTGGAAAGTGTAA; Hmgcs1, F-AGGTGCCCGTGACTGCTGCTC/R-AGTGCCCTGCCCATCCCTCCTA; Pcsk9, F-TTAGTCTTCGCCCAGAGCAT/R-CTCCTCAGGCACACTGTTGA; Insig1, F-TACTGACCAGCCCAGGACAACACAA/R-AACGCGAAATGAATGCCTGCTGAG.

FCS analysis.
FCS analysis was performed with the software downloaded from http://www.geneontology.org/GO.tools.microarray.shtml#ermine and implemented in a JAVA environment. In FCS analysis, all of the expressed genes in a particular GO class were examined (37). The expressed genes were present (determined by "P calls" with Affymetrix GCOS v1.2 analysis) in at least three of the four chips of either vehicle control or D3T group. The intensities of the expressed genes were analyzed by unpaired t-test to determine the significance of change. The false discovery rate (FDR) was controlled by Benjamini-Hochberg procedure (43) using GeneSpring v7.0 (Agilent Technologies). The adjusted P values (called q-values) of all the expressed genes were used in the FCS analysis (38). For repeated occurrence of a gene (a gene was represented by 2 or more probe sets in the chips), only the best (minimum) q-value was used.

GO classes of the three ontologies (biological process, molecular function, and cellular component) were scrutinized equally (18). To reduce bias, GO classes with fewer than 8 or more than 200 genes were not included (14). Two GO terms were considered related if they were in parent-child relationships or contained large portions of identical genes. The significance of each GO class was determined by its P value.

Hierarchical clustering.
Expression values of genes in GO classes exhibiting significance were processed by dChip software (http://www.dchip.org/) for hierarchical clustering analysis. The default parameter of dChip was used in the clustering algorithm (30).

Biological network and pathway analysis.
For Ingenuity network and pathway analysis, the moderately changed or unregulated genes play an equal role with the highly significantly regulated genes in constructing gene networks and thus result in irrelevant networks. Therefore, a data set containing only the identifiers of the significantly up- or downregulated genes with their corresponding fold changes was uploaded as a tab-delimited text file into the Ingenuity software (http://www.ingenuity.com). Stringent criteria combining P calls (present in at least 3 of the 4 chips in either vehicle control or D3T group), fold changes (≥1.7-fold change), and unpaired t-test followed by Benjamini-Hochberg procedure controlling FDR (q ≤ 0.05) were used to select the differentially expressed genes.

This web-delivered application makes use of the Ingenuity Pathways Knowledge Base (IPKB) containing large amounts of individually modeled relationships between gene objects (e.g., genes, mRNAs, and proteins) to dynamically generate significant biological networks and pathways. The submitted genes that are mapped to the corresponding gene objects in the IPKB are called "focus genes." The focus genes are used as the starting point for generating biological networks. To start building networks, Ingenuity software queries the IPKB for interactions between focus genes and all the other genes stored in the IPKB and generates a set of networks with a maximum network size of 35 genes. A P value for each network and canonical pathway is calculated according to the fit of the user's set of significant genes. This is done by comparing the number of focus genes that participate in a given network or pathway, relative to the total number of occurrences of those genes in all networks or pathways stored in the IPKB. The score of network is displayed as the negative log of the P value, indicating the likelihood of the focus genes in a network being found together due to random chance. Therefore, scores of 2 have at least 99% confidence of not being generated by chance alone. In the current study, a score of 12 or higher was used to select highly significant biological networks regulated by D3T.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
GRANTS
 REFERENCES
 
Microarray data processing and validation.
All rats remained in good health and without difference in chow consumption throughout the experimental period. No difference in body weight gain of D3T and control groups was observed. No gross or microscopic changes in liver were observed. The levels of serum alanine aminotransferase and aspartate aminotransferase activity did not differ significantly between the two groups (Karuri A, Huang Y, and Sutter TR, unpublished results).

We performed microarray analysis to identify transcriptional networks and pathways regulated by D3T. The ".cel" files of the eight Affymetrix RG-U34A chips were analyzed in Affymetrix GCOS v1.2 to determine the intensities and "P" or "A" calls of each probe set. The probe sets not present in at least 3 of the 4 chips in either vehicle control or D3T group were considered as meaningless and therefore were eliminated to reduce data complexity. The remaining 3,153 probe sets represented the expressed genes in the liver of either vehicle- or D3T-treated rats. The FDR q-values of these 3,153 probe sets were used in FCS analysis. A list of 292 probe sets was generated using the combined criteria of fold change >1.7 and q <0.05 in expressed genes. On the basis of the annotation in NCBI's Unigene and Gene database, the 292 probe sets represented 248 unique genes, including 67 upregulated genes and 181 downregulated genes (Supplemental Table S1, A and B; Supplemental Material is available at the Physiological Genomics web site).1 These results are consistent with the findings of Kwak et al. (27) that fewer gene transcripts remain elevated after feeding of a D3T-containing diet for several days compared with the number of transcripts elevated by a single oral dose.

In the current study, we adopted very strict criteria to select the D3T-regulated genes by combining P calls of probe sets, fold changes, and FDR q-values. Although false positives might still be unavoidable, at least three lines of evidence support the general validity of our microarray data. First, the previously reported Nrf2-regulated genes, such as glutathione-S-transferase-{alpha} type 2, NAD(P)H:quinone oxidoreductase 1, and UDP-glucose dehydrogenase (3), were also induced in our D3T-treated animals 117.2-, 3.5-, and 2.0-fold, respectively. Second, we evaluated the internal consistency of our microarray data using the data of differentially expressed genes represented by two different (duplicate) probe sets on the chip. In the 248 differentially expressed genes, there were a total of 25 genes represented by duplicate probe sets. As shown in Fig. 1A, all the duplicate probe sets displayed consistent direction of alteration with similar extent of fold change, indicating very high confidence of our gene selection strategy. Third, the qPCR confirmed not only a strongly induced gene, Afar (23), but also the moderately repressed genes Hmgcs1, Pcsk9, and Insig1 (Fig. 1B). The fold changes of these four genes detected by qPCR were consistent with but slightly larger than that detected by microarray. This might be due to the higher sensitivity of qPCR compared with microarray.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 1. Cross-validation of microarray result. A: 25 differentially expressed genes were represented by 2 different probe sets for each gene. These probe sets were examined in respect to their direction and extent of response. B: to validate the microarray data, the fold changes of 4 genes analyzed by quantitative, real-time, 2-step RT-PCR (qPCR) were compared with their corresponding fold changes detected by microarray.

 
FCS analysis ranked the GO classes based on their response to D3T treatment.
By microarray analysis, thousands of genes were analyzed simultaneously. This identifies not only individual genes whose expression levels are altered by the experimental condition but also reveals the functional and regulatory relationship of these genes. By combining the expression values and functional annotations of all expressed genes, we can identify the most interesting functional categories of genes and thus uncover the major effects of the treatment. This goal can be achieved with a semi-supervised method called FCS, by ranking the GO classes based on their P values (37). The P value of each GO class was calculated from the FDR q-values of individual genes in that GO class. All of the 3,153 expressed probe sets with their q-values and functional annotations (GOs) were analyzed simultaneously by FCS. This allows genes that might not have been selected as differentially expressed genes to still have a positive impact on the GO class and thus preserves information contained in the genewise scores (q-values).

The top 18 significant GO classes influenced by treatment were ranked according to the P value of each GO class (Table 1). However, many GO classes are overlapping or redundant. First, in the GO database, there are three GO categories (biological process, molecular function, and cellular component) describing different aspects of a gene product (18). Therefore, many GO classes in different categories overlap with each other regarding the gene products they contain. Second, in each GO category, the GO classes are structured in parent-child relationships. One parent GO class may be subdivided into several child GO classes, while one child GO class may inherit from multiple parent GO classes (4). Therefore, many child GO classes might be identical or largely overlap with their parent classes. The GO classes with large portions of overlapping genes were combined into the same functional class. As listed in Table 2, three major functional classes were identified from the GO classes with low P values. The largest functional class consisted of the GO classes containing large portions of genes with transferase and oxidoreductive activities, which were the known functions mediated by D3T (3).


View this table:
[in this window]
[in a new window]
  		Table 1. Functional class scoring analysis of the expressed genes in rat liver

 

View this table:
[in this window]
[in a new window]
  		Table 2. Significant canonical pathways (P ≤ 0.05) involved by D3T-regulated genes

 
D3T treatment inhibited the hepatic cholesterol synthesis and enhanced the hepatic cytosolic ribosome constituents.
The presence of glutathione transferase activity, response to xenobiotic stimuli, and aldehyde metabolism as the top significant GO classes in Table 1 illustrated that FCS analysis correctly identified functional classes that have previously been demonstrated to be modulated by D3T (3, 26) and provided a greater context for screening the impact of D3T treatment on other functional classes. Surprisingly, at least two functional classes, the steroid/cholesterol synthesis (GO: 0016126/GO: 0006695) (P = 0.00075/0.00121) and cytosolic ribosome constituents (GO: 0005830) (P = 0.00222) were also assigned low P values that were comparable with those of transferase activity (GO: 0016758) (P = 0.00081) and oxidoreductase activity (GO: 0016616) (P = 0.00225).

It is important to note that a low P value for a GO class does not imply that all genes in the class are significantly altered. The rationale for including genes displaying subtle alteration in expression levels is based on the concept that such genes may be highly relevant to the biological function of the treatment when viewed in a large context of interacting genes. Because the alterations of most genes in the GO classes of cholesterol biosynthesis and cytosolic ribosome constituents are not significant according to our criteria for differentially expressed genes, we displayed the expression profile instead of fold change of each unique gene across the eight samples using dChip. Strikingly, >90% (11/12) of the unique genes in the cholesterol biosynthesis class were inhibited (Fig. 2, cluster D), whereas >90% (56/62) of the genes in the cytosolic ribosome class were enhanced (Supplemental Fig. S1, cluster U), implying decreased cholesterol biosynthesis and increased cytosolic ribosome activity under D3T treatment. Moreover, the eight samples were correctly classified into vehicle control group and D3T group based on the expression pattern of all cytosolic ribosome genes (Supplemental Fig. S1). This finding indicated that the expression pattern of cytosolic ribosome genes is a transcriptional signature to distinguish the vehicle control and D3T group. Such a sample clustering cannot be performed based on the expression pattern of cholesterol biosynthesis genes due to the limited number of genes (only 12 genes). While these findings are consistent with the previous reports showing that a set of genes for cholesterol/lipid biosynthesis, including sterol regulatory element-binding protein-1 (Srebp1), are inhibited in mouse liver (23), and some ribosome proteins such as L18a and S16 are induced by D3T treatment in rat liver (40), none of these earlier studies could establish the relative pathway significance of such an observation.



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 2. Hierarchical clustering of cholesterol biosynthesis genes. All of the expressed genes of cholesterol biosynthesis (GO: 0006695) in the RG-U34A chip were clustered based on their expression patterns across the 8 samples using dChip software. There were 12 expressed genes belonging to the GO class of cholesterol biosynthesis. More than 91% of them were classified into the downregulated cluster D. GO, Gene Ontology.

 
The microarray and qPCR results demonstrated that three genes (Hmgcs1, Pcsk9, and Insig1) involved in the regulation of cholesterol synthesis and metabolism were also inhibited by D3T treatment (Fig. 1B). Hmgcs1 (3-hydroxy-3-methylglutaryl-CoA synthase) catalyzes the very early step of cholesterol biosynthesis by condensing acetoacetyl-CoA with another molecule of acetyl-CoA to form hydroxamethylglutaryl-CoA (HMG-CoA). Chemicals inhibiting Hmgcs1 can repress the hepatic sterol synthesis in vivo (36). Pcsk9 (proportion convertase subtilisin kexin-9) is a member of the subtilisin serine protease family, which has an important role in cholesterol metabolism. Overexpression of Pcsk9 in mice leads to increased total and low-density lipoprotein (LDL) cholesterol levels by decreasing hepatic LDL receptor protein levels (34). Conversely, inhibitors of Pcsk9 may act synergistically with statins to enhance LDL receptor levels and reduce plasma cholesterol (42). Insig1 (insulin-induced gene-1), a membrane protein on endoplasmic reticulum, is responsive to insulin level (39) and plays a central role in cholesterol homeostasis. When cellular sterol rises, Insig1 binds to Scap and retains the Scap/Srebp1 complex in endoplasmic reticulum, thus inhibiting cholesterol synthesis (55). With excess sterol, Insig1 also binds to HMG-CoA reductase, facilitating its ubiquitination and proteasomal degradation (49). The downregulation of Insig1 upon D3T treatment might be due to reduced hepatocellular sterol/cholesterol level and/or the inhibited insulin signaling pathway. Collectively, the influence of D3T treatment displayed on the regulation of sterol/cholesterol biosynthesis is very intriguing and warrants further study.

Ingenuity pathway analysis validated glutathione metabolism as a significant pathway induced by D3T treatment.
Although identification of a list of individual genes that show expression changes is important, there is an increasing need to move beyond this level of analysis. Instead of simply enumerating a list of genes, we want to know how they interact as parts of complexes, pathways and biological networks. For this purpose, the 248 differentially expressed genes were imported into the Ingenuity pathway analysis software to identify biological networks and pathways. The networks described functional relationships between gene products based on known interactions in the literature. Biological functions were assigned to each gene network, and these networks were then associated with canonical pathways. Nine highly significant networks with score ≥12 were identified from the 248 genes regulated by D3T treatment (Supplemental Table S2).

We first asked whether Ingenuity pathway analysis can identify a known pathway induced by D3T treatment. As shown in Fig. 3A, this analysis rapidly validated glutathione metabolism (10) as the major pathway in the first network. All the D3T-regulated genes involved in glutathione metabolism were increased (Fig. 3A). These included Gsta1 (3.5-fold), Gsta2 (117.2-fold), Gstp1 (30.8-fold), Gstt1 (1.8-fold), G6pd (2.8-fold), and Gclc (2.1-fold). Gclc is the catalytic subunit of glutamate-cysteine ligase. It catalyzes the rate-limiting reaction in glutathione biosynthesis in an Nrf2-dependant manner (54).



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 3. Glutathione metabolism regulated by 3H-1,2-dithiole-3-thione (D3T) treatment. A: glutathione metabolism regulated by Nrf2 (the node labeled as NFE2L2) is highlighted in blue. Uncommon gene symbols: Gsta1, glutathione-S-transferase A1; Gstt1, glutathione-S-transferase-{theta}1; Gstp1, glutathione-S-transferase-{pi}; Gclc, catalytic subunit of glutamate-cysteine ligase; FTL, ferritin. The intensity of node color indicates the degree of up- (red) or downregulation (green). B: explanation of the symbols, the edges, and their labels. The length of an edge reflects the evidence supporting that node-to-node relationship, in that edges supported by more articles from the literature are shorter.

 
Previous studies have shown that D3T induces carcinogen detoxication and antioxidant genes via an Nrf2-dependent mechanism (41). In our microarray data analysis, Nrf2 (referred to as NFE2L2 in the 1st network) was not selected as a differentially expressed gene. This is consistent with the finding of Kwak et al. (26) that Nrf2 expression increases at 6 h but returns to normal levels by 24 h after D3T treatment. Other Nrf2-regulated genes involved in glutathione metabolism, including Gsr (glutathione reductase, 2.3-fold) in the fifth and Gstm5 (glutathione-S-transferase M5, 2.6-fold) in the second network, were also increased by D3T treatment. Moreover, canonical pathway analysis confirmed that the glutathione metabolism was a significant pathway (P = 0.00111) induced by D3T treatment (Table 2).

Ingenuity analysis revealed lipid and tryptophan metabolism as the most significant pathways influenced by D3T.
Because Ingenuity pathway analysis successfully identified the known function of D3T, we further asked whether Ingenuity pathway analysis could reveal novel functions of D3T. As shown in Table 2, at least four of the nine significant canonical pathways were related to lipid metabolism, including fatty acid metabolism (P = 0.000216), androgen and estrogen metabolism (P = 0.000551), sterol biosynthesis (P = 0.0261), and bile acid biosynthesis (P = 0.0311). The strongest induction was observed in Acaa1 (acetyl-CoA acyltransferase-1, 4.0-fold), the gene encoding the enzyme catalyzing the ß-oxidation of the fatty acid moiety of acyl-CoA in the peroxisomal ß-oxidation of fatty acids. The enzyme cleaves long chain fatty acyl-CoA to generate acetyl-CoA and shortened acyl-CoA (47). In addition, seven cytochrome P450 genes involved in fatty acid metabolism were significantly regulated by D3T. Six of them were downregulated, including Cyp1a2 (–2.7-fold), Cyp2c40 (–2.4-fold), Cyp3a2 (–3.2-fold), Cyp3a3 (–1.7-fold), Cyp3a7 (–2.0-fold), and Cyp51a1 (–1.9-fold). The repression of the cytochrome P450 genes, especially Cyp3a, might inhibit {alpha}-hydroxylation of fatty acids in liver (7). The negative influence on multiple cytochrome P450 genes might also interfere with pharmacokinetics of other drugs undergoing the hepatic metabolism and thus increase their potential toxicities. All the D3T-regulated genes involved in bile acid synthetic pathway were increased (Table 2). This outcome might facilitate detoxification by increasing excretion of conjugated metabolites through bile acids. The genes involved in estrogen and androgen biosynthesis, such as Smp2a (rat senescence marker protein 2A), Sult2a2 (sulfotransferase family 2A, member 2), and Nsdhl [NAD(P)-dependent steroid dehydrogenase-like] (50, 8, 31) were inhibited, while five UDP-glucuronosyltransferase genes involved in the clearance of the sex hormone (5) were induced by D3T treatment. Fatty acid metabolism was the most significantly influenced canonical pathway on D3T treatment (with the lowest P value).

The second most significantly influenced canonical pathway influenced by D3T was the tryptophan pathway (P = 0.000331). There are several known tryptophan metabolic pathways, including its degradation to serotonin (32). While melatonin, a metabolite of serotonin, is known to be effective as a free radical scavenger and may have anticarcinogenic effects, its metabolite 6-hydroxymelatonin may exhibit carcinogenic potential through enhancement of oxidative DNA damage (46). At least seven cytochrome P450 genes that were involved in the metabolism of melatonin into 6-hydroxymelatonin were regulated by D3T. Six of them, including Cyp1a2 (–2.7-fold) (15), were significantly inhibited. Tryptophan-to-kynurenine transformation is one of the alternative tryptophan metabolic pathways, involved in the biosynthesis of NAD coenzyme from tryptophan (17). In rat, tryptophan is mainly metabolized along the kynurenine pathway (2). Our data showed that D3T inhibited the expression of kynureninase (–1.9-fold), the enzyme catalyzing the cleavage of L-kynurenine and L-3-hydroxykynurenine into anthranilic and 3-hydroxyanthranilic acids, respectively (1). Therefore, D3T treatment might also have a negative effect on the NADH salvage pathway in the liver.

Antigen presentation and interferon regulatory factor-1 pathways were inhibited by D3T treatment.
Activation of the immune response is an important protective mechanism. However, it may also lead to tissue damage under some conditions (19, 53). The three main types of cells in the liver are hepatocytes, sinusoidal endothelial cells, and the bone marrow-derived Kupffer cells. Previous studies have shown that all three cell types present antigens (24). Major histocompatibility complex class I (MHC-I) is expressed by all nucleated cells including hepatocytes, whereas MHC-II is expressed by antigen-presenting cells such as Kupffer cells and sinusoidal endothelial cells (9). As shown in Table 2, the antigen presentation pathway was significantly affected by D3T (P = 0.000476). In the third network, at least six genes (Cd74, Canx, Hla-E, H2-Aa, Hla-DRB1, and Tap1) involved in the antigen presentation pathway were inhibited by D3T (Fig. 4). Another gene (Psmb9) in the antigen presentation pathway was identified in the sixth network and also inhibited by D3T (Supplemental Table S2). The global canonical pathway of antigen presentation is shown in Fig. 5. In summary, D3T treatment suppresses multiple genes participating in both the MHC-I- and MHC-II-associated antigen presentation pathways.



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 4. Immune response and regulation network mediated by D3T. Six D3T-regulated genes (CD74, CANX, HLA-E, H2-Aa, HLA-DRB1, and TAP1) in the 3rd network (score = 18, see Supplemental Table S2) were involved in the antigen presentation pathway. All of them were significantly suppressed by D3T treatment. Uncommon gene symbols: CANX, calnexin; TAP1, transporter-1, ATP-binding cassette, subfamily B. For explanation of the symbols, letters, and edges, see legend to Fig. 3B.

 


View larger version (39K):
[in this window]
[in a new window]
  		Fig. 5. Antigen presentation pathway inhibited by D3T treatment. Canonical antigen presentation pathways (P = 0.000476, see Table 2) are displayed. Gray nodes represent D3T-regulated genes. Gene symbols equivalent to those in Fig. 4: major histocompatibility complex class I{alpha} (MHC-I{alpha}) = HLA-E, MHC-II{alpha} = H2-Aa, MHC-IIß = HLA-DRB1, LMP2 = PSMB9, CLIP = CD74. For explanation of symbols, see legend to Fig. 3B.

 
Antigen recognition can increase the retention of activated T cells, including MHC-I- activated CD8+ and MHC-II-activated CD4+ T cells (35, 12). Immune-mediated liver injury occurs in a wide variety of systemic immune/inflammatory responses (45, 21, 29). Such injury of hepatocytes is triggered by MHC-restricted recognition of specific antigenic peptides on hepatocytes, e.g., during immunopathology associated with the CD8+ T cell response to hepatitis B virus antigens (6), and during the ischemia/reperfusion-induced inflammatory response mediated by CD4+ T cells (56). Identification of new targets to suppress the overstimulated immune/inflammatory response will significantly contribute to the development of hepatoprotective treatments. The analysis of the current microarray data indicated that D3T could be a candidate for this purpose. However, because the main focus of research on dithiolethiones is currently for chemoprevention, where such drugs will be chronically administered, care should be taken in future studies to assess both beneficial and potential immunotoxic effects of D3T-induced suppression of antigen presentation.

Other evidence for immune/inflammatory suppression by D3T was derived from the inhibitory effect of D3T on Irf1-regulated genes. Irf1 functions as a transcription activator of genes induced by interferon (IFN)-{alpha}, -ß, and -{gamma} (52). Although the expression level of Irf1 itself was not altered, many target genes regulated directly or indirectly by Irf1 in the third network were inhibited by D3T treatment (Fig. 4). For example, Hrasls3 (HRAS-like suppressor-3), Cybb (cytochrome b-245, ß-polypeptide), and Ctss (cathepsin S) (48, 13, 51) were downregulated –3.4-, –2.9-, and –1.7-fold, respectively. It is known that some activators of inflammation such as lipopolysaccharide (LPS) can induce upregulation of the MHC class I and II response in hepatocytes, which is accompanied by increased levels of IFN-{gamma} in plasma (20). Moreover, the toxic effect of LPS can be inhibited by cyclosporine and a monoclonal antibody against IFN-{gamma} (20) by suppressing the immune/inflammatory response. The bioinformatics analysis in the current study suggested that D3T could suppress the immune/inflammatory response by inhibiting both MHC-I- and MHC-II-mediated antigen presentation and also the Irf1-regulated genes. As predicted, a study in our laboratory (unpublished data) demonstrated the protective effect of D3T in an LPS-induced inflammatory model. Both alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels in plasma were significantly lower after LPS injection in D3T-treated rats compared with the vehicle-treated rats (Karuri A, Huang Y, and Sutter TR, unpublished observations), indicating the hepatoprotective effects of D3T against severe inflammation.

Our study of transcriptional networks was carried out on rats, where Nrf2 knockout animals are not available. Thus we were unable to clarify whether all of our reported networks are Nrf2 dependant. However, a recent cell culture study indicates that activation of Nrf2 signaling exerts profound protective effects through anti-inflammatory mechanisms, e.g., inhibiting inducible nitric oxide synthase and blocking the formation of nitrite in LPS-treated cells. This effect is abolished in cells in which Nrf2 has been disrupted (11).

In conclusion, our study was designed to understand the global functional profiles of D3T. We used sophisticated tools for microarray data analysis to identify novel functional classes, biological networks, and canonical pathways induced or repressed by D3T. The inhibition of cholesterol synthesis and the enhancement of cytosolic ribosome constituents are demonstrated as one of the primary responses to D3T treatment. Our pathway analysis revealed the profound effects of D3T on lipid and tryptophan metabolism. In addition, D3T treatment also repressed the immune response by inhibiting both MHC-I- and MHC-II-mediated antigen presentation pathways.

In addressing a pharmacological response, the biological representation of a set of genes is more interesting than the genes themselves. In the current study, the biological representations of genes are presented in GO classes, transcriptional networks, and canonical pathways associated with weight of significance (P values). The two approaches taken here are complementary in identifying functional profile regulated by D3T treatment and therefore can serve as a model for dissection of many other more complex pharmacological responses.


   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
REFERENCES
 
This work was supported by the National Cancer Institute Contract N01-CW-95114-MAO and Grant R01-CA39416 and the W. Harry Feinstone Center for Genomic Research.


   FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: T. R. Sutter, W. Harry Feinstone Center for Genomic Research, Univ. of Memphis, 3774 Walker Ave., Memphis, TN 38152 (e-mail: tsutter{at}memphis.edu)

10.1152/physiolgenomics.00258.2005.

1 The Supplemental Material for this article (Supplemental Fig. S1 and Supplemental Tables S1, A and B, and S2) is available online at http://physiolgenomics.physiology.org/cgi/content/full/00258.2005/DC1. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 

  1. Alberati-Giani D, Buchli R, Malherbe P, Broger C, Lang G, Kohler C, Lahm HW, and Cesura AM. Isolation and expression of a cDNA clone encoding human kynureninase. Eur J Biochem 239: 460–468, 1996.[ISI][Medline]
  2. Allegri G, Ragazzi E, Bertazzo A, Costa CV, and Rocchi R. Tryptophan metabolism along the kynurenine pathway in rats. Adv Exp Med Biol 527: 481–496, 2003.[Medline]
  3. Anderson L, Steele VK, Kelloff GJ, and Sharma S. Effects of oltipraz and related chemoprevention compounds on gene expression in rat liver. J Cell Biochem Suppl 22: 108–116, 1995.[Medline]
  4. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, and Sherlock G. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 25: 25–29, 2000.[CrossRef][ISI][Medline]
  5. Belanger A, Hum DW, Beaulieu M, Levesque E, Guillemette C, Tchernof A, Belanger G, Turgeon D, and Dubois S. Characterization and regulation of UDP-glucuronosyltransferases in steroid target tissues. J Steroid Biochem Mol Biol 65: 301–310, 1998.[CrossRef][Medline]
  6. Bertoletti A, Ferrari C, and Fiaccadori F. Role of the cell-mediated immune response in the pathogenesis of hepatitis B virus infection: possible immune-therapeutic strategies. Acta Biomed Ateneo Parmense 67: 87–93, 1996.[Medline]
  7. Boocock DJ, Maggs JL, White IN, and Park BK. Alpha-hydroxytamoxifen, a genotoxic metabolite of tamoxifen in the rat: identification and quantification in vivo and in vitro. Carcinogenesis 20: 153–160, 1999.[Abstract/Free Full Text]
  8. Chatterjee B, Song CS, Kim JM, and Roy AK. Androgen and estrogen sulfotransferases of the rat liver: physiological function, molecular cloning, and in vitro expression. Chem Biol Interact 92: 273–279, 1994.[CrossRef][Medline]
  9. Chen M, Tabaczewski P, Truscott SM, Van Kaer L, and Stroynowski I. Hepatocytes express abundant surface class I MHC and efficiently use transporter associated with antigen processing, tapasin, and low molecular weight polypeptide proteasome subunit components of antigen processing and presentation pathway. J Immunol 175: 1047–1055, 2005.[Abstract/Free Full Text]
  10. Davidson NE, Egner PA, and Kensler TW. Transcriptional control of glutathione S-transferase gene expression by the chemoprotective agent 5-(2-pyrazinyl)-4-methyl-1,2-dithiole-3-thione (oltipraz) in rat liver. Cancer Res 50: 2251–2255, 1990.[Abstract/Free Full Text]
  11. Dinkova-Kostova AT, Liby KT, Stephenson KK, Holtzclaw WD, Gao X, Suh N, Williams C, Risingsong R, Honda T, Gribble GW, Sporn MB, and Talalay P. Extremely potent triterpenoid inducers of the phase 2 response: correlations of protection against oxidant and inflammatory stress. Proc Natl Acad Sci USA 102: 4584–4589, 2005.[Abstract/Free Full Text]
  12. Doherty DG and O'Farrelly C. Innate and adaptive lymphoid cells in the human liver. Immunol Rev 174: 5–20, 2000.[CrossRef][Medline]
  13. Eklund EA, Jalava A, and Kakar R. PU.1, interferon regulatory factor 1, and interferon consensus sequence-binding protein cooperate to increase gp91(phox) expression. J Biol Chem 273: 13957–13965, 1998.[Abstract/Free Full Text]
  14. Erraji-Benchekroun L, Underwood MD, Arango V, Galfalvy H, Pavlidis P, Smyrniotopoulos P, Mann JJ, and Sibille E. Molecular aging in human prefrontal cortex is selective and continuous throughout adult life. Biol Psychiatry 57: 549–558, 2005.[Medline]
  15. Facciola G, Hidestrand M, von Bahr C, and Tybring G. Cytochrome P450 isoforms involved in melatonin metabolism in human liver microsomes. Eur J Clin Pharmacol 56: 881–888, 2001.[CrossRef][ISI][Medline]
  16. Ferrari CK and Torres EA. Biochemical pharmacology of functional foods and prevention of chronic diseases of aging. Biomed Pharmacother 57: 251–260, 2003.[CrossRef][Medline]
  17. Gaertner FH, Shetty AS. Kynureninase-type enzymes and the evolution of the aerobic tryptophan-to-nicotinamide adenine dinucleotide pathway. Biochim Biophys Acta 482(2):453–60, 1977.
  18. Gene Ontology Consortium. The Gene Ontology (GO) database and informatics resource. Nucleic Acids Res 32: D258–D261, 2004.[Abstract/Free Full Text]
  19. Hines IN and Wheeler MD. Recent advances in alcoholic liver disease. III. Role of the innate immune response in alcoholic hepatitis. Am J Physiol Gastrointest Liver Physiol 287: G310–G314, 2004.[Abstract/Free Full Text]
  20. Jephthah-Ochola J, Urmson J, Farkas S, and Halloran PF. Regulation of MHC expression in vivo. Bacterial lipopolysaccharide induces class I and II MHC products in mouse tissues by a T cell-independent, cyclosporine-sensitive mechanism. J Immunol 141: 792–800, 1988.[Abstract]
  21. Ju C. Immunological mechanisms of drug-induced liver injury. Curr Opin Drug Discov Devel 8: 38–43, 2005.[ISI][Medline]
  22. Kensler TW, Egner PA, Dolan PM, Groopman JD, and Roebuck BD. Mechanism of protection against aflatoxin tumorigenicity in rats fed 5-(2-pyrazinyl)-4-methyl-1,2-dithiol-3-thione (oltipraz) and related 1,2-dithiol-3-thiones and 1,2-dithiol-3-ones. Cancer Res 47: 4271–4277, 1987.[Abstract/Free Full Text]
  23. Kensler TW, Groopman JD, Eaton DL, Curphey TJ, and Roebuck BD. Potent inhibition of aflatoxin-induced hepatic tumorigenesis by the monofunctional enzyme inducer 1,2-dithiole-3-thione. Carcinogenesis 13: 95–100, 2001.
  24. Kmiec Z. Cooperation of liver cells in health and disease. Adv Anat Embryol Cell Biol 161: III–XIII, 1–151, 2001.[Medline]
  25. Kwak MK, Itoh K, Yamamoto M, and Kensler TW. Enhanced expression of the transcription factor Nrf2 by cancer chemopreventive agents: role of antioxidant response element-like sequences in the nrf2 promoter. Mol Cell Biol 22: 2883–2892, 2002.[Abstract/Free Full Text]
  26. Kwak MK, Itoh K, Yamamoto M, Sutter TR, and Kensler TW. Role of transcription factor Nrf2 in the induction of hepatic phase 2 and antioxidative enzymes in vivo by the cancer chemoprotective agent, 3H-1, 2-dimethiole-3-thione. Mol Med 7: 135–145, 2001.[ISI][Medline]
  27. Kwak MK, Wakabayashi N, Itoh K, Motohashi H, Yamamoto M, and Kensler TW. Modulation of gene expression by cancer chemopreventive dithiolethiones through the Keap1-Nrf2 pathway. Identification of novel gene clusters for cell survival. J Biol Chem 278: 8135–8145, 2003.[Abstract/Free Full Text]
  28. Lee JM, Calkins MJ, Chan K, Kan YW, and Johnson JA. Identification of the NF-E2-related factor-2-dependent genes conferring protection against oxidative stress in primary cortical astrocytes using oligonucleotide microarray analysis. J Biol Chem 278: 12029–12038, 2004.
  29. Leevy CB and Elbeshbeshy HA. Immunology of alcoholic liver disease. Clin Liver Dis 9: 55–66, 2005.[Medline]
  30. Li C and Hung Wong W. Model-based analysis of oligonucleotide arrays: model validation, design issues and standard error application. Genome Biology 2: research0032, 2001.
  31. Liu XY, Dangel AW, Kelley RI, Zhao W, Denny P, Botcherby M, Cattanach B, Peters J, Hunsicker PR, Mallon AM, Strivens MA, Bate R, Miller W, Rhodes M, Brown SD, and Herman GE. The gene mutated in bare patches and striated mice encodes a novel 3beta-hydroxysteroid dehydrogenase. Nat Genet 22: 182–187, 1999.[CrossRef][ISI][Medline]
  32. Llambias EB, Aldonatti C, and San Martin de Viale LC. Tryptophan metabolism via serotonin in rats with hexachlorobenzene experimental porphyria. Biochem Pharmacol 66: 35–42, 2003.[Medline]
  33. Maxuitenko YY, Curphey TJ, Kensler TW, and Roebuck BD. Protection against aflatoxin B1-induced hepatic toxicity as short-term screen of cancer chemopreventive dithiolethiones. Fundam Appl Toxicol 32: 250–259, 1996.[CrossRef][ISI][Medline]
  34. Maxwell KN, Fisher EA, and Breslow JL. Overexpression of PCSK9 accelerates the degradation of the LDLR in a post-endoplasmic reticulum compartment. Proc Natl Acad Sci USA 102: 2069–2074, 2005.[Abstract/Free Full Text]
  35. Mehal WZ, Juedes AE, and Crispe IN. Selective retention of activated CD8+ T cells by the normal liver. J Immunol 163: 3202–3210, 1999.[Abstract/Free Full Text]
  36. Nagashima H, Kumagai H, Tomoda H, and Omura S. Inhibition of hepatic cholesterol biosynthesis by a 3-hydroxy-3-methylglutaryl coenzyme A synthase inhibitor, 1233A, in mice. Life Sci 52: 1595–1600, 1993.[Medline]
  37. Pavlidis P, Lewis DP, and Noble WS. Exploring gene expression data with class scores. Pac Symp Biocomput: 474–485, 2002.
  38. Pavlidis P, Qin J, Arango V, Mann JJ, and Sibille E. Using the gene ontology for microarray data mining: a comparison of methods and application to age effects in human prefrontal cortex. Neurochem Res 29: 1213–1222, 2004.[CrossRef][ISI][Medline]
  39. Peng Y, Schwarz EJ, Lazar MA, Genin A, Spinner NB, and Taub R. Cloning, human chromosomal assignment, and adipose and hepatic expression of the CL-6/INSIG1 gene. Genomics 43: 278–284, 1997.[CrossRef][ISI][Medline]
  40. Primiano T, Gastel JA, Kensler TW, and Sutter TR. Isolation of cDNAs representing dithiolethione-responsive genes. Carcinogenesis 17: 2297–2303, 1996.[Abstract/Free Full Text]
  41. Ramos-Gomez M, Kwak MK, Dolan PM, Itoh K, Yamamoto M, Talalay P, and Kensler TW. Sensitivity to carcinogenesis is increased and chemoprotective efficacy of enzyme inducers is lost in nrf2 transcription factor-deficient mice. Proc Natl Acad Sci USA 98: 3410–3415, 2001.[Abstract/Free Full Text]
  42. Rashid S, Curtis DE, Garuti R, Anderson NN, Bashmakov Y, Ho YK, Hammer RE, Moon YA, and Horton JD. Decreased plasma cholesterol and hypersensitivity to statins in mice lacking Pcsk9. Proc Natl Acad Sci USA 102: 5374–5379, 2005.[Abstract/Free Full Text]
  43. Reiner A, Yekutieli D, and Benjamini Y. Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics 19: 368–375, 2003.[Abstract/Free Full Text]
  44. Roebuck BD, Curphey TJ, Li Y, Baumgartner KJ, Bodreddigari S, Yan J, Gange SJ, Kensler TW, and Sutter TR. Evaluation of the cancer chemopreventive potency of dithiolethione analogs of oltipraz. Carcinogenesis 24: 1919–1928, 2003.[Abstract/Free Full Text]
  45. Russell JQ, Morrissette GJ, Weidner M, Vyas C, Aleman-Hoey D, and Budd RC. Liver damage preferentially results from CD8(+) T cells triggered by high affinity peptide antigens. J Exp Med 188: 1147–1157, 1998.[Abstract/Free Full Text]
  46. Sakano K, Oikawa S, Hiraku Y, and Kawanishi S. Oxidative DNA damage induced by a melatonin metabolite, 6-hydroxymelatonin, via a unique non-o-quinone type of redox cycle. Biochem Pharmacol 68: 1869–1878, 2004.[Medline]
  47. Schram AW, Goldfischer S, van Roermund CW, Brouwer-Kelder EM, Collins J, Hashimoto T, Heymans HS, van den Bosch H, Schutgens RB, Tager JM, et al. Human peroxisomal 3-oxoacyl-coenzyme A thiolase deficiency. Proc Natl Acad Sci USA 84: 2494–2496, 1987.[Abstract/Free Full Text]
  48. Sers C, Husmann K, Nazarenko I, Reich S, Wiechen K, Zhumabayeva B, Adhikari P, Schroder K, Gontarewicz A, and Schafer R. The class II tumour suppressor gene H-REV107–1 is a target of interferon-regulatory factor-1 and is involved in IFNgamma-induced cell death in human ovarian carcinoma cells. Oncogene 21: 2829–2839, 2002.[CrossRef][Medline]
  49. Sever N, Song BL, Yabe D, Goldstein JL, Brown MS, and DeBose-Boyd RA. Insig-dependent ubiquitination and degradation of mammalian 3-hydroxy-3-methylglutaryl-CoA reductase stimulated by sterols and geranylgeraniol. J Biol Chem 278: 52479–52490, 2003.[Abstract/Free Full Text]
  50. Song CS, Kim JM, Roy AK, and Chatterjee B. Structure and regulation of the senescence marker protein 2 gene promoter. Biochemistry 29: 542–551, 1990.[CrossRef][Medline]
  51. Storm van's Gravesande K, Layne MD, Ye Q, Le L, Baron RM, Perrella MA, Santambrogio L, Silverman ES, and Riese RJ. IFN regulatory factor-1 regulates IFN-gamma-dependent cathepsin S expression. J Immunol 168: 4488–4494, 2002.[Abstract/Free Full Text]
  52. Taniguchi T, Ogasawara K, Takaoka A, and Tanaka N. IRF family of transcription factors as regulators of host defense. Annu Rev Immunol 19: 623–655, 2001.[CrossRef][ISI][Medline]
  53. Vergani D and Mieli-Vergani G. Autoimmune hepatitis. Minerva Gastroenterol Dietol 50: 113–123, 2004.[Medline]
  54. Wild AC, Moinova HR, and Mulcahy RT. Regulation of {gamma}-glutamylcysteine synthetase subunit gene expression by the transcription factor Nrf2. J Biol Chem 274: 33627–33636, 1999.[Abstract/Free Full Text]
  55. Yang T, Espenshade PJ, Wright ME, Yabe D, Gong Y, Aebersold R, Goldstein JL, and Brown MS. Crucial step in cholesterol homeostasis: sterols promote binding of SCAP to INSIG-1, a membrane protein that facilitates retention of SREBPs in ER. Cell 110: 489–500, 2002.[CrossRef][ISI][Medline]
  56. Zwacka RM, Zhang Y, Halldorson J, Schlossberg H, Dudus L, and Engelhardt JF. CD4(+) T-lymphocytes mediate ischemia/reperfusion-induced inflammatory responses in mouse liver. J Clin Invest 100: 279–289, 1997.[ISI][Medline]



This article has been cited by other articles:


Home page
 Physiol. GenomicsHome page
C. M. Olsen, Y. Huang, S. Goodwin, D. C. Ciobanu, L. Lu, T. R. Sutter, and D. G. Winder
Microarray analysis reveals distinctive signaling between the bed nucleus of the stria terminalis, nucleus accumbens, and dorsal striatum
Physiol Genomics, February 19, 2008; 32(3): 283 - 298.
[Abstract] [Full Text] [PDF]

Home page
 Mol. Cell. Biol.Home page
M. K. Cho, W. D. Kim, S. H. Ki, J.-I. Hwang, S. Choi, C. H. Lee, and S. G. Kim
Role of G{alpha}12 and G{alpha}13 as Novel Switches for the Activity of Nrf2, a Key Antioxidative Transcription Factor
Mol. Cell. Biol., September 1, 2007; 27(17): 6195 - 6208.
[Abstract] [Full Text] [PDF]

This Article
Free upon publication Free Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figure and Tables
Right arrowFree Article All Versions of this Article:
24/2/144    most recent
00258.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Huang, Y.
Right arrow Articles by Sutter, T. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Huang, Y.
Right arrow Articles by Sutter, T. R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2006 by the American Physiological Society.