PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Mol Carcinog. Author manuscript; available in PMC 2010 May 4.
Published in final edited form as:
PMCID: PMC2863309
NIHMSID: NIHMS87946

Genome wide transcriptional profiling in breast cancer cells reveals distinct changes in hormone receptor target genes and chromatin modifying enzymes after proteasome inhibition

Abstract

Steroid hormone receptors, like glucocorticoid (GR) and estrogen receptors (ER), are master regulators of genes that control many biological processes implicated in health and disease. Gene expression is dependent on receptor levels which are tightly regulated by the ubiquitin-proteasome system. Previous studies have shown that proteasome inhibition increases GR, but decreases ER-mediated gene expression. At the gene expression level this divergent role of the proteasome in receptor-dependent transcriptional regulation is not well understood. We have used a genomic approach to examine the impact of proteasome activity on GR and ER-mediated gene expression in MCF-7 breast cancer cells treated with dexamethasone (DEX) or 17β-estradiol (E2), the proteasome inhibitor MG132 (MG) or MG132 and either hormone (MD or ME2) for 24h. Transcript profiling reveals that inhibiting proteasome activity modulates gene expression by GR and ER in a similar manner in that several GR and ER target genes are up-regulated and down-regulated after proteasome inhibition. In addition, proteasome inhibition modulates receptor-dependent genes involved in the etiology of a number of human pathological states, including multiple myeloma, leukemia, breast/prostate cancer, HIV/AIDS and neurodegenerative disorders. Importantly, our analysis reveals that a number of transcripts encoding histone and DNA modifying enzymes, prominently histone/DNA methyltransferases and demethylases, are altered after proteasome inhibition. As proteasome inhibitors are currently in clinical trials as therapy for multiple myeloma, HIV/AIDs and leukemia, the possibility that some of the target molecules are hormone regulated and by chromatin modifying enzymes is intriguing in this era of epigenetic therapy.

Keywords: Proteasome inhibitor, receptors: glucocorticoid, estrogen, gene expression profiling, microarray analysis

Introduction

Glucorticoids and estrogens play a crucial role in regulating transcription of many genes that are important regulators of diverse physiological processes, including development, reproduction, bone formation/resorption, energy metabolism, cholesterol mobilization and immunity. The physiological actions of glucocorticoids and estrogens are mediated primarily through the glucocorticoid receptor (GR) and estrogen receptor (ER). Glucocorticoid and estrogen receptors are ligand-dependent transcription factors and members of the nuclear hormone receptor super family [1]. Upon hormone binding these receptors localize in the nucleus where they associate with specific hormone response elements within promoter sequences embedded in chromatin [2]. To activate or repress target genes, steroid hormone receptors recruit various co-regulator complexes, including chromatin remodeling complexes to modify local chromatin structure [3,4]. Receptor and coregulator levels play key roles in controlling appropriate physiological outcomes in specific target tissues. Similar to other steroid hormone receptors, GR and ER are tightly regulated by the ubiquitin proteasome system (UPS) [reviewed in [5,6]. Additionally, levels of nuclear hormone receptor co-regulators are also regulated by the UPS [7-9]. Briefly, the UPS plays an important role in a variety of cellular functions primarily via its proteolytic activity, although recent studies implicate the components of the pathway in direct regulation of specific transcriptional processes [reviewed in [10,11]. The 26S proteasome is the principal biochemical machinary that degrades short lived cellular proteins and rids the cell of damaged and misfolded polypeptides, in addition to providing basic housekeeping functions [12]. The 26S proteasome is a multi-enzyme complex made of a 20S catalytic ‘core’, capped by the 19S regulatory complex [13,14]. The 19S complex is composed of two sub-complexes: the lid and the base composed of six AAA-type ATPases and two non-ATPase subunits. Proteolysis of a target protein by the 26S proteasome, involves two intricate steps [13,14]. First, the protein is tagged with ubiquitin (Ub), a conserved 76 amino acid polypeptide, or, more precisely, with a poly-Ub chain of defined length and topology to generate the polyubiquitin degradation signal [14]. Secondly, the tagged protein is degraded by the 26S proteasome complex. Conjugation of ubiquitin to the protein substrate is mediated by a multi-enzyme cascade consisting of an Ub-activating enzyme (E1), an Ub-conjugating enzyme (E2), and an Ub ligase (E3) [15].

Control of cellular protein levels by the ubiquitin–proteasome system is essential for various cellular functions and ultimately dysregulation of the system is associated with many pathological conditions [16,17]. Although the role of the ubiquitin-proteasome system in regulating many transcription factors, such as p53, is well established, the system has only recently been linked to steroid hormone receptor function. There is a general agreement that the ubiquitin-proteasome system and particularly the proteolytic activity of the proteasome is critical for promoting the exchange of transcriptional factors on chromatin and possibly facilitating multiple rounds of transcription initiation, hence controlling receptor mediated gene expression [6,10,11,18,19. In addition, a number of ubiquitin proteasome pathway enzymes, such as E6 associated protein (E6-AP) and the marine double minute-2 (Mdm2), have been identified as steroid receptor co-activator [reviewed by {Kinyamu, 2005 #388]. Furthermore, specific components of the proteasome, such as the 19S subunit, thyroid interacting protein 1 (TRIP1/Sug1) and the 20S beta subunit low molecular mass polypeptide 2 (LMP2) are implicated in receptor-mediated transcriptional regulation [20,21]. Consequently, receptor turnover is tightly linked to receptor-mediated transcription.

Two main observations led us to the current study. First, our laboratory and others showed that proteasome inhibitors, such as MG132, increase GR mediated transcriptional activation of the mouse mammary tumor virus promoter (MMTV) in breast cancer cells [22,23]. Secondly, other groups showed that proteasome inhibitors were inhibitory to nuclear receptor function particularly that of the ER [19,24]. These findings suggested that proteasome activity differentially modulates gene transcription in a receptor dependent manner. This divergent role of the proteasome in receptor-dependent transcriptional regulation is not well understood. Since previous experiments suggesting a requirement for proteasome activity in ER, but not GR were conducted using specific model genes, we used microarray analysis to test the requirement for proteasome activity in the regulation of global gene expression mediated by these two receptors. Data from the global gene expression analysis show that inhibiting proteasome activity modulates gene expression mediated by GR and ER in a similar manner. Specifically, the requirement for proteasome activity is gene, but not receptor specific. Proteasome activity modulates receptor dependent genes involved in the etiology of a number of diseases, including leukemia, HIV/AIDS and neurodegenerative disorders. Intriguingly, proteasome inhibition modulates a subset of transcripts that encode factors that regulate RNA polymerase II and DNA/histone modifying enzymes. Our study provides a snapshot of global gene expression after proteasome inhibition in breast cancer cells treated with either dexamethasone or 17β-estradiol. These data provide a useful tool particularly since proteasome inhibitors are currently in clinical trials as potential therapeutics for various diseases.

Materials and Methods

Cell Culture

The generation of MCF-7 cells stably expressing the GR and endogenous ERα has been described previously [25]. Briefly, parental MCF-7 cells (American Type Culture Collection, Manassas, Va.) were co-transfected with pGR-NEO and a neomycin resistance plasmid, pRSV-NEO, using the calcium phosphate precipitation method (GIBCO-BRL Life Technologies, Grand Island, NY) [26]. The resulting cell line which expresses both GR and ER shows similar gene expression profiles in response to17β-estradiol compared to MCF-7 from other laboratories [27-29]. Similar to ER, the GR in MCF-7 cells activates known exogenous and endogenous GR target genes [25,30,31].

For the current study, cells were grown in a humidified incubator at 37°C with 5% CO2 in MEM supplemented with 2 mM glutamine, 100 μg/mL penicillin/streptomycin, 10 mM HEPES, 10% FBS and 300 μg/mL G418. For glucocorticoid treatment, cells were seeded overnight in phenol red-free MEM supplemented with 5% charcoal-stripped calf serum and 2 mM glutamate. Cells treated with 17β-estradiol were cultured in MEM media with 5% charcoal stripped serum for 3 days and then seeded for experiments as described for microarray analysis.

Antibodies and Western Blotting

After washing twice with PBS, cells were pelleted by centrifugation. For whole cell extracts, cells were lysed as previously described [25]. Twenty to 50 μg of protein was resolved on 4-12 % SDS-PAGE and transferred to a PVDF membrane (Amersham). Proteins were immunoblotted using the following antibodies: anti-GR-BUGR2 (Dr. B. Gametchu, Medical College of Wisconsin, Milwaukee, WI), ERα-H-184 Santa Cruz Biotechnology, β-Actin (Sigma), GAPDH (Research Diagnostics Inc).

Gene Expression Profiling and Analysis

Gene expression analysis was performed using Agilent Human1A array (pattern id = 01152) (Agilent Technologies, Palo Alto, CA). Total RNA samples were prepared from two biological replicates of MCF-7 cells treated with vehicle, 1 nM dexamethasone or 10 nM 17β-estradiol (24 hr), 1 mM MG132 (24 hr) or MG132 and dexamethasone or 17β-estradiol (24 hr) using RNeasy Midi Kits (Invitrogen). Total RNA was labeled with Cyanine (Cy) 3- or Cy5-dCTP (Amersham, Piscataway, NJ) using the Agilent Fluorescent Direct Label Kit protocol with a slight modification in the starting amount (10 μg was used rather than 20 μg). Each RNA pair (vehicle and either dexamethasone, 17β-estradiol, MG132, MG132 and dexamethasone, or 17β-estradiol and dexamethasone) was mixed and hybridized to an array at two separate times employing fluor reversal. Hybridizations were performed for 17 hours in a rotating hybridization oven using the Agilent 60-mer oligo microarray processing protocol. Slides were washed as indicated in this protocol and then scanned with an Agilent Scanner.

Data were retrieved with the Agilent Feature Extraction software (v7.1), using defaults for all parameters, except the Ratio terms. To account for the use of the Direct Label protocol, error terms were changed as suggested by Agilent as follows: Cy5 multiplicative error = 0.15, Cy3 multiplicative error = 0.25, Cy5 additive error = 20, Cy3 additive error = 20. The Agilent Feature Extraction Software adjusted the data to account for additive and multiplicative noise in the array data acquisition process. The resulting ratio intensity value for each gene feature on the array was averaged across technical and biological replicates as follows: the log base 10 ratio values from all four arrays for each comparison [two biological replicates, each with a fluor reversal (technical replicate)] were averaged in the Rosetta Resolver® system (Rosetta Biosoftware, Kirkland, WA) using the error-weighted approach [32]. Briefly, letting x(i) represent the ith log base 10 ratio value for a gene and σx(i) the measurement error, the error-weighted average for a gene feature is

x¯=iw(i)x(i)iw(i),wherew(i)=1σx2,i=1:N,andNis the number of replicates.

A p-value for each gene feature is computed based upon the reproducibility of the expression measurements across the four arrays (biological and technical replicates). Gene features with p < 0.001 for a given comparison were considered significantly and differentially expressed.

Validation of microarray results by real-time RT-PCR

The microarray data trends were verified by examining a subset of representative classes of genes after treatment with hormone and proteasome inhibitor for 24 hr. To establish whether the genes were direct targets of the hormone or proteasome inhibitor, expression of select genes was monitored after treating the cells for 2 hr. Because MG132 is known to inhibit targets other than the 26S proteasome, expression of a subset of genes was also determined after a similar treatment with the highly specific proteasome inhibitor epoxomicin. After removing genomic DNA, total RNA (1-2 μg) from cells treated with the vehicle, hormone or the proteasome inhibitor (MG132 or epoxomicin) in the presence or absence of hormone were reverse transcribed using oligo-dt as described in the Superscript Kit (Invitrogen Corp.). The cDNA was treated with ribonuclease H (Invitrogen Corp.) to remove RNA:DNA hybrids. The cDNA was diluted 5-fold with DNAse-free water and used for real-time PCR analysis.

Real-Time PCR Analysis

cDNA levels were detected using the STRATAGENE, Mx3000P™ real time PCR system and SYBR Green I dye (STRATAGENE, Cedar Creek, TX). Primers were created using Applied Biosystems Primer Express Software version 2.0. For cDNA amplification, 2-5 μL of cDNA was combined with SYBR Green PCR mix as described by the manufacturer (STRATAGENE, Cedar Creek, TX). GAPDH mRNA expression was used as the endogenous control for normalization of initial RNA levels. Data is expressed as relative expression.

Chromosome Map

Genes that were found to be significant in Rosetta Resolver (p < 0.001) following treatment by MG132, MG132 + DEX, and MG132 + E2 were displayed in the Physical Position View for the Agilent Human 1A array (011521) in Agilent's GeneSpring GX software (version 7.3.1).

Microarray data accession number

The microarray data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO Series accession number GSE8383″ [33].

Results

Global transcriptional changes in glucocorticoid and estrogen receptor targets after proteasome inhibition

It is well known that cellular levels of steroid hormone receptors including those of the glucocorticoid (GR) and estrogen receptors (ER) are tightly regulated by proteosomal degradation. Consequently proteasome inhibition by widely used proteasome inhibitors such as MG132, block ligand dependent degradation and stabilize receptor levels (Figure 1A and and2A).2A). However, previous studies using model reporter gene assays have shown that proteasome inhibition increased GR-mediated gene transcription, whereas ER-mediated gene transcription is decreased. Since receptor levels, should correlate with gene expression, the divergent effect of proteasome inhibition on gene expression mediated by the two receptors is not well understood. To examine the global role of proteasome activity, we turned to transcript profiling to provide genome wide view of gene expression in response to proteasome inhibitor and hormone in MCF-7 cells. We compared transcripts from RNA treated with vehicle (Con) vs. dexamethasone (D or DEX) or 17β-estradiol (E2) vs. those treated with proteasome inhibitor MG132 (MG) vs. MG132 plus dexamethasone (MD) or MG132 plus 17β-estradiol (ME2). Those genes differentially expressed were clustered and displayed in dendograms (Figure 1B and and2B).2B). In all figures MD designates treatment with MG132 and dexamethasone (D), whereas ME2 designates treatment with MG132 and 17β-estradiol (E2).

Figure 1Figure 1Figure 1Figure 1
Global transcriptional profile from MCF-7 cells treated with dexamethasone or proteasome inhibitor. (A) Proteasome inhibition blocks ligand dependent GR turnover. Whole cell extracts from untreated cells (lane 1), cell treated with DEX for 4 (+) or 24 ...
Figure 2Figure 2Figure 2Figure 2
Global transcriptional profile from MCF-7 cells treated with 17β-estradiol or proteasome inhibitor. (A). Proteasome inhibition blocks ligand dependent ER turnover. Whole cell extracts from untreated cells (lane 1), cells treated with E2 for 4 ...

Proteasome inhibition has a synergistic and antagonistic effect on glucocorticoid-induced gene expression

In the first set of analysis we concentrated on genes affected by treatment with DEX alone or with DEX and MG. Clustering analysis revealed 4 broad transcript categories. The first category represents genes affected by glucocorticoid treatment only. Of the over 20,000 genes on the Agilent human 1a array, 268 genes were up-regulated and 118 down-regulated when cells were treated with DEX alone (Figure 1B and C). In the second category, 209 genes (131 + 78) were similarly affected by DEX and MG treatment (Figure 1C); of these 131 genes were up-regulated and 78 were down-regulated. In a third category, although 48 transcripts were affected in common by DEX and MG, the effect of the treatment on a specific gene was antagonistic; e.g., treatment with MG blocked DEX induction or repression of the gene (Figure 1C). A fourth category consisting of a total of 2945 genes that were affected when cells were treated with MG and DEX in a hormone independent manner, 1290 and 1655 gene transcripts were increased and decreased, respectively. We further explored the transcripts in the 3 categories where the hormone response is affected by proteasome inhibition. Genes from the fourth category are primarily affected by proteasome inhibition and are discussed in section 3. It is important to note that transcript profiling resulting in microarray analysis, as carried out in this study, only deciphers ‘relative’ changes among genes and not genome wide gene expression. While validation of all the genes identified was not practical, we chose a representative sample that was subsequently analyzed by quantitative RT-PCR to verify the microarray trends.

Among the genes in the first category affected by DEX alone were bona fide GR targets. These include 11-β-hydroxysteroid dehydrogenase type 2 (HSD11β2), msh homeobox homolog 2 (MSX2), dual specificity phosphatase 6 (DUSP6) and sin 3A associated protein (SAP 30) (Figure 1D and Table 1-1). Some genes known to be repressed by GR like neurturin (NRTN), adhesion molecule with Ig like domain 1 (Amigo1), heterogeneous nuclear ribonucleoprotein A2/B1 (HNRPA2B1) and melanoma antigen family D4 (MAGED4) were down-regulated by DEX alone (Figure 1D, Table 1-1). HSD11β2 is a well established target of GR mediated activation. As predicted from the microarray analysis, treatment with DEX (D) for 24 hr increases HSD11β2 expression over 100-fold (Figure 1D, 24hr), whereas treatment with MG132 alone (MG) or with dexamethasone (MD) had no significant effect HSD11β2 expression compared to control. Furthermore, the HSD11β2 mRNA expression increased (6-fold) within 2 hr after dexamethasone treatment, indicating direct regulation of this gene by the GR (Figure 1D, 2hr). In a similar manner, treatment with dexamethasone decreased NTRN expression by 90% compared to control as predicted from microarray analysis (Figure 1D, 24hr). Compared to DEX treatment, treatment with proteasome inhibitor did not significantly affect NTRN expression, suggesting DEX-dependent repression of this gene at 24hr. This repression was not detected at an earlier time point in which DEX treatment increased NTRN expression 2-fold (Figure 1D, 2hr). Notably, treatment with proteasome inhibitor does not significantly changed NTRN expression compared to DEX.

The second category of transcripts was synergistically altered by MG and DEX (Figure 1E, Table 1-2). As demonstrated previously for model genes in vitro, proteasome inhibition enhanced glucocorticoid-mediated gene expression [22,23]. Similar to the effect observed with MMTV-LUC and CAT reporter gene, proteasome inhibition enhanced expression of some well characterized GR target genes [34-38]. These include S100 calcium binding protein (S100P), regulator of G protein signaling (RGS2) also known as G0S8, RNA Pol II elongation factor 2 (ELL2) and dual specificity phosphatase 1 (DUSP1) (Figure 1E, Table 1-2). Among the genes in this category were genes not previously shown to be glucocorticoid inducible, such as alpha B crystallin (CRYAB) and N-Myc downstream regulated gene 1 (NDRG1) which are mildly activated by DEX, but highly up-regulated after proteasome inhibition. Other genes in this category include collagen type VI, alpha 1 (COL6A1), musculoaponeurotic fibrosarcoma oncogene B (MAFB) and annexin 1 (ANXA1) (Figure 1E, Table 1-2). For this class of genes we validated expression of S100 P after treatment with DEX (D) or inhibitor and DEX (MG, MD). At 24 hr, treatment with DEX (D) increased S100P expression by 30-fold, MG alone was not significantly different from control. Treatment with MG and DEX (MD) synergistically increased S100P expression 120-fold, an effect significantly larger than the sum of the individual effect of hormone or inhibitor alone. (Figure 1E-24hr). A similar effect is observed when the cells were treated with DEX or MG for 2 hrs. DEX induced S100P expression 3-fold at early time points and this effect was potentiated by proteasome inhibition (6-fold) (Figure 1E- 2hr).

Conversely, proteasome inhibition facilitates glucocorticoid-mediated repression as seen for the GR target adhesion molecule with an Ig-like domain 2 (AMIGO2), 2-5-oligoadenylate synthetase 2 (OAS2), interferon-responsive protein 28 or receptor transporting protein 4 (RTP4/IFRG28), androgen-induced basic leucine zipper (AIBZIP/CREB3L4), neuronal cell adhesion molecule (NCAM2) and other transcripts, such as fasciculation and elongation protein zeta 1 (FEZ1) and hedgehog acyltransferase (HHAT) and transforming growth factor beta 3 (TGFB3) (Figure 1E, Table 1-2). Expression of TGFB3 was validated as an example of those genes repressed. At 24 hr, treatment with DEX (D) decreased TGFB3 expression by 50 percent. Treatment with MG and DEX (MD) synergistically decreased TFGB3 expression by over 90%, an effect significantly larger than the sum of the individual effect of hormone or inhibitor alone. (Figure 1E-24hr). Significant TGFB3 repression did not occur at shorter time points under these experimental conditions, although a trend to decrease was observed (Figure 1E- 2hr).

For the third category, treatment with either proteasome inhibitor or hormone had an antagonistic effect on gene expression. An antagonistic response was viewed as one where the inhibitor blocks hormone induction or repression of a transcript and vise versa. This third category of genes was different from that described in Figure 1D (Table 1-1). In the first category, the hormone exerts its main effect on gene expression, whereas in the third category the hormone or proteasome inhibitor have an independent effect on gene expression, which is reversed in the opposite manner by either agent; i.e. antagonism. Proteasome inhibition attenuates DEX induction of a number of bona fide GR targets including, galanin (GAL), baculoviral IAP repeat-containing 3 (BIRC3) and B-Cell CLL/lymphoma 6 (BCL6) (Figure 1F, Table 1-3). For some genes DEX-induced changes in the levels of certain transcripts, but these transcripts were completely repressed by proteasome inhibition. These included transcripts for calcium binding protein A8 (S100A8), prolactin inducible protein (PIP), TAR (HIV) RNA binding protein (TARBP1) and transcripts encoding interferon genes IFIH1 and IFIT2 (Figure 1F, Table 1-3). The results from the microarray analysis were confirmed by RTPCR using GAL and IFIT2 as a representative gene for this class (Figure 1F). GAL expression increased 26-fold after treatment with DEX (D) for 24 hr, and this effect was reduced 7-fold by MG, which was very similar to microarray analysis (Table 1-3). A short time treatment with DEX induced GAL expression only 2-fold, and proteasome inhibition did not affect this induction, suggesting an indirect effect of inhibitor observed at 24 hr. A second example of antagonistic response was detected when DEX-mediated repression was abrogated by proteasome inhibition. Treatment with dexamethasone reduced IFIT2 expression by 85%, whereas treatment with MG alone increased IFIT2 expression 4-fold compared to control (Figure 1F). Co-treatment with dexamethasone and inhibitor reversed DEX-mediated repression by 8-fold as predicted by microarray analysis (Table 1-3). A short treatment time with DEX decreased IFIT2 expression by 60% with a smaller but consistent effect of the proteasome inhibitor compared to 24 hr treatment (Figure 1F-2hr).

Because MG132 has targets other than the 26S proteasome, we validated a select number of gene targets after treatment with a second proteasome inhibitor, epoxomicin. Gene expression profiles for HSD11B2, S100P and GAL following epoxomicin exposure were similar to those observed after MG132 treatment (Figure S1 A-C).

Proteasome inhibition has a synergistic and antagonistic effect on estrogen response

Previous studies suggested that proteasome inhibition repressed ER-mediated gene expression [19,24]. We therefore examined the effect of proteasome inhibition on estrogen response (Figure 2B, Table 2-1-4). We compared transcripts treated with E2 to those from cells treated with MG alone or MG plus E2. Genes were classified into 4 categories as carried out for the glucocorticoid response. The first category of genes was specifically altered by E2 treatment; 272 transcripts were up-regulated and 126 down-regulated, respectively (Figure 2C). Among those transcripts up-regulated by E2 were bona fide ER targets including early growth response 3 (EGR3), retinoblastoma binding protein 8 (RBBP8) and low density lipoprotein receptor related 8 (LRP8) (Figure 2D, Table 2-1). Transcripts repressed included grainyhead like protein 1 (GRHL1) or leader-binding protein 32 (LBP-32), transcripts encoding histone H2A (H2AFA) and H2B (H2BFQ) (Figure 2 D, Table 2-1). EGR3 is a well established target of ER. As predicted from the microarray analysis, treatment with E2 for 24 hr increased EGR3 expression 65-fold (Figure 2 D, 24hr), whereas treatment with MG132 alone (MG) led to a significant increase in expression compared to control. However, co-administration of drug and hormone (ME2) resulted in a smaller increase than seen with E2 alone. EGR3 mRNA expression increased (52-fold) within 2 hr after E2 and the inhibitor had no significant effect alone (MG) or on the ER-mediated induction (ME2), confirming EGR3 is primarily an ER target gene (Figure 2D, 2hr), In contrast to EGR3, LBP-32 was repressed (70%) by E2 at both time points (Figure 2D). Treatment with MG132 alone or with MG132 and E2 did not lead to a significant change in expression compared to control or E2.

The second category of genes were those synergistically up-regulated (66) or down-regulated (122) by treatment with MG and E2 (Figure 2E, Table 2-2). Among ER targets up-regulated after E2 and MG treatment was a GTP binding protein over expressed in skeletal muscle (GEM), tubulin beta 2 (TUBB2A), DEAD (Asp-Glu-Ala-Asp) box polypeptide 10 (DDX10) and cofilin 2 (CFL2). Proteasome inhibition also synergistically repressed ER targets including the well characterized ER target, thioredoxin interacting protein (TXNIP), calcium/calmodulin dependent kinase II inhibitor 1 (CANK2N1), SRY (sex determining region Y) box 13 (Sox 13), neuronal cell adhesion molecule (NCAM2), cadherin 10 type 2 (CDH10) CREB3L4/AIBZIP, AMIGO2 and S100 A8 (Figure 2E, Table 2-2). For this class of genes DDX10 and AMIGO2 expression were validated as representative genes. Treatment with E2 or inhibitor MG and E2 (MG, ME2) for 24 hr increased DDX10 expression by 2-fold; MG alone was only 6-fold. Treatment with MG and E2 (ME2) increases DDX10 expression 7.5-fold (Figure 2E-24hr). The synergistic action of proteasome inhibition of E2-mediated increase in DDX10 expression was more evident at 2 hr, whereas treatment with E2 induced DDX10 (13-fold) and treatment with MG and E2 led to a 26-fold induction (Figure 2E-2hr). As an additional positive control, we observed that proteasome inhibition increased E2 induction of pS2, a known ER target gene (Figure S2 A-B).

In the third category, as shown for the glucocorticoid response, proteasome inhibition antagonized the effects of estrogen response. Proteasome inhibition abrogated the effect of E2 on amphiregulin (AREG), epiregulin (EREG) and retinol binding protein 7 (RBP7) (Figure 2F, Table 2-3). A classic example of the previously reported repression of proteasome inhibition on ER-mediated regulation is the effect on the progesterone receptor (PGR), which is increased by E2, but repressed by MG (validation data not shown). Additionally, other ER targets including stromal derived factor 1 (SDF-1/CXCL12), collagen, type XII, alpha 1 (COL12A1), minichromosome maintenance deficient 6 (MCM6), DNA (cytosine-5) methyltransferase 1 (DNMT1) are induced by E2, but significantly repressed by MG (Figure 2F, Table 2-3). Other targets were repressed by E2, but up-regulated by proteasome inhibition (Figure 2F, Table 2-3). These included the lipocalin-2 (LCN2), a putative in vivo estrogen target gene and paracrine factor that mediates the growth regulatory effects of estrogen in normal breast epithelium. Additionally, tribbles homolog 3 (TRIB3), a negative regulator of NF-kappaB, interferon–induced protein with tetrapeptide repeats 2 (IFIT2) and sel-1-suppressor of lin-12 like (SEL1L), which plays a role in pancreatic carcinoma and breast cancer (Figure 2F, Table 2-3). There were also transcripts repressed by E2, but the repression dampened by proteasome inhibition, for example the immunoglobin-like domain counter receptor 1 (ILDR1) (Figure 2F, Table 2-3). Expression of SDF-1 was validated as example a gene that was activated by E2, but repressed by inhibitor (Figure 2F). SDF-1 expression increased 12-fold after treatment with E2 for 24 hr, and this effect is inhibited 3-fold by MG, very similar to what was observed in the microarray analysis (Figure 2F-24 hr, Table 2-3). SDF-1 is a direct target of ER and a short treatment time with E2 induces SDF-1 expression 8-fold. The impact of proteasome inhibition is observed at 24 hr suggesting an indirect effect of the inhibitor (Figure 2F-2 hr). In another characteristic antagonism, treatment with E2 for 24 hr decreased expression 30%, whereas treatment with MG alone increased IFIT2 expression 4-fold compared to control (Figure 1F). Co-treatment with E2 and inhibitor reversed E2-mediated repression, thereby increasing IFT2 expression by 7-fold, which was similar to that observed in microarray analysis (Table 2-3). A short treatment time with E2 induced IFIT2 repression by 30% with a smaller, but consistent antagonistic effect of the proteasome inhibitor (Figure 2F-2 hr). Interestingly, the effect of proteasome inhibition on ER-mediated induction and repression of SDF-1 and IFIT2, respectively, was very similar to that observed for the GR targets GAL and IFIT2 (Figure 1F). Furthermore IFIT2 is a target of both hormones and proteasome inhibition has similar inhibition effect on DEX and E2 mediated repression (Figure 1F and and2F).2F). This observation solidifies the idea that the two receptors behave in a similar manner when the proteasome is inhibited. We further show that proteasome inhibition by epoxomicin on ER-dependent gene expression is similar to that observed with MG132 treatment (Figure S3, A-C).

Specific effect of proteasome inhibitor on gene expression

The fourth category of genes represents those primarily affected by proteasome inhibition (MG). The transcripts activated in this class presumably do not require proteasome activity, while it may be required for the repressed transcripts. Some genes in this category were not significantly changed by either hormone acting alone, but significant changes in gene expression were observed after treatment with proteasome inhibitor and hormone. To pinpoint transcripts only affected by MG, we compared transcripts from MG alone with those affected by MG plus DEX or MG plus E2 (Figure 3A). A total of 583 genes were altered by MG alone. Of these genes, 294 were up-regulated and 289 down-regulated. Among the specific genes increased by proteasome inhibitor exclusively were replication factor C1 (activator 1) (RFC1), 5-azacystidine induced gene 2 (AZI2), proteasome subunits PSMB1 and PSMD12, CD44, DNA damage inducible beta GADD45B, p300/CBP associated factor (PCAF), SET and MYD domain containing (SMYD1), and TAF7 RNA polymerase II TATA box binding protein (TAF7). A number of transcripts were repressed by proteasome inhibition, including breast cancer 1 (BRCA1), jumonji containing 2D (JMJD2D) and jumonji AT rich interactive domain 2 (JARID2) (Figure 3B, Table 3-1).

Figure 3Figure 3Figure 3Figure 3Figure 3
Proteasome inhibition results in broad changes in gene expression. (A) Venn diagrams showing the number of genes up- or down-regulated by proteasome inhibitor alone and in common with either dexamethasone (MD) or 17β-estradiol (ME2). (B) Cluster ...

A total of 913 transcripts were changed by MG and DEX, 487 up-regulated and 426 down-regulated. Key transcripts regulated in this manner are heat shock protein 70 (HSPA6), Kruppel-like factor 6 (KLF6) also known as core promoter element binding protein (COPEB), activating transcription factor 3 (ATF3), growth differentiation factor 15 (GDF15) also known as placental bone morphogenetic protein (PLAB) or nonsteroidal anti-inflammatory drug-activated gene (NAG-1), myeloid/lympoid or mixed lineage leukemia translocation 11 (AF1Q), GTP binding protein or gene expressed in mitogen stimulated T cells (GEM), and DNA damage inducible transcript 1 (GADD45A) (Figure 3C, Table 3-2). Conversely, some transcripts were repressed by MG plus DEX, including chloride intracellular channel 3 (CLIC3), lin-28 homolog of C elegans (lin 28), interferon induced transmembrane protein 2 (IFITM2), SOX 13, nuclear receptor type 1 (COUPTF11), S100 calcium binding protein A4 (S100A4) and transcription elongation factor A (SII) 2 and 3 (TCEA2 and 3). The microarray analyses were confirmed by RT-PCR of a representative genes, HSPA6 and S100A4 (Figure 3C). Treatment with proteasome inhibitor alone induced HSPA6 gene expression at both 2 hr and 24 hr, indicating HSPA6 is a direct target of proteasome inhibitor. Conversely, treatment with proteasome inhibitor results in the repression of S100A4 transcript at 24 hr, but not at 2 hr suggesting the effect of inhibitor on S100A4 gene is mediated in the long term (Figure 3C). To verify the effect of the inhibitor we demonstrated that treatment with epoxomicin increased expression of HSPA6 (Figure S1-D).

A total of 618 genes were altered by MG and E2, 290 were up-regulated and 328 down-regulated. The key transcripts activated by MG and E2 were HSPA6, KLF6/COPEB, ATF3, GDF15, AF1Q and GADD45A. Some transcripts were repressed by MG and E2, including CLIC3, lin 28, IFITM2, SOX 13, NR2F1 and 2, S100A4, TCEA2 and 3, zinc finger protein 467 (ZNF467), solute carrier family 40 (SLC40A1) and prolactin induced protein (PIP). Most these genes are also changed by MG and DEX; however, a number were specifically changed after treating with MG plus E2, including dehydrogenase/reductase (SDR family) member 10 (DHRS10), DNA damage inducible transcript 3 (DDIT3), DEAD (Asp-Glu-Ala-Asp) box polypeptide 43 (DDX43) and interleukin 8 (IL8) (Figure 3 D, Table 3-3). The microarray analyses were confirmed by RT-PCR of representative genes, ATF3 and Lin 28 (Figure 3D). Treatment with proteasome inhibitor alone induces ATF3 gene expression at both time points, indicating ATF3 is a direct target of proteasome inhibitor, but not E2. Treatment with proteasome inhibitor leads to decreased expression of Lin28 at 2 hr and 24 hr (Figure 3D). E2 alone, independent of inhibitor, led to a diminution in Lin 28 after 2 hr treatment (a result to be further investigated). For each category of genes the effect of the proteasome inhibitor on gene expression was verified by gene expression after treating with epoxomicin (Figure S3-D).

Approximately 1700 genes were common between MG plus DEX and MG plus E2, 699 transcripts up-regulated and 988 repressed, whereas 10 genes were differentially expressed. Common activated genes include CRYAB, NDRG1, GADD45A, DUSP1, KLF6/COPEB, HSPA6, GEM, TUBB2A, ATF3 and AF1Q; and examples of genes repressed include S100A8, COL12A1, CLIC3, AMIGO2, NR2F1, NCAM2, cAMP responsive element binding protein 3-like 4 (CREB3L4/AIBZIP), PIP, CXXC finger 4 (CXXC4/IDAX), SOX13 and lin 28 (Figure 3E, Table 3-4). The microarray analyses were confirmed by RT-PCR of a representative gene, CRYAB (Figure 3E). Treatment with proteasome inhibitor alone induces CRYAB gene expression at both 2 hr and 24 hr, indicating CRYAB is a direct target of proteasome inhibitor, but not DEX; however, treatment with DEX and MG132 highly induced CRYAB (Figure 3-MD). In contrast to DEX, treatment with E2 and inhibitor did not affect CRYAB expression (Figure 3E- ME2). In addition, prolactin-induced protein (another gene in this class) is repressed by inhibitor alone and with hormone (Figure 3E-PIP). The observation that CRYAB expression increases after treatment with proteasome inhibitor was confirmed after treatment with another inhibitor, epoxomicin (Figure S3-E).

Proteasome inhibition modulates transcripts encoding RNA polymerase II transcriptional regulators

To better understand the biological and molecular functions of the transcripts regulated after proteasome inhibition and hormone, we performed gene ontology classification. The analysis revealed that many of the transcripts changed after proteasome inhibition and hormone are characteristic of genes involved in transcription and transcription factor activity (Figure 4). Apart from transcripts encoding transcription factors, such as ATF3 and zinc finger-binding proteins, two prominent classes of transcripts emerged from further analysis. These included transcripts encoding factors that drive RNA polymerase II transcription and modify chromatin. Among transcripts changed by proteasome inhibitor that regulate RNA polymerase II transcription included PTEFb complex Cdk9 and cyclin K that regulates RNA polymerase carboxy-terminus phosphorylation. We note that treatment with DEX alone repressed CDK9 transcript, but treatment with MG and DEX increased Cdk9, whereas the treatment with E2 increased CDK9 transcript (2-fold) and MG plus E2 decreased Cdk9 transcript (Figure 4C). Transcripts encoding carboxy terminus phosphatase (CTD) including SSU72, CTDSP1 and CTDSPL were repressed by proteasome inhibition except CTDP1 (FCP1), which increased with proteasome inhibition (Figure 4C, Table 4-1).

Figure 4Figure 4Figure 4
Functional classification of genes changed after treatment of MCF-7 with proteasome inhibitor and dexamethasone (A) or 17β-estradiol (B), X- axis represents functional category shown on table, Y-axis represents percent of total genes in the category ...

Proteasome inhibition had significant effects on other RNA polymerase II regulators. Transcripts that encode the TATA box binding protein (TBP)-associated factors, TAF10 and TAF1B (TAFI63) were repressed by proteasome inhibition, whereas TAF1A, TAF2, TAF7, TAF9 and TAF 13 increased with proteasome inhibition (Figure 4C, Table 4-2). Transcripts that encode mediator subunits, MED10, MED28 and MED6 increased with proteasome inhibition (Figure 4C, Table 4-1). Genes that regulate the elongation rate of RNA polymerase II, RNA polymerase II elongation factor 2 (ELL2), which is also a GR target,, ELL and cell division cycle 73 (CDC73/PAF1) increased, whereas RNA polymerase II elongation factor-like 3 (ELL3) decreased.

Further analysis showed that proteasome inhibition had a substantial effect on transcripts encoding transcription elongation and translation initiation factors (Figure 4D, Table 4B). Transcription elongation factor A (SII) (TCEA) factors were all repressed by proteasome inhibition. MG plus DEX significantly decreased transcription elongation factor A (SII) like 1 (TCEAL1) and TCEAL4, while TCEA1 remained unchanged (Figure 4D). Proteasome inhibition alone or in addition to either dexamethasone or E2 significantly repressed TCEA2, TCEA3, TCEAL8 and TCEAL5. A number of transcripts encoding eukaryotic translation factors were significantly increased by proteasome inhibition including EIF1, EIF1B, EIF2A and EIF2C3 (Argonaute3), whereas those transcripts that encode negative regulators of the translation factors, such as eukaryotic translation initiation factor 2- alpha kinase (EIF2AK2) an interferon induced kinase that phosphorylates EIF2A and eukaryotic translation initiation factor 4E binding protein 2 (EIF4EBP2) a protein that binds to EIFE to inhibit protein translation, are repressed by proteasome inhibition (Figure 4D, Table 4-2).

Proteasome inhibition modulates expression of chromatin regulators including histone and DNA modifying enzymes

Proteasome inhibition alters transcripts encoding enzymes or factors that modify DNA and histones. Nuclear receptors utilize a number of coregulators to modulate transcription. To date the best characterized histone modifying enzymes are those that mediate histone acetylation (HATs) and de-acetylation (HDACs), activating and repressing transcription, respectively. Proteasome inhibition increased some common nuclear receptor coactivators including NCOA6 also known as activating signal cointegrator (ASC2), NCOA7 also known as estrogen receptor activation protein 140 (ERAP140), thyroid interacting protein 4 (TRIP4) also known as ASC-1 and TRIP12. Conversely transcripts encoding co-repressors were decreased by proteasome inhibition including nuclear receptor co-repressor 2 (NCOR2 or SMRT) and histone deacetylases, HDAC1 and 8, although HDAC3 transcript was significantly increased when proteasome is inhibited in the presence of dexamethasone. Most strikingly, sin 3A associated protein (SAP30) is induced by DEX, but inhibited by MG alone and in the presence of DEX (Figure 4E and Table 4-3).

Apart from acetylation and deacetylation of histone N-terminal tails, another modification gaining interest with respect to gene regulation by a nuclear receptor is histone methylation. Examination of transcripts changes by proteasome inhibition revealed a number of histone methyltransferases and recently discovered demethylases were altered by proteasome inhibition. Transcripts encoding histone methyltransferases particularly associated with histone H3-Lysine 4 were increased by proteasome inhibition, including MLL and MLL translocation partners namely, MLLT2/AFF1/AF4/FMR2, MLLT11/AF1Q, SETD1A and SMYD1. Transcripts encoding other MLL translocation partners, MLLT3/AF9 and MLLT1/ENL decreased (Figure 4F and Table 4-4). Transcripts encoding histone methyltransferases specific for histone H3-lysine 9, euchromatin-lysine N-methyltransferase 1 (EHMT1 or G9 like protein, GLP) and EHMT2 (G9a), and the testis specific H3K9 methyltransferase SUV39H2 decreased, whereas the KAP-1 associating SET domain bifurcated 1 also known as ERG associated protein (ESET) increased after proteasome inhibition. Of note, EHMT1 increased by DEX, but repressed by MG and DEX, whereas SETDB1 is repressed by E2, but increased after MG and E2. In addition proteasome inhibition alters transcripts encoding methyltransferases targeting histone H3 lysine 36. These include Wolf-Hirschhorn syndrome candidate 1 (WHSC1) also known as multiple myeloma SET domain protein (MMSET) or nuclear SET domain-containing protein 2 (NSD2), Wolf-Hirschhorn syndrome candidate 1-like 1 (WHSC1L1/NSD3) and SMYD2 which decreased by proteasome inhibition. In a number of cases the hormone component is involved, for example SMYD2 increased by hormone but decreased by proteasome inhibition. Transcripts encoding recently identified Jumonji-containing histone demethylases were also affected by proteasome inhibition including JARID2, JMJD2D and RBP2, which were repressed by proteasome inhibition whereas JMD1A transcript increased (Figure 4F and Table 4-4).

Protein arginine methylation has an important role in hormone regulated transcription [39] Proteasome inhibition alters expression of protein arginine methyltransferases (PRMT), including PRMT3 a ribosomal protein arginine methyltransferase that regulates ribosome biosynthesis, PRMT8 a membrane-associated and tissue-specific arginine methyltransferase and PRMT6 a methyltransferase shown to possess auto-methylation activity and methylated the non-histone chromatin protein HMGA1 (Figure 4F, Table 4-4). Finally DNA methyltransferase, DNMT1, DNMT3B and 3L were significantly repressed by proteasome inhibition (Figure 4F, Table 4-4).

Among chromatin factors that are affected by proteasome inhibition were transcripts encoding various histone proteins. The major histone transcripts affected were those encoding histone H2A and H2B family members. These family members were all decreased by proteasome inhibition (Figure 4G, Table 4-5). Transcripts for histone H2AFL, H2AFY2, H2AFA, H2BFF, H2BFD, H2BFH, H2BFQ, H2BFE, H2BFB and H2BFK were repressed 2- to 4-fold by proteasome inhibition. Interestingly, histone H2AFY2 increased by E2 was inhibited by proteasome inhibition. Histone H2BFQ is highly down-regulated by E2, but this effect is reversed by proteasome inhibition. Variants of histone H3, H3FT and H3F1 were also down-regulated by proteasome inhibition. Histone H1F4 (H1.2), which is predicted to maintain low methylation state, was repressed up to 4-fold. Histone H1F0 (H1.0) was up-regulated by DEX, but repressed by MG and DEX (Figure 4G, Table 4-5).

Effect of proteasome inhibition on transcription of developmental genes, proteasome subunits and stress proteins

Because there were very significant changes in transcripts encoding MLL and MLL translocation partners, we investigated whether transcripts encoding clustered homeobox (Hox) genes were affected by proteasome inhibition. Knockout experiments have previously identified Hox genes as targets of MLL. Of the transcripts encoding HOX genes, HOXA1 which was down regulated by hormone alone (DEX or E2) was highly up-regulated by MG and either hormone. Other Hox genes were down-regulated by proteasome inhibition including those of HOXC8, HOXA10, HOX D9, B2, C13 and C9 (Figure 5A, Table 5-1).

Figure 5
Proteasome inhibition alters transcription of developmental, proteasome subunits and stress response genes. (A). Cluster analysis of genes encoding developmental genes. (B) Cluster analysis of genes encoding proteasome subunits. (C) Cluster analysis of ...

Analysis of transcripts regulated by proteasome inhibition showed an increase in transcripts encoding lin-7 homolog A and C (Lin7A and C), but a decrease in Lin 7B was seen. Lin-28 was highly repressed by proteasome inhibition, where as sel-1 suppressor of lin-12-like increased by proteasome inhibition (Figure 5A, Table 5-1).

Among other targets of the proteasome are the proteasome subunits themselves. Our transcript profiling analysis shows that proteasome inhibition up-regulated 19S proteasome ATPase subunits PSMC1, -4, -5, and -6, but not PSMC2 and non-ATPase subunits, PSMD1, -2, -8, -9, -11, -12 and -14. Proteasome inhibition also increases transcripts encoding the 20S subunits, alpha subunits PSMA1, -3, -4, -5, and -7 and beta subunits 1, 2,3,4,5, 6 and 7. On the other hand, proteasome inhibition repressed transcripts encoding antigen presenting, immunoassembly proteasomes PSMB10, PSME1 and -2 (Figure 5B, Table 5-2).

Previous studies have shown that proteasome inhibition increased stress response factors, particularly heat shock proteins. Proteasome inhibition induced a global increase heat shock protein transcripts, including hsp90, -70 and -40 families. These changes are among the most pronounced changes of proteasome inhibition; for example, proteasome inhibition induced HSPA6 transcript (Hsp70B) up to 40-fold and DNAJB1 (Hsp40, subfamily B) up to 14-fold, whereas another member of this family DNAJC19 (Hsp40, subfamily C) is repressed (Figure 5C, Table 5-3).

Proteasome inhibition affects transcription of genes associated in the pathogenesis of neurodegenerative diseases, leukemia, multiple myeloma, breast/prostate cancer and HIV/AIDS

Proteasome inhibitors, such as bortezomib, are currently in clinical trials as potential therapeutic agents. In particular, protein inhibitors plus DEX have been used to treat relapsed multiple myeloma. Using a chromosome tool, we aligned the 1697 genes affected in common by MG, MG plus DEX and MG plus E2 to chromosome loci (Figure 6, see also Figure 3A). Examination of chromosome loci showed specific clustering of genes or hot spots on chromosomes 1, 6, 11, 19 and on the X chromosome. Genes clustered on the hot spots marked in a black line on the specific chromosome are associated with leukemia, Kaposi sarcoma, severe combined immunodeficiency, non-Hodgkin's B-cell lymphoma, acute myeloid leukemia, breast cancer and Sjogren syndrome antigen among other diseases. Genes clustered in chromosome 19 encode a number of zinc finger proteins. This observation is interesting, considering that 50% of all human KRAB-ZNF genes are located on chromosome 19 and recent data shows that the specific domain harboring these genes is heterochromatic and marked by elevated binding of heterochromatin protein 1 (HP1) [40].

Figure 6
Proteasome inhibition affects genes at specific chromosome loci. A chromogram showing genes affected by exclusively by proteasome inhibition (red up-regulated; green down-regulated). Proposed hot spots within chromosomes are indicated by a black line. ...

Discussion

A number of studies indicated that inhibiting proteasome degradation increased transcriptional activity of some, but not all nuclear receptors suggesting a receptor specific effect of proteasome inhibition [19,22,24,41,42]. Specifically blocking proteasome degradation with the proteasome inhibitor MG132 elevated GR, but diminished ER-mediated gene activation, suggesting that proteasome degradation is required for transactivation at least by the estrogen receptor [19,22-24]. However, these studies were based on either reporter gene constructs or limited individual receptor target genes [19,22-24]. We have taken a genomic approach to show that the requirement for proteasome activity is gene specific rather than receptor specific. Our data provides new information indicating that proteasome inhibition has both synergistic and antagonistic effects on GR and ER-mediated gene expression. Proteasome inhibition enhances GR-mediated gene expression of endogenous targets (S100P), but other known GR targets like galanin, BCL6 and TGFB3 are repressed [35-38].

We confirm previous reports that proteasome inhibition decreases E2-mediated progesterone receptor gene expression, but also show that E2 targets, such as DDX10, are synergistically induced by E2 and a proteasome inhibitor, whereas TXNIP, SOX13 and IFIT2 were synergistically repressed.

Gene expression profiles observed in this study are similar to those reported by others in MCF-7 cells treated with E2 [27-29]. With respect to the GR response, the gene profiling signature from the GR/ER positive MCF-7 cell line is similar to that observed in other cell lines in response to dexamethasone [35-38].

Present analysis suggests some negative cross-talk between GR and ER [25]. A number of gene transcripts are differentially regulated by GR and ER, when proteasome activity is inhibited. For example, the gene NDRG1 is activated by DEX and MG, but repressed by E2 and inhibitor. A similar trend follows for KLF6, SMYD2 and S100A8 genes. NDRG1 is markedly expressed in the placenta and it is the most ubiquitous member of the NDRG family genes (NDRG 1-4) [43]. Over expression of NDRG1 in colon, breast or prostate cell lines decreases proliferation rate, enhances differentiation and suppresses the metastatic potency of the tumor [44,45]. KLF6 or core promoter element binding protein is a Krüppel family of C2H2-type zinc finger protein involved in regulation and maintenance of the basal expression of TATA box-less genes. It is highly expressed in the placenta [46]. KLF6 is an inhibitor of cell proliferation, suggesting a role of KLF6 as a potential tumor suppressor [47]. SYMD2 has a role in cell proliferation since it was shown recently to methylate p53 [45,47,48]; S100 A8 is strongly up-regulated only in ductal carcinoma in situ [49]. For these genes, repression by E2 favors cell proliferation, whereas activation by DEX inhibits proliferation. It is of particular interest that some the genes differentially expressed after proteasome inhibition and hormone treatment are highly expressed in various type of breast tumors [50-52]. Proteasome inhibitors are currently applied in the therapy of hormone responsive cancers; however, the negative crosstalk between GR and ER can influence the outcome of therapeutic application.

A novel finding from the genomic profiling is the regulation of transcripts encoding genes for RNA polymerase II transcriptional regulators (transcription elongation/translation initiation factors) and chromatin modifying enzymes (DNA and histone methyltransferases/histone demethylases/acetyltransferases/deacetylases). The profound impact of proteasome inhibition on transcriptional regulators suggests that proteasome activity can regulate transcription at multiple steps, initiation, elongation and even mRNA processing. Key molecules, such as TAFs, mediator subunits and KLF6 that impact transcriptional initiation/activation and confer gene specific activation, are altered by proteasome inhibition. TAFs and KLF6 play a role in regulation of TATA less promoters [46,53]. Perhaps these factors can account for differential regulation of receptor target genes after proteasome inhibition.

Additionally, proteasome inhibition alters transcripts encoding RNA polymerase II CTD phosphatases and transcriptional elongation factors (TCEA (SII), ELL). These factors can enhance or repress RNA Pol II elongation rate, supporting a role of the proteasome in transcriptional elongation. We have reported recently that proteasome activity at least in part regulates transcription by modulating the phosphorylation of RNA polymerase II, a hallmark of the elongating polymerase [31]. Other gene transcripts, such as transcriptional translation initiation factors and genes regulated by micro-RNAs (Lin 28, Lin 7), suggest proteasome activity might be required in the regulation of mRNA processing and translation [54-56].

DNA methylation and histone modifications have crucial roles in the control of gene activity. Changes in expression of enzymes that modify DNA or histones after proteasome inhibition can impact on gene expression. Proteasome inhibition alters expression of transcripts that encode DNA methyltransferases (DNMT1, 3L and 3B). DNA methylation is normally associated with gene silencing, but also provides multiple layers of gene control; for example, tissue specific gene expression. Proteasome activity may impact on genes tightly regulated by DNA methylation: for example, the melanoma antigen (MAGE) family of cancer testis genes and the S100 calcium binding protein A4 (S100A4), which is over-expressed in colon cancer, are tightly regulated by DNA methylation and in this study they are altered by proteasome inhibition [57,58](Figure 1C and 1 F).

Another level of transcriptional regulation by proteasome activity can be achieved by modification of chromatin architecture. Several gene transcripts encoding histone proteins and histone modifying enzymes are changed after proteasome inhibition. Histones are no longer considered to be simple DNA-packaging proteins: they are recognized as dynamic regulators of chromatin architecture and gene transcription. In this study we found changes in transcripts encoding specific histones and histone variants, providing an opportunity for proteasome activity in the regulation of chromatin architecture. We demonstrated that the H1.2 (H1F4) isoform, which is proposed to maintain low DNA methylation state, is significantly repressed by proteasome inhibition. In mammals, histone H1 is expressed in at least 8 isoforms. Though we do not know the direct effect of this isoform on receptor mediated-transcription, we have previously showed that prolonged DEX treatment effectively dephosphorylated the H1.3, H1.4, and H1.5 isoforms to repress MMTV transcription indicating that histone H1 isoforms directly influence the transcriptional activation/repression of specific genes [59].

Proteasome inhibition results in changes in expression of transcripts encoding a number of histone modifying enzymes, especially those resulting in arginine and lysine methylation. Transcripts encoding histone methyltransferases targeting histone H3-K4 and H3-K36 previously associated with active chromatin are significantly changed by proteasome inhibition. The changes in histone modifying enzymes, methyltransferases and demethylases seen after proteasome inhibition offer an exciting mechanism to explain differential regulation of hormone mediated gene expression. Indeed, recent studies have shown that specific histone methyltransferases can regulate hormone response and impose gene specific functions [60].

Apart from transcriptional regulation, a number of the transcripts encoding histone modifying enzymes are particularly interesting because of their established or putative roles in human diseases. Proteasome inhibition alone or in the presence of either DEX or E2 leads to an increase mixed lineage leukemia (MLL) specific methyltransferase for histone H3 Lys4 (H3K4). In addition a number of MLL translocation partners, for example RNA polymerase II elongation factor 2 (ELL2) and AF1Q, are increased when cells are treated with proteasome inhibitor. Mixed-lineage leukemia 1 (MLL1) gene is disrupted by chromosomal translocation in acute leukemia and is a master regulator of Hox genes [61], which have been recognized as oncogenes in leukemia. Additionally, the oncogenic potential of Hox genes is implicated in various cancers [62]. For example, HOX A1 is up-regulated in cervical cancer and we found that it is altered by proteasome inhibition and estradiol [63]. The Hox cluster, C10, -11 and -13 are implicated in metastatic melanoma [64]. Hox C8 is over expressed in prostate cancer [65]. Interestingly, proteasome inhibition decreases most of Hox gene expression perhaps offering a clue on how proteasome inhibitors act as a therapeutic application in leukemia. Our studies reveal an interesting avenue to pursue as both the proteasome and steroid hormone receptors are targets for therapy in the treatment of leukemia [66,67].

Disruption of MLL function by translocation is recently implicated in the promiscuous regulation of cell cycle regulators (cyclin dependent kinases and kinase inhibitors) and a cluster of miRNAs involved in cancer, supporting a role of MLL in tumor formation and suppression [61,68,69]. Our cluster analysis after proteasome inhibition reveals a set of developmental genes that are regulated by miRNAs are altered by proteasome inhibition. Lin 28 encodes a RNA binding protein of which functional mutations results in abnormal development of various cell lineages [70]. Lin 28 is regulatory target of mir-125 cluster which function in neuronal development [56]. Lin7A, Lin7B, and Lin7C, which each encodes a protein that is required for generation and maintenance of neuroepithelial cell junctions is a proposed target for mir22 and mir365 (http://microrna.sanger.ac.uk/).

The genes encoding Wolf-Hirschhorn syndrome candidate 1 (WHSC1) also known as multiple myeloma SET domain (MMSET) or nuclear receptor-binding SET domain-containing protein 2 (NSD2) and Wolf-Hirschhorn syndrome candidate 1-like 1 (WHSC1L1) also known as NSD3, the putative histone methyltransferase targeting both histone H3-K36 and H4-K20 are down-regulated in the cells treated with proteasome inhibitor and hormone. Translocations between multiple myeloma SET domain (MMSET) and fibroblast growth factor receptor 3 (FGFR3) result in multiple myeloma [71]. Additionally a set of recently discovered histone demethylases in the Jumonji and Jarid family are altered by proteasome inhibition. These changes in molecules that impact on multiple myeloma are interesting especially since in clinical trials proteasome inhibitors are used to treat multiple myeloma patients with glucocorticoid resistance who have undergone relapse, where treatment with dexamethasone and proteasome inhibitor restores clinical outcome [72].

Finally, given the potential of proteasome inhibitors in antiviral therapy, an interesting candidate in this regard is the estrogen-dependent gene stromal cell-derived factor (SDF-1 or CXCL12) a ligand of CCRX4 chemokine receptor, which is involved in diseases including AIDS and cancer cell metastasis [73,74]. Other molecules involved in HIV transcription are altered by proteasome inhibition include NR2F1, the proteasome subunit PSMC4 which interacts with HIV TAT and the protein arginine methyltransferase PRMT6 which methylates and modulates TAT-mediated transactivation [75-77].

Proteasome inhibition modulates transcripts encoding genes involved in protein folding, cell migration, cell cycle regulation, apoptosis, inflammatory responses, cell adhesion, antigen presentation and ion transport to name a few. Importantly, our genome-wide transcript profiling analysis and chromosome mapping shows that proteasome inhibition impacts on expression of many genes involved in the pathogenesis of various human diseases including many cancers, HIV/AIDs and neurodegenerative disorders, Alzheimer's, Parkinson's and Huntington's [17,78]. Many proteasome targets, such p53, MDM2 and ER, play critical roles in cell growth and proliferation and can contribute to survival of tumor cells. Not surprisingly, inhibitors of the proteasome, such as Velcade/Bortezomib have been showed to inhibit tumor growth in clinical trials of multiple myeloma, breast, pancreatic, lung, and ovarian cancers [79,80]. The precise mechanisms of how proteasome inhibitors, such as Velcade, work as anti-tumor agents are unknown. The predominant view attributes the outcome of the therapy to the degradation of specific tumor suppressors or cell cycle regulators or in-activation of the NFkB due to its anti-apoptotic activity [81]. Our analysis of proteasome/hormone receptor mediated gene transcription suggests alternative pathways that may provide a mechanistic explanation for therapeutic outcomes of proteasome inhibitors. Our studies imply that proteasome activity modulates NR function via changes in chromatin enzymes, there by implicating the proteasome in epigenetic contribution to human disease. Presently, there is evidence to show that disruption in the balance of epigenetic networks can cause pathological disease states, such as leukemia and inhibitors for chromatin modifying enzymes, offer future prospects for epigenetic therapy [82,83]. Proteasome inhibitors join other classes of therapy, such as DNA demethylating agents and HDACs that change epigenetic marks.

Supplementary Material

supplemental figures

supplemental legends

Acknowledgments

We are deeply grateful to Dr. Pierre Bushel (Biostatistics Branch, NIEHS) for providing help with the statistical analysis and re-writing the methods section to answer the reviewers concerns. We thank Wendy Jefferson and Sylvia Hewitt for helpful comments in organizing the paper.

This research was supported by the Intramural Research Program of NIH and NIEHS.

References

1. Mangelsdorf DJ, Thummel C, Beato M, et al. The nuclear receptor superfamily: the second decade. Cell. 1995;83(6):835–839. [PubMed]
2. Yamamoto KR. Steroid receptor regulated transcription of specific genes and gene networks. Annu Rev Genet. 1985;19:209–252. [PubMed]
3. Kinyamu HK, Archer TK. Modifying chromatin to permit steroid hormone receptor-dependent transcription. Biochim Biophys Acta. 2004;1677(1-3):30–45. [PubMed]
4. Kishimoto M, Fujiki R, Takezawa S, et al. Nuclear receptor mediated gene regulation through chromatin remodeling and histone modifications. Endocr J. 2006;53(2):157–172. [PubMed]
5. Kinyamu HK, Chen J, Archer TK. Linking the ubiquitin-proteasome pathway to chromatin remodeling/modification by nuclear receptors. J Mol Endocrinol. 2005;34(2):281–297. [PubMed]
6. Nawaz Z, O'Malley BW. Urban renewal in the nucleus: is protein turnover by proteasomes absolutely required for nuclear receptor-regulated transcription? Mol Endocrinol. 2004;18(3):493–499. [PubMed]
7. Hoang T, Fenne IS, Cook C, et al. cAMP-dependent protein kinase regulates ubiquitin-proteasome-mediated degradation and subcellular localization of the nuclear receptor coactivator GRIP1. J Biol Chem. 2004;279(47):49120–49130. [PubMed]
8. Li X, Lonard DM, Jung SY, et al. The SRC-3/AIB1 coactivator is degraded in a ubiquitin- and ATP-independent manner by the REGgamma proteasome. Cell. 2006;124(2):381–392. [PubMed]
9. Yan F, Gao X, Lonard DM, Nawaz Z. Specific ubiquitin-conjugating enzymes promote degradation of specific nuclear receptor coactivators. Mol Endocrinol. 2003;17(7):1315–1331. [PubMed]
10. Baker SP, Grant PA. The proteasome: not just degrading anymore. Cell. 2005;123(3):361–363. [PubMed]
11. Collins GA, Tansey WP. The proteasome: a utility tool for transcription? Curr Opin Genet Dev. 2006;16(2):197–202. [PubMed]
12. Goldberg AL. Protein degradation and protection against misfolded or damaged proteins. Nature. 2003;426(6968):895–899. [PubMed]
13. Glickman MH, Ciechanover A. The ubiquitin-proteasome proteolytic pathway: destruction for the sake of construction. Physiol Rev. 2002;82(2):373–428. [PubMed]
14. Pickart CM. Back to the future with ubiquitin. Cell. 2004;116(2):181–190. [PubMed]
15. Fang S, Weissman AM. A field guide to ubiquitylation. Cell Mol Life Sci. 2004;61(13):1546–1561. [PubMed]
16. Ciechanover A. Intracellular Protein Degradation: From a Vague Idea thru the Lysosome and the Ubiquitin-Proteasome System and onto. Human Diseases and Drug Targeting Hematology. Am Soc Hematol Educ Program. 2006:1–12. [PubMed]
17. Schwartz AL, Ciechanover A. The ubiquitin-proteasome pathway and pathogenesis of human diseases. Annu Rev Med. 1999;50:57–74. [PubMed]
18. Lipford JR, Smith GT, Chi Y, Deshaies RJ. A putative stimulatory role for activator turnover in gene expression. Nature. 2005;438(7064):113–116. [PubMed]
19. Reid G, Hubner MR, Metivier R, et al. Cyclic, proteasome-mediated turnover of unliganded and liganded ERalpha on responsive promoters is an integral feature of estrogen signaling. Mol Cell. 2003;11(3):695–707. [PubMed]
20. Lee JW, Ryan F, Swaffield JC, Johnston SA, Moore DD. Interaction of thyroid-hormone receptor with a conserved transcriptional mediator. Nature. 1995;374(6517):91–94. [PubMed]
21. Zhang H, Sun L, Liang J, et al. The catalytic subunit of the proteasome is engaged in the entire process of estrogen receptor-regulated transcription. Embo J. 2006;25(18):4223–4233. [PubMed]
22. Deroo BJ, Rentsch C, Sampath S, Young J, DeFranco DB, Archer TK. Proteasomal inhibition enhances glucocorticoid receptor transactivation and alters its subnuclear trafficking. Mol Cell Biol. 2002;22(12):4113–4123. [PMC free article] [PubMed]
23. Wallace AD, Cidlowski JA. Proteasome-mediated glucocorticoid receptor degradation restricts transcriptional signaling by glucocorticoids. J Biol Chem. 2001;276(46):42714–42721. [PubMed]
24. Lonard DM, Nawaz Z, Smith CL, O'Malley BW. The 26S proteasome is required for estrogen receptor-alpha and coactivator turnover and for efficient estrogen receptor-alpha transactivation. Mol Cell. 2000;5(6):939–948. [PubMed]
25. Kinyamu HK, Archer TK. Estrogen receptor-dependent proteasomal degradation of the glucocorticoid receptor is coupled to an increase in mdm2 protein expression. Mol Cell Biol. 2003;23(16):5867–5881. [PMC free article] [PubMed]
26. Fryer CJ, Nordeen SK, Archer TK. Antiprogestins mediate differential effects on glucocorticoid receptor remodeling of chromatin structure. J Biol Chem. 1998;273(2):1175–1183. [PubMed]
27. Frasor J, Danes JM, Komm B, Chang KC, Lyttle CR, Katzenellenbogen BS. Profiling of estrogen up- and down-regulated gene expression in human breast cancer cells: insights into gene networks and pathways underlying estrogenic control of proliferation and cell phenotype. Endocrinology. 2003;144(10):4562–4574. [PubMed]
28. Lin CY, Strom A, Vega VB, et al. Discovery of estrogen receptor alpha target genes and response elements in breast tumor cells. Genome Biol. 2004;5(9):R66. [PMC free article] [PubMed]
29. Lobenhofer EK, Bennett L, Cable PL, Li L, Bushel PR, Afshari CA. Regulation of DNA replication fork genes by 17beta-estradiol. Mol Endocrinol. 2002;16(6):1215–1229. [PubMed]
30. Hebbar PB, Archer TK. Chromatin-dependent cooperativity between site-specific transcription factors in vivo. J Biol Chem. 2007;282(11):8284–8291. [PMC free article] [PubMed]
31. Kinyamu HK, Archer TK. Proteasome Activity Modulates Chromatin Modifications and RNA Polymerase II Phosphorylation To Enhance Glucocorticoid Receptor-Mediated Transcription. Mol Cell Biol. 2007;27(13):4891–4904. [PMC free article] [PubMed]
32. Weng L, Dai H, Zhan Y, He Y, Stepaniants SB, Bassett DE. Rosetta error model for gene expression analysis. Bioinformatics. 2006;22(9):1111–1121. [PubMed]
33. Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002;30(1):207–210. [PMC free article] [PubMed]
34. Rhen T, Grissom S, Afshari C, Cidlowski JA. Dexamethasone blocks the rapid biological effects of 17beta-estradiol in the rat uterus without antagonizing its global genomic actions. Faseb J. 2003;17(13):1849–1870. [PubMed]
35. Rogatsky I, Wang JC, Derynck MK, et al. Target-specific utilization of transcriptional regulatory surfaces by the glucocorticoid receptor. Proc Natl Acad Sci U S A. 2003;100(24):13845–13850. [PubMed]
36. Stojadinovic O, Lee B, Vouthounis C, et al. Novel genomic effects of glucocorticoids in epidermal keratinocytes: inhibition of apoptosis, interferon-gamma pathway, and wound healing along with promotion of terminal differentiation. J Biol Chem. 2007;282(6):4021–4034. [PubMed]
37. Wan Y, Nordeen SK. Overlapping but distinct gene regulation profiles by glucocorticoids and progestins in human breast cancer cells. Mol Endocrinol. 2002;16(6):1204–1214. [PubMed]
38. Wang JC, Derynck MK, Nonaka DF, Khodabakhsh DB, Haqq C, Yamamoto KR. Chromatin immunoprecipitation (ChIP) scanning identifies primary glucocorticoid receptor target genes. Proc Natl Acad Sci U S A. 2004;101(44):15603–15608. [PubMed]
39. Lee DY, Teyssier C, Strahl BD, Stallcup MR. Role of protein methylation in regulation of transcription. Endocr Rev. 2005;26(2):147–170. [PubMed]
40. Vogel MJ, Guelen L, de Wit E, et al. Human heterochromatin proteins form large domains containing KRAB-ZNF genes. Genome Res. 2006;16(12):1493–1504. [PubMed]
41. Blanquart C, Barbier O, Fruchart JC, Staels B, Glineur C. Peroxisome proliferator-activated receptor alpha (PPARalpha) turnover by the ubiquitin-proteasome system controls the ligand-induced expression level of its target genes. J Biol Chem. 2002;277(40):37254–37259. [PubMed]
42. Lin HK, Altuwaijri S, Lin WJ, Kan PY, Collins LL, Chang C. Proteasome activity is required for androgen receptor transcriptional activity via regulation of androgen receptor nuclear translocation and interaction with coregulators in prostate cancer cells. J Biol Chem. 2002;277(39):36570–36576. [PubMed]
43. Zhou RH, Kokame K, Tsukamoto Y, Yutani C, Kato H, Miyata T. Characterization of the human NDRG gene family: a newly identified member, NDRG4, is specifically expressed in brain and heart. Genomics. 2001;73(1):86–97. [PubMed]
44. Bandyopadhyay S, Pai SK, Gross SC, et al. The Drg-1 gene suppresses tumor metastasis in prostate cancer. Cancer Res. 2003;63(8):1731–1736. [PubMed]
45. Lachat P, Shaw P, Gebhard S, van Belzen N, Chaubert P, Bosman FT. Expression of NDRG1, a differentiation-related gene, in human tissues. Histochem Cell Biol. 2002;118(5):399–408. [PubMed]
46. Koritschoner NP, Bocco JL, Panzetta-Dutari GM, Dumur CI, Flury A, Patrito LC. A novel human zinc finger protein that interacts with the core promoter element of a TATA box-less gene. J Biol Chem. 1997;272(14):9573–9580. [PubMed]
47. Narla G, Friedman SL, Martignetti JA. Kruppel cripples prostate cancer: KLF6 progress and prospects. Am J Pathol. 2003;162(4):1047–1052. [PubMed]
48. Huang J, Perez-Burgos L, Placek BJ, et al. Repression of p53 activity by Smyd2-mediated methylation. Nature. 2006;444(7119):629–632. [PubMed]
49. Carlsson H, Petersson S, Enerback C. Cluster analysis of S100 gene expression and genes correlating to psoriasin (S100A7) expression at different stages of breast cancer development. Int J Oncol. 2005;27(6):1473–1481. [PubMed]
50. Perou CM, Jeffrey SS, van de Rijn M, et al. Distinctive gene expression patterns in human mammary epithelial cells and breast cancers. Proc Natl Acad Sci U S A. 1999;96(16):9212–9217. [PubMed]
51. Perou CM, Sorlie T, Eisen MB, et al. Molecular portraits of human breast tumours. Nature. 2000;406(6797):747–752. [PubMed]
52. Zhu Y, Wang A, Liu MC, et al. Estrogen receptor alpha positive breast tumors and breast cancer cell lines share similarities in their transcriptome data structures. Int J Oncol. 2006;29(6):1581–1589. [PubMed]
53. Wright KJ, Marr MT, 2nd, Tjian R. TAF4 nucleates a core subcomplex of TFIID and mediates activated transcription from a TATA-less promoter. Proc Natl Acad Sci U S A. 2006;103(33):12347–12352. [PubMed]
54. Jiang HY, Jiang L, Wek RC. The eukaryotic initiation factor-2 kinase pathway facilitates differential GADD45a expression in response to environmental stress. J Biol Chem. 2007;282(6):3755–3765. [PubMed]
55. Stavreva DA, Kawasaki M, Dundr M, et al. Potential roles for ubiquitin and the proteasome during ribosome biogenesis. Mol Cell Biol. 2006;26(13):5131–5145. [PMC free article] [PubMed]
56. Wu L, Belasco JG. Micro-RNA regulation of the mammalian lin-28 gene during neuronal differentiation of embryonal carcinoma cells. Mol Cell Biol. 2005;25(21):9198–9208. [PMC free article] [PubMed]
57. De Smet C, Lurquin C, Lethe B, Martelange V, Boon T. DNA methylation is the primary silencing mechanism for a set of germ line- and tumor-specific genes with a CpG-rich promoter. Mol Cell Biol. 1999;19(11):7327–7335. [PMC free article] [PubMed]
58. Nakamura N, Takenaga K. Hypomethylation of the metastasis-associated S100A4 gene correlates with gene activation in human colon adenocarcinoma cell lines. Clin Exp Metastasis. 1998;16(5):471–479. [PubMed]
59. Banks GC, Deterding LJ, Tomer KB, Archer TK. Hormone-mediated dephosphorylation of specific histone H1 isoforms. J Biol Chem. 2001;276(39):36467–36473. [PubMed]
60. Garcia-Bassets I, Kwon YS, Telese F, et al. Histone methylation-dependent mechanisms impose ligand dependency for gene activation by nuclear receptors. Cell. 2007;128(3):505–518. [PMC free article] [PubMed]
61. Hess JL. Mechanisms of transformation by MLL. Crit Rev Eukaryot Gene Expr. 2004;14(4):235–254. [PubMed]
62. Grier DG, Thompson A, Kwasniewska A, McGonigle GJ, Halliday HL, Lappin TR. The pathophysiology of HOX genes and their role in cancer. J Pathol. 2005;205(2):154–171. [PubMed]
63. Shim C, Zhang W, Rhee CH, Lee JH. Profiling of differentially expressed genes in human primary cervical cancer by complementary DNA expression array. Clin Cancer Res. 1998;4(12):3045–3050. [PubMed]
64. Cillo C, Cantile M, Mortarini R, Barba P, Parmiani G, Anichini A. Differential patterns of HOX gene expression are associated with specific integrin and ICAM profiles in clonal populations isolated from a single human melanoma metastasis. Int J Cancer. 1996;66(5):692–697. [PubMed]
65. Alami Y, Castronovo V, Belotti D, Flagiello D, Clausse N. HOXC5 and HOXC8 expression are selectively turned on in human cervical cancer cells compared to normal keratinocytes. Biochem Biophys Res Commun. 1999;257(3):738–745. [PubMed]
66. Harris HA. Estrogen receptor-beta: recent lessons from in vivo studies. Mol Endocrinol. 2007;21(1):1–13. [PubMed]
67. Musto P, Rossini F, Gay F, et al. Efficacy and safety of bortezomib in patients with plasma cell leukemia. Cancer. 2007;109(11):2285–2290. [PubMed]
68. Guenther MG, Jenner RG, Chevalier B, et al. Global and Hox-specific roles for the MLL1 methyltransferase. Proc Natl Acad Sci U S A. 2005;102(24):8603–8608. [PubMed]
69. Milne TA, Hughes CM, Lloyd R, et al. Menin and MLL cooperatively regulate expression of cyclin-dependent kinase inhibitors. Proc Natl Acad Sci U S A. 2005;102(3):749–754. [PubMed]
70. Richards M, Tan SP, Tan JH, Chan WK, Bongso A. The transcriptome profile of human embryonic stem cells as defined by SAGE. Stem Cells. 2004;22(1):51–64. [PubMed]
71. Rasmussen T, Hudlebusch HR, Knudsen LM, Johnsen HE. FGFR3 dysregulation in multiple myeloma: frequency and prognostic relevance. Br J Haematol. 2002;117(3):626–628. [PubMed]
72. Jagannath S, Barlogie B, Berenson J, et al. A phase 2 study of two doses of bortezomib in relapsed or refractory myeloma. Br J Haematol. 2004;127(2):165–172. [PubMed]
73. Vergote D, Butler GS, Ooms M, et al. Proteolytic processing of SDF-1alpha reveals a change in receptor specificity mediating HIV-associated neurodegeneration. Proc Natl Acad Sci U S A. 2006;103(50):19182–19187. [PubMed]
74. Wang Z, Ma Q. beta-Catenin is a promising key factor in the SDF-1/CXCR4 axis on metastasis of pancreatic cancer. Med Hypotheses. 2007 [PubMed]
75. Frankel A, Yadav N, Lee J, Branscombe TL, Clarke S, Bedford MT. The novel human protein arginine N-methyltransferase PRMT6 is a nuclear enzyme displaying unique substrate specificity. J Biol Chem. 2002;277(5):3537–3543. [PubMed]
76. Rohr O, Schwartz C, Hery C, Aunis D, Tardieu M, Schaeffer E. The nuclear receptor chicken ovalbumin upstream promoter transcription factor interacts with HIV-1 Tat and stimulates viral replication in human microglial cells. J Biol Chem. 2000;275(4):2654–2660. [PubMed]
77. Xie B, Invernizzi CF, Richard S, Wainberg MA. Arginine methylation of the human immunodeficiency virus type 1 Tat protein by PRMT6 negatively affects Tat Interactions with both cyclin T1 and the Tat transactivation region. J Virol. 2007;81(8):4226–4234. [PMC free article] [PubMed]
78. Reinstein E, Ciechanover A. Narrative review: protein degradation and human diseases: the ubiquitin connection. Ann Intern Med. 2006;145(9):676–684. [PubMed]
79. Orlowski RZ, Dees EC. The role of the ubiquitination-proteasome pathway in breast cancer: applying drugs that affect the ubiquitin-proteasome pathway to the therapy of breast cancer. Breast Cancer Res. 2003;5(1):1–7. [PMC free article] [PubMed]
80. Richardson PG, Mitsiades C. Bortezomib: proteasome inhibition as an effective anticancer therapy. Future Oncol. 2005;1(2):161–171. [PubMed]
81. Roccaro AM, Hideshima T, Richardson PG, et al. Bortezomib as an antitumor agent. Curr Pharm Biotechnol. 2006;7(6):441–448. [PubMed]
82. Esteller M. Cancer epigenomics: DNA methylomes and histone-modification maps. Nat Rev Genet. 2007;8(4):286–298. [PubMed]
83. Jones PA, Baylin SB. The epigenomics of cancer. Cell. 2007;128(4):683–692. [PMC free article] [PubMed]