RhoH is a member of the Rho family of small GTP-binding proteins that lacks GTPase activity. Since RhoH is constantly bound by GTP, it is thought to be constitutively active and controlled predominantly by changes in quantitative expression. RhoH is produced specifically in haematopoietic cells and aberrant expression has been linked to various forms of leukaemia. Transcription of the RHOH gene is the first level at which the quantitative levels of the RhoH protein are regulated. Previous studies have demonstrated that RHOH gene transcription is initiated by three distinct promoter regions designated P1, P2 and P3 that define the 5′ end of exons 1, 2 and 4 respectively. In the present study we report that the P3 promoter is largely responsible for RHOH gene transcription in the B-lymphocytic cell line Raji. The P3 promoter contains a minimal promoter region and a repressor region extending from −236 to +67 and +68 to +245 respectively, relative to the 5′ end of exon 4. Chromatin immunoprecipitation demonstrated that two AP1 (activator protein 1) sites in the minimal promoter region bind JunD. When JUND is overexpressed, the endogenous RHOH gene is repressed; however, when JUND is inhibited, expression of endogenous RHOH is induced both in the Raji cell line and AML (acute myeloid leukaemia) cells. In the HCL (hairy cell leukaemia) cell line JOK-1, induction of RHOH increases expression of the α isoform of protein kinase C. This downstream target of RHOH is also induced in AML cells by JUND inhibition. Collectively, these data indicate that JunD is an inhibitor of RHOH gene expression.
- differential promoter utilization
- JUND proto-oncogene
- reciprocal expression patterns
- repression of transcription
- RhoH gene promoter
RhoH is a small GTP-binding protein that is specifically but differentially expressed in haematopoietic cells. RhoH lacks the ability to hydrolyse GTP. Therefore, its function is not controlled by cycling between forms bound by GTP or GDP. Rather, the function of RhoH appears to be controlled by phosphorylation and its quantitative level of expression [1–5]. Mature B-lymphocytes and T-lymphocytes exhibit high levels of RHOH gene expression, in contrast with erythroid cells, mature myeloid cells and haematopoietic progenitors where RHOH expression is low. An increasing body of evidence indicates that RhoH plays an important role in a range of normal physiological processes and pathological conditions. The physiological functions of RhoH include the regulation of intracellular signal transduction in mast cells, T-lymphocytes and B-lymphocytes [5–9], integrin activation , leukotriene production , chemotaxis [3,12,13] and haematopoiesis [3,5,6,14]. A role for RhoH in the pathogenesis of malignancy has been implied from the rearrangement of its gene in follicular non-Hodgkin's lymphoma [15,16], nodal marginal zone lymphoma , splenic lymphoma with villous lymphocytes  and multiple myeloma . In addition, the RHOH gene is subject to somatic hypermutation in a diverse set of B-lymphoid proliferative disorders. These include AIDS-related non-Hodgkin's lymphoma , DLBCL (diffuse large B-cell lymphoma) [20,21], B-cell post-transplant lymphoproliferative disorders , primary central nervous system lymphoma , MALT (mucosa-associated lymphoid tissue) lymphoma , primary cutaneous follicular center lymphoma and primary cutaneous large B-cell lymphoma, leg type . As well as indirect evidence linking RhoH with malignancy, more direct linkage was first established in HCL (hairy cell leukaemia) where reconstitution of aberrant under-expression of RhoH limited disease progression in a xenograft mouse model . Subsequently, low expression of RHOH mRNA was found to be an independent unfavourable prognostic factor in AML (acute myeloid leukaemia) . Furthermore, high expression of RHOH mRNA has been shown to be associated with B-cell CLL (chronic lymphocytic leukaemia) and deletion of the RHOH gene in a genetic mouse model of CLL has been demonstrated to limit pathogenesis .
Within T-lymphocytes, the quantity of RhoH protein is regulated by a lysosomal degradation pathway that is activated following stimulation of the T-cell receptor complex . In addition, the steady-state levels of RHOH mRNA are reduced in Th1 T-helper cells when stimulated with an anti-CD3ϵ (cluster of differentiation 3ϵ) antibody and in Jurkat T-lymphocytic cells when activated with PMA . Changes in the steady-state levels of mRNA can be caused by mechanisms that regulate either gene transcription or transcript stability. If both or just one of these mechanisms control RhoH levels in T-lymphocytes is presently unknown.
Besides T-lymphocytes, mature B-lymphocytes represent the other major site of RhoH expression within the haematopoietic lineage [1–4]. However, in contrast with T-lymphocytes, very little is known about the mechanisms controlling RhoH expression in B-lymphocytes. The only insight gained to date is our finding that in B-lymphocytes transcription of the RHOH gene can be initiated at three distinct promoter regions and that one of these regions exhibits differential activity in the B-lymphocyte cell lines Raji and EH [1,26]. The Raji cell line was derived from a patient with Burkitt's lymphoma and expresses high steady-state levels of RHOH mRNA. EH cells were derived from a patient with HCL and mimic the under-expression of RhoH that characterizes this disease . Our previous studies demonstrated that the P3 promoter of the RHOH gene, extending from nucleotide −236 to +67 relative to the 5′ end of exon 4, is 7.4-fold more active in Raji compared with EH cells . Consequently, these results indicated that transcriptional repression of the RHOH gene could be the cause of RhoH under-expression in HCL.
In the present study, we report that, of the three promoters that can drive expression of the RHOH gene in B-lymphocytic cells, it is the P3 promoter that is preferentially utilized. Deletion studies indicate that the cis-acting elements critical for P3 promoter activity lie in a repressor region spanning nucleotides +68 to +245 relative to the 5′ end of exon 4 and an adjacent minimal promoter region spanning nucleotides −236 to +67. Within the minimal promoter region, two evolutionarily conserved binding sites for the transcription factor JunD were identified. Over-expression of JunD was found to inhibit transcription driven by the minimal promoter region when isolated and cloned into a plasmid vector. In contrast, knock-down of JUND either within Raji or AML mononuclear cells caused induced expression of endogenous RHOH mRNA. Finally, we determined that expression of PKCα (protein kinase Cα) is induced by either RhoH expression or JUND knockdown. Taken together, these data indicate that JunD binding to the P3 promoter of the RHOH gene represents a mechanism by which the quantitative levels of RhoH are controlled to effect downstream signalling pathways.
Interrogation of the GEO (Gene Expression Omnibus) database indicates that JUND mRNA is overexpressed in patients with HCL and AML where RHOH mRNA is low, but under-expressed in CLL where RHOH mRNA is high [37,38,42]. Furthermore, previous studies have demonstrated that, during physiological T-lymphocyte activation, JunD expression is induced, whereas that of RhoH is repressed [2,4,64]. Consequently, our finding that JunD represses transcription of the RHOH gene provides a mechanistic insight into their apparent reciprocal pattern of expression both in leukaemia and during normal physiological processes.
The B-lymphocytic cell lines Raji and Namalwa were obtained from A.T.C.C. JOK-1 B-lymphocytic cells stably expressing the linearized plasmid pMEP-RhoH (JOK-RhoH) or the corresponding parental plasmid pMEP4 (JOK-Empty) were obtained as previously described . The growth medium for all cell lines was composed of RPMI 1640, 10% (v/v) FBS (fetal bovine serum), 100 units/ml penicillin and 100 μg/ml streptomycin sulphate. However, JOK-1 growth medium also contained 150 μg/ml of hygromycin, whereas Raji growth medium was supplemented with 1% (v/v) non-essential amino acids and 1mM sodium pyruvate. Cultures were maintained in a humidified atmosphere of 5% CO2 at 37 °C and a density of 0.5–1.0×106 viable cells per ml. Complete medium changes were performed every 2–3 days.
Blood samples from two patients with AML were obtained from the Service des Maladies du Sang, Hôpital Huriez, Lille, France. These patients, designated AML#1 and AML#2, were classified by the French-American-British system as having stage 5 and stage 4 disease respectively. Both patients exhibited a normal karyotype and blood leucocyte counts ranging from 35–45×106 per millilitre. Patients AML#1 and AML#2 had 71% and 68% of their leucocytes assessed as blastic respectively. Peripheral blood was drawn after informed consent, then total mononuclear cells were isolated by Ficoll gradient centrifugation and frozen in liquid nitrogen. Donors provided informed consent in accordance with the Declaration of Helsinki and this study was approved by the Internal Review Board at the Tumorotheque of the Centre Hospitalier et Universitaire de Lille, Centre de Biologie, Lille, France.
The activity of the P3 promoter of the human RHOH gene was assessed using the expression vector pGL4.14[Luc2/Hygro] that contains a promoter-less firefly LUCIFERASE reporter gene (Promega). PCR that employed Platinum® Taq DNA Polymerase High Fidelity (Invitrogen) was used to generate eight fragments of the P3 promoter. These fragments spanned nucleotides −1519 to +245, −583 to +245, −583 to +67, −380 to +67, −236 to +67, −167 to +67, −103 to +67 and −24 to +67 relative to the 5′ end of exon 4 . Fragments with 3′ ends that mapped to nucleotide +67 were generated using a primer with the nucleotide sequence 5′-ATACAACAGTTGGGCTGGCAG-3′. Fragments with 3′ ends that mapped to nucleotide +245 were generated with a primer with the nucleotide sequence 5′-GAATGCCGAAGAAATGTCTTGACAG-3′. The 5′ ends of the P3 promoter fragments mapping to nucleotides −1519, −583, −380, −236, −167, −103 and −24 were generated with primers that had the respective nucleotide sequences; 5′-CAGCTCACTGCAACCTCAACC-3′, 5′-GGTAACTCCCAGCCCAGGTG-3′, 5′-ACCAAGATTGGTCAGTGGCATG-3′, 5′-GAAACATAGTCTCGGAGTAGGGTG-3′, 5′-TCTGCCTCTGATCTGGGGAAG-3′, 5′-CCTAAACCCACACAAATGAAACC-3′ and 5′-CTCTTTTGTCACATTCGGATTGC-3′. All the P3 promoter fragments generated by PCR were sub-cloned into the ‘filled-in’ HindIII site of pGL4.14[Luc2/Hygro] then validated by DNA sequencing (GATC Biotech, SARL, Mulhouse, France). The transfection control plasmid pRSV-βGal that directs constitutive expression of β-galactosidase was purchased from Promega. The plasmid pRSV-hjD that directs constitutive expression of JunD was provided by Yosef Shaul (Weizmann Institute of Science, Rehovot, Israel). The plasmid pRSV-Ng was used to control for any effects of pRSV-hjD not attributable to JunD overexpression. This negative control plasmid was produced by eliminating the entire JunD coding region from pRSV-hjD by digesting with SmaI followed by re-ligation of the isolated vector backbone. In previous studies, we have used pRSV-hjD and pRSV-Ng to elucidate molecular mechanisms that control expression of the CD11C gene .
siRNA (small interfering RNA)
Repression of JUND gene expression was achieved using siRNA purchased from Eurogentec. This siRNA has previously been shown to repress human JUND gene expression and consists of the nucleotide sequence 5′-ACGCGAACCUGAGCAGCUAUU-3′ . As a control for the siRNA targeted against transcripts encoding JunD, a siRNA was utilized with the nucleotide sequence 5′-AGACGGUGGUCAUUACCUAUU-3′. This control siRNA has no significant homology with any known human gene. The control siRNA was purchased from Eurogentec. Stock solutions (100 μM) of siRNA were stored at −80 °C. Prior to use, stocks were thawed and diluted to a final concentration of 10 μM in cell culture medium. The siRNA stock solutions were subjected to either one or two freeze-thaw cycles.
Evaluation of the transcriptional activity of the RHOH P3 promoter was achieved by utilizing nine luciferase reporter constructs. The first of these constructs was the control parental vector pGL4.14[Luc2/Hygro]. The additional eight constructs comprised the pGL4.14[Luc2/Hygro] vector containing the P3 promoter regions extending from −1519 to +245, −583 to +245, −583 to +67, −380 to +67, −236 to +67, −167 to +67, −103 to +67 and −24 to +67 relative to the 5′ end of exon 4 . A quantity (4.5 μg) of each of the P3 promoter constructs was combined with 0.5 μg of pRSV-βGal, then this mixture was transfected by electroporation into the Raji cell line using program M13 of the Nucleofector™ II apparatus and Cell Line Nucleofector Kit V (Lonza Group Ltd, Basel, Switzerland). In parallel to these experimental transfections, control transfections were performed with 0.5 μg of pRSV-βGal mixed with 4.5 μg of pGL4.14[Luc2/Hygro] empty of P3 promoter sequences. The influence of overexpression of JunD on recombinant RHOH was evaluated using the vector pGL4.14[Luc2/Hygro] into which was cloned the P3 promoter extending from nucleotides −236 to +67 relative to the 5′ end of exon 4. A quantity (2.5 μg) of this construct along with 0.5 μg of pRSV-βGal were mixed with 2.5 μg of the expression construct pRSV-hjD and in parallel with 2.5 μg of its empty parental vector pRSV-Ng. As controls for these experiments, parallel transfections were performed with the empty parental vector pGL4.14[Luc2/Hygro] from which the P3 promoter construct was derived. Raji cells were lysed 48 h after electroporation by a 5 min incubation at room temperature (25 °C) with 250 μl of lysis buffer [91 mM K2HPO4, 8 mM KH2PO4, 0.2% Triton X-100 and 1 mM DTT (dithiothreitol)]. The lysate was cleared by centrifugation (18000 g at 20 °C) and 50 μl of supernatant was mixed with 100 μl of luciferase buffer [25 mM Tris phosphate, pH 7.5, 8 mM MgCl2, 1 mM DTT, 1% (v/v) Triton X-100, 15% (v/v) glycerol, 1.8 mM ATP and 45 μM luciferin]. Luciferase activity was then quantified using a Lumat LB 9501 luminometer (Berthold Technologies). This same apparatus was used to quantify the level of β-galactosidase activity with a Luminescent β-galactosidase Detection Kit II (Clontech). All luciferase and β-galactosidase quantifications were determined in duplicate. The level of β-galactosidase activity was taken as reflective of transfection efficiency and used to correct the luciferase assay results. The influence of JunD overexpression on the endogenous RHOH gene was evaluated in the same way as its influence on recombinant RHOH, except that LUCIFERASE plasmids were omitted and 5 μg not 2.5 μg of each of the plasmids pRSV-hjD and pRSV-Ng were used. In addition, β-galactosidase levels were used to correct Q-RT–PCR (quantitative-reverse-transcription–PCR) analyses of total RNA and not luciferase activities. The influence of JunD repression on transcription of the endogenous RHOH gene in Raji cells was evaluated by electroporation using 500 nM of siRNAs targeting transcripts encoding JunD or a control siRNA with no human gene homology. Each siRNA was mixed with 500 nM of BLOCK-iT™ Fluorescent Oligo for electroporation (Invitrogen). At 48 h after siRNA electroporation, cells were washed once in PBS and fluorescence intensity was measured using a COULTER® EPICS® XL-MCL™ flow cytometer (Coulter). This intensity was taken as reflecting the level of transfection efficiency and used to correct Q-RT–PCR analyses of total RNA that was subsequently extracted. Transfection of total mononuclear cells isolated from AML patients was undertaken by first thawing the stored frozen cells, then washing them twice in RPMI 1640 containing 10% FBS. Next, cells were cultured for 2 h in the same medium to increase viability and transfection efficiency. Cells (5×106) were mixed with 500 nM of BLOCK-iT™ Fluorescent Oligo for electroporation (Invitrogen) and either 500 nM of a siRNA directed against JUND mRNA or 500 nM of a control siRNA with no human gene homology. These cells were then electroporated using the Human CD34+ Cell Nucleofector® kit from Lonza according to the manufacturer's instructions. At 48 h after siRNA electroporation, fluorescent intensity was measured and used to correct Q-RT–PCR analyses of subsequently extracted total RNA.
Proteins were isolated from cell cultures using M-PER® Mammalian Protein Extraction Reagent (Thermo-Fisher Scientific) and their concentrations were evaluated using the Protein Assay from Bio-Rad Laboratories. Proteins were reduced using the NuPAGE® Sample Reducing Reagent and 30 μg was immediately subjected to electrophoresis through 12% (v/v) NuPAGE® Gels using NuPAGE® Mops SDS Running Buffer (Invitrogen). After electrophoresis, proteins were transferred on to nitrocellulose filters using the iBlot® Dry Blotting System from Invitrogen. Next, filters were soaked for 2 h at room temperature in PBS containing 3% (w/v) non-fat dried skimmed milk powder then incubated overnight at 4 °C in this same blocking buffer containing rabbit polyclonal antibody sc-74 directed against JunD (Santa Cruz Biotechnology) or either rabbit monoclonal antibody 60A8 or C37F9 directed against c-Jun or JunB respectively (Cell Signaling Technology). Filters were then washed and incubated for 1 h at room temperature with a donkey anti-rabbit secondary antibody conjugated with HRP (horseradish peroxidase) that was purchased from Amersham Biosciences. Filters were again washed and HRP was visualized using the ECL(enhanced chemiluminescence) Western Blotting System (Amersham Biosciences). The protein loading of gels was evaluated by hybridization with the mouse monoclonal antibody AC-15 directed against β-actin (Sigma–Aldrich) and a sheep anti-mouse secondary antibody conjugated with HRP (Amersham Biosciences).
ChIP (chromatin immunoprecipitation)
Evaluation of the occupancy of the RHOH gene promoter by JunD was achieved using ChIP assays performed using the ChIP-IT™ kit (Active Motif). Unless otherwise stated, all reagents, buffers and supplies were included in the kit. Briefly, Raji cells were cross-linked with 1% (v/v) formaldehyde for 10 min at room temperature. After washing and treatment with glycine Stop-Fix solution, the cells were resuspended in lysis buffer and incubated for 30 min on ice. The cells were homogenized, and nuclei were resuspended in shearing buffer and subjected to optimized sonication shearing conditions (15–25 pulses of 20s each; between pulses the samples were allowed to cool on ice for 30s) to yield DNA fragments of 200–1000 bp in size. The sonicator instrument employed was a Vibra Cell™ (Sonics & Materials). Sheared chromatin was pre-cleared with Protein G beads and incubated overnight at 4 °C with 2 μg of a negative control human IgG antibody, an antibody directed against human RNA Polymerase II or the sc-74-x rabbit polyclonal antibody directed against human JunD (Santa Cruz Biotechnology). Protein G beads were then added to the antibody–chromatin mixtures and incubated for 1.5 h at 4 °C. After extensive washings, the immunoprecipitated DNA was removed from the beads with an elution buffer. To reverse cross-links and remove RNA, NaCl was added to a final concentration of 0.2 M and RNase A to a final concentration of 100 ng/μl. Samples were then incubated overnight at 65 °C. Next, the samples were treated with 180 ng/μl of proteinase K for 2 h at 42 °C and DNA purified using DNA purification mini-columns. The diluted purified DNA was subjected to semi-quantitative PCR analysis. Binding of human RNA Polymerase II to the appropriate region of the GAPDH gene promoter was used as a positive control for the efficiency of the ChIP procedure as recommended by the manufacturer of the ChIP-IT™ kit (Active Motif).
Total RNA was extracted from the mononuclear cells of AML patients and from the cell lines Raji, JOK-RhoH, JOK-Empty and Namalwa by using the High Pure RNA Isolation Kit (Roche Diagnostics). The procedure included a DNAse-I treatment to eliminate genomic DNA contamination. The amounts and quality of RNA were evaluated by UV spectroscopy using Nanodrop technology (Thermo Scientific). The cDNA templates for Q-RT–PCR were prepared as previously described . Evaluation of PKCα expression and the relative expression of RHOH gene transcripts initiated by promoters P1, P2 and P3 was achieved using SYBR® Green technology (Applied Biosystems). The quantity and quality of the cDNA templates were evaluated by analysis of ABL mRNA in accordance with the recommendations of Europe Against Cancer . PCR efficiencies were evaluated by generating linear regression curves using cDNAs from Raji or mononuclear cells. Each PCR reaction was performed in a 10 μl total volume using 0.4 μl of each of the sense and anti-sense primers at 10 μM, approximately 50 ng of cDNA template and 5 μl of 2×UDG Master Mix (Life Technologies). Transcripts of the PKCα gene were quantified using the primers PKCα-S and PKCα-AS that have the respective nucleotide sequences 5′-GATGGATGGAGTCACGACCAG-3′ and 5′-AGGACGCCATAGGCCCA-3′. Previous studies have validated the efficacy of this primer pair in determining PKCα expression . Transcripts specifically initiated by the P1 promoter were quantified using primer 21 located inside exon 1 coupled with primer 482 located inside exon 6. These primers had the respective nucleotide sequences 5′-CCCTTGACCCAAAGATACTGCTC-3′ and 5′-TCTCTGTCGGCTTCTACTCCAAG-3′. Transcripts initiated by promoter P2 were quantified using primer 482 coupled with primer 125 located inside exon 2 that had the sequence 5′-AACACTGCAGGGCCATTTG-3′. Transcripts initiated by promoter P3 were quantified using primer 41 located inside exon 4 coupled with primer 299 located inside exon 6. These primers had the respective nucleotide sequences 5′-CCCTGCCAGCCCAACTG-3′ and 5′-AGAGCGTTCTCCACGGCTT-3′. The primer 41 sequence is only present in transcripts initiated at promoter P3 since it is removed by splicing in transcripts initiated at promoter P1 or P2. Transcripts originating from the ABL gene were quantified using the primers ABL-F and ABL-R with the respective sequences 5′-TGGAGATAACACTCTAAGCATAACTAAAGG-3′ and 5′-GATGTAGTTGCTTGGGACCCA-3′ . The relative expression level of RHOH gene transcripts initiated specifically upstream of exons 1, 2 and 4 was calculated by dividing their quantitative level of expression by the quantitative level of ABL gene expression. Relative expression levels were determined from two independent cultures of Raji cells and each cDNA synthesis was analysed by three independent quantifications performed in triplicate. Evaluation of the sum total of all RHOH gene transcripts was achieved as we have previously described using Q-RT–PCR and TaqMan technology (Applied Biosystems) . The B-cell line Namalwa that expresses high levels of ABL and RHOH mRNA was used to produce calibration curves that evaluated PCR efficiency for each primer pair. Semi-quantitative PCR was used to analyse immunoprecipitated chromatin. In these studies, the degree to which JunD binds RHOH at the consensus AP1 (activator protein 1) site named AP1#1 was assessed using the primer pair 5′-AGAAACATAGTCTCGGAGTAGGGT-3′ and 5′-AAGAGATTAGCAAGTTCCTGT-3′. JunD binding to the AP1 site named AP1#2 was assessed using the primer pair 5′-GGCACAGGAACTTGCTAATCTC-3′ and 5′-CGAGAATATTAACTGATTCACACTGGAAGAACTG-3′. The binding of RNA polymerase II to the promoter region of the GAPDH gene was assessed using a sense-strand primer with the nucleotide sequence 5′-TACTAGCGGTTTTACGGGCG-3′ coupled with an antisense-strand primer with the nucleotide sequence 5′-TCGAACAGGAGGAGCAGAGAGCGA-3′. The products of semi-quantitative PCR were visualized by agarose gel electrophoresis.
Total RNA was isolated from the B-lymphocytic cell lines JOK-RhoH and JOK-Empty using the RNAeasy kit from Qiagen that included a DNAse I treatment. Control of quality was evaluated using an Agilent 2100 Bioanalyser (Agilent Technologies). Total RNA (0.5 μg) was then processed and analysed on Human Whole Genome Agilent 44K 60-mer oligonucleotide microarrays using the kits for cRNA amplification, labelling, fragmentation, hybridization and washing recommended by the manufacturer for the Two-Color Microarray-Based Gene Expression protocol (Agilent Technologies). For each cell line, two biological samples were studied by dye-swap analysis. Microarrays were scanned using an Agilent DNA Microarray G2505B scanner, the expression data was extracted by Feature Extraction (version 18.104.22.168) and then processed by GeneSpring version 7.3 (Agilent Technologies) for normalization, filtering and statistical analysis using the Student's t test. Significance was assessed at a P value of 0.05.
Cross-species sequence homology analysis of the RHOH gene was performed using the computer program GenomeVISTA . This is an automatic server that allows the user to find candidate orthologous regions for a draft or finished DNA sequence from any species on a base genome and provides detailed comparative analysis. A computational strategy is used where query sequences are anchored on the base genome by local alignment matches and then globally aligned to candidate regions by the program AVID . The VISTA family of tools are developed and hosted by the Genomics Division of Lawrence Berkeley National Laboratory and the United States Department of Energy Joint Genome Institute (Walnut Creek, CA, U.S.A.). The sequence of the P3 promoter of humans was aligned to the equivalent sequence in the genomes of mouse (chr5:66254636–66255114 [+]), rat (chr14:45055696–45056166 [−]), rhesus monkey (chr5:35224232–35224710 [+]), horse (chr3:80658339–80658821 [−]) and dog (chr3:75142670–75143127 [−]). Putative transcription factor binding sites within the P3 promoter were identified using the analytic program Match™ version 1.0 Public (BIOBASE Biological Databases, Wolfenbüttel, Germany). This software uses a library of mononucleotide weight matrices from the database TRANSFAC® 6.0 that contains experimentally proven eukaryotic transcription factor-binding sites. Interrogation of the GEO database determined the relative expression of C-JUN, JUNB and JUND in Raji cells, the relative expression of RHOH and JUND in HCL patients, the expression of RHOH in DLBCL compared with normal B-lymphocytes and the relative expression of JUND in AML and CLL mononuclear cells compared with normal controls [35,36]. The GEO database was created and is administered by the National Center for Biotechnology Information and serves as a public repository for a wide range of high-throughput experimental data. These data include single and dual channel microarray-based experiments measuring mRNA, miRNA (microRNA), genomic DNA and protein abundance, as well as non-array techniques such as serial analysis of gene expression. The specific DataSet interrogated for expression of JUND mRNA in AML patients compared with normal volunteers was GDS1059 . The equivalent DataSet interrogated for CLL patients was GDS1454 . The DataSets interrogated for the relative expression of C-JUN, JUNB and JUND in Raji cells were GDS181 and GDS989 [39,40]. The DataSet interrogated for the expression of RHOH in DLBCL compared with normal B-lymphocytes was GDS3516 . The Series record interrogated for relative expression of JUND and RHOH in HCL patients was GSE2350 .
Preferential usage of the P3 promoter of the RHOH gene by B-lymphocytic cells
Previous studies have established that transcription of the RHOH gene can be initiated at three distinct sites  (Figure 1A). These sites define the 5′ ends of non-coding exons 1, 2 and 4. The promoter regions responsible for these initiation events are designated P1, P2 and P3 respectively. We used Q-RT–PCR to determine whether the different promoters of the RHOH gene exhibit preferential usage in the B-lymphocytic cell line Raji (Figure 1B). This analysis demonstrated that promoter P3 is <8-fold more active than the P1 and P2 promoters combined.
The P3 promoter is controlled by both repressor and activator elements
The human P3 promoter extending from −1519 to +245 relative to the 5′ end of exon 4 was cloned into a LUCIFERASE reporter plasmid to identify individual cis-acting elements that control its function. The −1519/+245 region was chosen as a starting point for our analysis since −1519 to −1 contains the most evolutionarily conserved section of the intron that separates exon 4 from exon 3 and +1 to +245 represents all but the last eight nucleotides of exon 4. These last eight nucleotides were excluded since they contain nucleotides involved in RNA splicing and therefore have the potential to confound reporter gene studies. The LUCIFERASE reporter plasmid containing the −1519/+245 region was found to direct transcription in Raji cells only 4-fold above the background level directed by the plasmid empty of RHOH gene sequences (Figure 2A). A modest increase in expression to 6-fold above background occurred when nucleotides −1519 to −584 were deleted, indicating this region does not contain cis-acting elements of significance. On the basis of this finding, we now concentrated on the P3 promoter region spanning nucleotides −583 to +245. Deletion of nucleotides +68 to +245 dramatically increased transcriptional activity to 17-fold above background, indicating this region contains repressor cis elements. Consequently, the span of nucleotides mapping from +68 to +245 was designated a repressor region (Figure 2B). Next, serial 5′ truncation of the −583/+67 region was performed. The 5′ ends of this series mapped such that successive constructs represented the progressive deletion of distinct regions of evolutionary conservation. Significant and progressive reductions in transcriptional activity occurred when nucleotides downstream of −236 were deleted. Therefore the region spanning nucleotides −236 to +67 was taken to represent the minimal promoter region (Figure 2B). Deletion of nucleotides −167 to −104 dramatically reduced transcriptional activity, indicating this region contains activating cis elements. Consequently, the −167/−104 span of nucleotides was designated an activator region (Figure 2B).
Identification of evolutionarily conserved elements within the RHOH P3 promoter that represent potential transcription factor-binding sites
Critical cis-acting control elements are often conserved during evolution. Therefore, in order to identify such elements within the P3 promoter, the human sequence was aligned with the equivalent DNA sequences from mouse, rat, rhesus monkey, horse and dog (Figure 3). The human sequence was then analysed for consensus transcription factor-binding sites and these sites mapped over the cross-species alignment. These in silico procedures focused on the repressor region spanning nucleotides +68 to +245 and the minimal promoter region spanning nucleotides −236 to +67 that included the −167/−104 activator region. Within the −167/−104 activator region, single evolutionarily conserved consensus binding sites were identified for Ets1 (E-twenty six 1), NFAT (nuclear factor of activated T-cells) and Sp1 (specificity protein 1) (Figure 3A). Each of these factors has been reported to have the potential to function as a transcription activator [43–45]. Outside the activator region, but within the minimal promoter region, single evolutionarily conserved potential binding sites for AML1 and Ets1 were identified along with two potential binding sites for NFAT and two for the AP1 family of transcription factors. The potential AP1-binding site mapping most distal to the 5′ end of exon 4 was designated AP1#1 and the proximal site was designated AP1#2 (Figure 3A). Within the +68/+245 repressor region, evolutionarily conserved consensus binding sites were identified for Ets1, NFAT, AML1, LYF1 (lymphoid factor 1) and members of the C/EBP (CCAAT/enhancer-binding protein) and AP1 families (Figure 3B). All these factors have the potential to act as transcription repressors [46–54]. Mutations sometimes found in the repressor region of patients with DLBCL disrupt binding sites for NFAT, AML1 and LYF1 while creating a potential binding site for MZF1 (myeloid zinc finger 1) [20,55]. Interrogation of the microarray studies deposited with the GEO database indicate that RHOH is expressed at higher levels in DLBCL compared with normal B-lymphocytes . Therefore it is intriguing to speculate that this induction is due to the disruption of repressor elements within the P3 promoter and their replacement with a binding site for the transcription activator MZF1.
JunD binds the endogenous P3 promoter
The putative cis-acting elements most commonly identified by in silico analysis of the P3 promoter are potential binding sites for the AP1 family of transcription factors. Two such elements map within the minimal promoter region and four map within the repressor region (Figure 3). AP1 is a group of dimers composed of members of the Jun, Fos and ATF (activating transcription factor) families of proteins . The majority of AP1 dimers contain c-Jun, JunB or JunD, which comprise the members of the Jun family. Analysis of the GEO database indicates that, in Raji cells, mRNA encoding JunD predominates over those encoding c-Jun and JunB [39,40]. Western blot analysis indicates that this relative expression profile is reflected at the protein level (Figure 4A). Since JunD appears to be the member of the AP1 family that is expressed at the highest level in Raji cells, ChIP was used to determine if this transcription factor binds the endogenous RHOH P3 promoter (Figure 4B). This analysis established that JunD binds each of the two AP1 sites within the minimal promoter region.
JunD is both necessary and sufficient to repress transcription driven by the P3 promoter
The binding of JunD to the P3 promoter of the endogenous RHOH gene suggests this transcription factor might play a role in controlling RHOH expression. In order to begin to determine this role, siRNA that inhibits expression of JunD was transfected into Raji cells and its affect on expression of both the cloned and endogenous RHOH gene was determined. These analyses indicated that, on average, JunD repression increased transcription driven by the cloned minimal promoter region, the endogenous P3 promoter and all RHOH promoters combined by 110%, 29% and 37% respectively (Figure 5). Consequently, it appears that JunD is necessary to affect repression of the RHOH gene. This conclusion was confirmed when a plasmid that directs constitutive expression of JunD was employed. For this, transcriptional activity of the cloned minimal promoter region, the endogenous P3 promoter and all RHOH promoters combined was reduced on average by 14%, 18% and 26% respectively (Figure 6). These results indicate that JunD is not only necessary but also sufficient to affect RHOH repression. The levels of RHOH repression that are affected by JunD induction are lower than those of RHOH induction elicited by JunD inhibition. This could indicate that JunD requires ancillary factors to exert its full function and that these factors are in limited supply.
JunD effects repression of the endogenous RHOH gene in mononuclear cells isolated from patients with AML
Immortalized cell lines cultured in vitro for many years may not reflect the function of native human cells. Consequently, we sought to validate our finding in the cell line Raji that JunD is a RHOH repressor with a functional analysis of leucocytes isolated from volunteer patients. RHOH mRNA has been shown to be expressed at abnormally low levels in the mononuclear cells of patients diagnosed with AML . In addition, interrogation of the GEO database indicates that, compared with normal mononuclear cells, those of AML patients tend to overexpress JUND mRNA . Therefore in AML cells there appears to be a correlation between low expression of RhoH and high expression of JunD. Consequently, AML cells were transfected with siRNA directed against JunD in order to determine if repression of JunD could affect induction of the sum total of RHOH transcripts irrespective of whether they initiated at the P1, P2 or P3 promoter (Figure 7). JunD siRNA caused mononuclear cells isolated from two different AML patients to both exhibit significant increases in expression of total RHOH mRNA. The increases of 132% and 89% in AML cells compared favourably with the 37% increase in total RHOH transcripts that was caused by JunD siRNA in Raji cells.
Both RhoH induction and JunD repression cause increased expression of PKCα
HCL is characterized by both overexpression of JunD and under-expression of RhoH [26,29]. JOK-1 is a B-lymphocytic cell line derived from a patient with HCL. In a previous study we determined that JOK-1 exhibits a low expression of RHOH mRNA . Pools of JOK-1 were generated that stably contain in their genome either an empty expression vector or the same vector constitutively expressing RHOH. These pools of JOK-1 were designated JOK-Empty and JOK-RhoH respectively. JOK-RhoH produces xenograft tumours that, on average, are 70% smaller than those produced by JOK-Empty . Comparison of the transcriptomes of JOK-Empty and JOK-RhoH by microarray analysis was performed in order to identify downstream targets of RhoH signalling. For each cell line, two biological samples were studied by dye-swap analysis. The result of this analysis identified a 56% increase in the expression of the mRNA encoding PKCα. Using a one-sided Student's t test, the probability value for this increase was calculated at P=0.016. Q-RT–PCR analysis confirmed the microarray study by demonstrating that expression of PKCα was increased by 85% in JOK-RhoH compared with JOK-Empty (Figure 8A). Given that JunD is a repressor of RhoH, it would be expected that inhibition of JunD would increase expression of PKCα due to increased expression of RhoH. Consistent with this expectation, increases of 70% and 37% in PKCα expression were elicited in mononuclear cells isolated from two AML patients when they were transfected with siRNA targeted against JUND mRNA (Figure 8B). Consequently, JunD appears to be a functionally relevant repressor of RhoH since its inhibition both upregulates RhoH and also at least one of its downstream signalling targets.
RHOH mRNA was originally identified in a NHL (non-Hodgkin's lymphoma) cell line as a fusion with transcripts originating from the BCL6/LAZ3 gene [15,63]. Since the fused transcripts resulted from a translocation of chromosomes 3 and 4, RhoH was initially named TTF for translocation three four. The TTF gene was subsequently found to encode a novel member of the Rho family of small GTP-binding proteins. Consequently, TTF was renamed RhoH to indicate a Rho protein following in the discovery sequence designated by the Gene Nomenclature Committee of the Human Genome Organization (European Bioinformatics Institute, Cambridge, U.K.) . That RhoH was also found to be expressed specifically in haematopoietic cells was conveniently coincident with its alphabetic designation.
Following its original identification as a participant in a NHL translocation, the association between aberrant RhoH expression and haematopoietic malignancy has strengthened. In AML, low expression of RHOH mRNA was shown to be a predictor of worse prognosis in both overall and disease-free survival . B-lymphocytes derived from patients with HCL also exhibit low expression of RHOH transcripts . A direct link between this low expression and HCL pathogenesis was demonstrated in a xenograft mouse model where reconstitution of high levels of RhoH reduced primary tumour burden by 70% and protected against splenomegaly and mortality . The B-lymphocytes of patients with CLL express abnormal levels of RHOH mRNA . However, unlike AML and HCL where RHOH mRNA is unusually low, in CLL it is unusually high. As with HCL, a direct link between the quantitative level of RhoH expression and disease pathogenesis has been established in CLL using an animal model. Specifically, deletion of the RHOH gene in the Eμ-TCL1Tg mouse model of CLL was found to delay accumulation of leukaemic cells in peripheral blood and the leukaemic burden in the peritoneal cavity, bone marrow and spleen . Taken together, the analyses of malignant leucocytes indicate that the function of RhoH is controlled to a significant degree by its quantitative level of expression. Indeed, such control has been proposed to be largely responsible for RhoH function, since the protein is expressed in a constitutively active conformation [2,14].
Despite the likelihood that RhoH levels are critical to its function, only one study to date has defined a mechanism by which these are regulated . In that study T-lymphocytes were shown to down-regulate RhoH protein levels by lysosomal degradation following stimulation of the T-cell receptor complex with an anti-CD3ϵ antibody. In the present study, we report for the first time a mechanism that controls RHOH expression in B-lymphocytic cells and mononuclear cells isolated from patients with AML. We find that in the B-lymphocytic cell line Raji and AML mononuclear cells RHOH is repressed by JunD binding its P3 promoter. Interestingly, RNA polymerase II exhibits binding to the JunD site that maps between only 14 and 20 nucleotides upstream of the transcription initiation site defined by the P3 promoter. Therefore, it is possible that JunD affects repression of the P3 promoter either by passive steric hindrance of RNA polymerase loading or active direct binding.
The finding that JunD represses RHOH is consistent with the observation in HCL of abnormal constitutive expression of JunD  but under-expression of RHOH . In silico analysis of microarray studies deposited in GEO confirms the reciprocal expression of JUND and RHOH mRNA in HCL . In addition, further analysis of the GEO database indicates that compared with normal mononuclear cells those of AML and CLL patients tend to over- and under-express JUND mRNA respectively [37,38]. Since AML and CLL patients respectively under- and over-express RHOH mRNA, a consistent inverse correlation appears to exist between RHOH and JUND expression. This correlation also holds in physiological rather than pathological circumstances, since RHOH expression is reduced during activation of normal T-lymphocytes [2,4], whereas JUND expression is induced . Consequently, our finding that JunD acts as a RHOH repressor provides a mechanistic connection between their reciprocal expression both in physiology and pathology. This connection indicates a reinforcing feedback loop may exist in which RhoH down-regulation by JunD relieves RhoH inhibition of Rac1 that is then able to activate JunD through PAK1 (p21-activating kinase 1) and MEK1 (mitogen-activated protein kinase/extracellular-signal-regulated kinase kinase 1) (Figure 9) [2,61,62]. However, a counter-balancing feedback loop may also exist between JunD and RhoH. This possibility arises since both RhoH induction and JunD inhibition induce PKCα. The kinase produced by PKCα can activate the AP1 family, of which JunD is a member [58–60]. Therefore, JunD could inhibit its own activator through RhoH repression (Figure 9). Our data cannot exclude the possibility that JunD could also directly inhibit PKCα by binding its cis-acting control elements. However, to date the only transcription factors implicated in PKCα regulation are AP2, Sp1, Ets1 and SOX9 [65,66].
RhoH plays an important role in haematopoiesis by driving leukocyte differentiation and inhibiting leukocyte survival and proliferation [3,5,6]. In myeloid cells, PKCα has also been shown to promote differentiation and inhibit proliferation [67,68]. In haematopoietic progenitor cells, PKCα has been shown to limit survival and decrease proliferation . Therefore, the finding that RhoH is a positive regulator of PKCα expression is consistent with an intracellular signalling cascade involved in producing short-lived differentiated leukocytes with a limited capacity to divide. Consequently, it is not surprising that leucocytes lacking these characteristics are hallmarks of AML where RhoH expression has been demonstrated to be low  and HCL where both RhoH and PKCα expression are low [26,70]. However, in CLL, RhoH is expressed at abnormally high levels . Initially, this would appear to contradict the apparent beneficial role played by RhoH in haematopoiesis. However, CLL is characterized by over-expression of active protein kinase CβII . Protein kinase Cβ antagonizes the function of PKCα . Therefore, even though RhoH expression is high in CLL, its function is likely confounded by inhibition of its down-stream effector.
Bruno Quesnel and Céline Berthon provided anonymized patient samples and clinical records. Carl Simon Shelley designed and interpreted experiments, provided research funds and wrote the paper. Jean-Pierre Kerckaert and Martin Figeac performed microarray experiments. Laure Delestré performed the majority of experiments, helped design and interpret experiments and helped write the paper. Sylvie Galiègue-Zouitina designed and interpreted experiments, provided research funds and helped write the paper.
This work was supported by The Gundersen Lutheran Medical Foundation, La Crosse, WI, U.S.A. and grants from the Hairy Cell Leukemia Research Foundation, Institut pour la Recherche sur le Cancer de Lille, comité du Nord and comité du Pas de Calais of Ligue Nationale Française Contre le Cancer, Institut National de la Santé et de la Recherche Médicale. L.D. was supported by fellowships from the Ministère de l'Enseignement Supérieur et de la Recherche and from the Institut pour la Recherche sur le Cancer de Lille.
We thank the Institut National de la Santé et de la Recherche Médicale IFR114, IMPRT, and EPER in Lille for use of their Applied Biosystems Abi-Prism 7900 quantitative PCR apparatus. In addition we would like to thank Sabine Quief and Céline Villenet for help with microarray hybridization, Nathalie Jouy for help with flow cytometry and Claude Denis for excellent technical assistance.
Abbreviations: AML, acute myeloid leukaemia; AP, activator protein; FBS, fetal bovine serum; CD3ϵ, cluster of differentiation 3ϵ; ChIP, chromatin immunoprecipitation; CLL, chronic lymphocytic leukaemia; DLBCL, diffuse large B-cell lymphoma; DTT, dithiothreitol; Ets1, E-twenty six 1; GEO, Gene Expression Omnibus; HCL, hairy cell leukaemia; HRP, horseradish peroxidase; LYF1, lymphoid factor 1; MEK1, mitogen-activated protein kinase/extracellular-signal-regulated kinase kinase 1; MZF1, myeloid zinc finger 1; NFAT, nuclear factor of activated T-cells; NHL, non-Hodgkin's lymphoma; PAK1, p21-activating kinase 1; Q-RT–PCR, quantitative-reverse-transcription–PCR; PKCα, protein kinase Cα; siRNA, small interfering RNA; Sp1, specificity protein 1; TTF, translocation three four
- © The Authors Journal compilation © 2011 Biochemical Society