Involvement of HTLV-I Tax and CREB in aneuploidy: a bioinformatics approach

1742-4690-3-43 1742-4690 Research Involvement of HTLV-I Tax and CREB in aneuploidy: a bioinformatics approach de la Fuente Cynthia bcmclf@gwumc.edu Gupta V Madhur bcmmvg@gwumc.edu Klase Zachary zklase@gwu.edu Strouss Katharine strouss@gwu.edu Cahan Patrick Patrick.cahan@gmail.com McCaffery Timothy mcc@gwu.edu Galante Anthony galante@phri.org Soteropoulos Patricia patricia@phri.org Pumfery Anne pumferan@shu.edu Fujii Masahiro fujiimas@med.niigata-u.ac.jp Kashanchi Fatah bcmfxk@gwumc.edu

The George Washington University Medical Center, Department of Biochemistry and Molecular Biology, Washington, DC 20037, USA

Center for Applied Genomics, Public Health Research Institute, Newark, NJ 07103, USA

The Institute for Genomic Research (TIGR), Rockville, MD 20850, USA

Department of Immunotherapeutics, Niigata University School of Medicine, Asahimachi-Dori, Niigata 951-8510, Japan

Department of Virology, Niigata University School of Medicine, Asahimachi-Dori, Niigata 951-8510, Japan

Retrovirology 1742-4690 2006 3 1 43 http://www.retrovirology.com/content/3/1/43 16822311 10.1186/1742-4690-3-43 10 1 2006 05 7 2006 05 7 2006 2006 de la Fuente et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background

Adult T-cell leukemia (ATL) is a complex and multifaceted disease associated with human T-cell leukemia virus type 1 (HTLV-I) infection. Tax, the viral oncoprotein, is considered a major contributor to cell cycle deregulation in HTLV-I transformed cells by either directly disrupting cellular factors (protein-protein interactions) or altering their transcription profile. Tax transactivates these cellular promoters by interacting with transcription factors such as CREB/ATF, NF-κB, and SRF. Therefore by examining which factors upregulate a particular set of promoters we may begin to understand how Tax orchestrates leukemia development.

Results

We observed that CTLL cells stably expressing wild-type Tax (CTLL/WT) exhibited aneuploidy as compared to a Tax clone deficient for CREB transactivation (CTLL/703). To better understand the contribution of Tax transactivation through the CREB/ATF pathway to the aneuploid phenotype, we performed microarray analysis comparing CTLL/WT to CTLL/703 cells. Promoter analysis of altered genes revealed that a subset of these genes contain CREB/ATF consensus sequences. While these genes had diverse functions, smaller subsets of genes were found to be involved in G2/M phase regulation, in particular kinetochore assembly. Furthermore, we confirmed the presence of CREB, Tax and RNA Polymerase II at the p97Vcp and Sgt1 promoters in vivo through chromatin immunoprecipitation in CTLL/WT cells.

Conclusion

These results indicate that the development of aneuploidy in Tax-expressing cells may occur in response to an alteration in the transcription profile, in addition to direct protein interactions.

Background

Human T-cell leukemia virus type 1 (HTLV-I) is a complex retrovirus that causes adult T-cell leukemia/lymphoma (ATLL), a CD4 lymphoproliferative disease 12. While endemic in Japan, South America, Africa, part of the Middle East and the Carribean, there is an increasing prevalence of HTLV-I seropositivity world wide 123. ATL develops in 2–5% of HTLV-I-infected individuals after a long latency period of about 20–30 years 456. Different clinical features have resulted in the division of this disease into four clinical subtypes characterized by increasing aggressiveness: smoldering, chronic, lymphoma, and acute ATL 7.

One important marker for the risk of ATL within patients is the percentage of abnormal T lymphocytes versus normal T lymphocytes within the peripheral blood 8. Binucleated lymphocytes or lymphocytes containing cleaved/cerebriform nuclei (also termed "flower" cells) have been observed in blood smears of HTLV-I infected individuals and in ATLL cells 79101112. These cells are representative of aneuploidy or abnormal chromosomal content which develops due to aberrant mitotic divisions 13. Since aneuploidy has been suggested to contribute to tumorigenesis, there is a growing interest in deciphering the events in late G2/mitosis phase and defects therein that lead to aneuploidy. Additionally, aneuploidy may be associated with an acquired resistance to chemotherapeutic agents such as imatinib or 5-fluorouracil 14; therefore, therapeutics disrupting aneuploidy development may improve upon current therapies for ATLL patients.

There is also a growing body of evidence to suggest that Tax, a 40 kDa viral oncoprotein encoded by HTLV-I, controls various aspects of cell cycle check points needed for aneuploidy. In fact, we were one of the first groups to show that Tax controls the G1/S check point 15, which was later confirmed by others 16, resulting in failure of G1 checkpoint and NER deficiency 17. For a more comprehensive review of the cell cycle and check point controls by Tax, we recommend some of the more relevant reviews published recently 18192021. In addition to disrupting checkpoints at the G1/S resulting in continuous cellular proliferation, Tax also directly targets a number of G2 and mitotic regulators. One of the first indication of Tax's involvement in the G2 and M phases was shown by Jin and colleagues 22 who discovered that Tax binds to hsMAD1. MAD1 and MAD2 are two of several genes that are involved in the activation of the mitotic spindle checkpoint function (MSC) following chromosomal missegregation. Tax interaction hindered the binding of MAD1/MAD2 complex to kinetochores by inducing translocation of these factors from the nucleus to the cytoplasm 23. Furthermore, recent reports have demonstrated that Tax promotes activation of the anaphase promoting complex (APC)- APC^Cdc20pleading to a reduction in Pds1p/securin and Clb2p/cyclin B levels in yeast, rodent and human cells 624. Overall, the degradation of these critical check point proteins results in a delay or failure in mitotic entry and progression, and is accompanied by a loss of cellular viability, resulting in aberrant anaphase progression, chromosomal instability and severe DNA aneuploidy 252627.

Concurrently, Tax has been shown to repress cellular DNA repair by binding to Chk2 2428 and Chk1, thus impairing kinase activities in vitro and in vivo 25. Moreover, Tax silences cellular checkpoints, which guard against DNA structural damage and chromosomal missegregation, thereby favoring the manifestation of a mutator phenotype in cells 18. In such cells, rapidly induced cytogenetic damage can be measured by a significant increase in the number of micronuclei (MN) in cells knocked-out for DNAPKcs 2930. Therefore, it is possible that Tax perturbs many dynamic complexes that coordinate the processes of cell cycle regulation and DNA repair.

Here we present evidence that cytotoxic T cells (CTLL) stably expressing wild type Tax (CTLL/WT) exhibited a higher incidence of aneuploidy when compared to a Tax clone deficient for CREB transactivation (CTLL/703) 31. Given the role of Tax as a strong activator of both viral and cellular transcription, we address the role of Tax-dependent transcription through the CREB/ATF pathway in the possible development of aneuploidy. We performed gene expression microarray analysis comparing CTLL/WT to CTLL/703 cells. Those genes that were either up or down-regulated in CTLL/WT cells were functionally annotated using the NIH's Database for Annotation, Visualization, and Integrated Discovery (NIH-DAVID). Next, we used an online database – PromoSer – to extract promoters of annotated genes to determine which of these genes contained CREB binding sites. Finally, chromatin immunoprecipitation was used to determine if DNA binding proteins such as Tax, Pol II and CREB were present at the promoters of the few selected genes. Our results clearly indicate that Tax/CREB binds to promoters of many genes, including Sgt1 and p97 (Vcp), which have functions in spindle formation and disassembly, respectively. The consequences of their over-expression and involvement in aneuploidy will be discussed.

Results and discussion

Aneuploidy prevalence in CTLL/WT cells

Tax deregulates the expression of genes that encode interleukin-2 (IL-2) and the multisubunit (IL-2R alpha, IL-2R beta, and IL-2R gamma) IL-2 receptor (IL-2R) in the early phases of HTLV-I induced ATL 32. In the later phases of ATL, cells no longer produce IL-2 but still continue to express the IL-2R 32. Fujii and colleagues developed an IL-2 independent system where Tax was stably expressed in a mouse IL-2-dependent T-cell line, CTLL-2, and examined the growth property of these cells in the absence of IL-2 31. While the Tax M47 (703) mutant activates NF-κB-dependent transcription but not CRE-dependent transcription, the reverse is true for the Tax M22 mutant 31. They also noted that in addition to Tax's role in cAMP-responsive element (CRE) and NFκB activation pathways, Tax also increases expression of mRNAs coding for various AP-1 transcription factor family members including c-Jun, JunB, JunD, c-Fos and Fra-1. Genes encoding AP-1 are immediate-early genes whose products play important roles in cell activation, proliferation and transformation. Thus, an alternate pathway, i.e., AP-1, may be involved in the dysregulated phenotypes of T cells expressing Tax (CTLL) or infected with HTLV-1 33.

Initially, these mutants were used to investigate the involvement of the transcription pathways in the transformation of CTLL-2 cells by Tax. Wild type Tax expression in CTLL-2 cells resulted in IL-2 independent growth. The Tax M47 mutant still activated the NF-κB-dependent transcription, and was able to support the growth of CTLL-2 in the absence of exogenously added IL-2. Therefore, the CREB dependent activity of Tax may not be as critical for IL-2 dependent growth, but may be needed for Tax induced transformation of cells. Tax M47 induced transformation may be accomplished through deregulation of cellular oncogenes, tumor suppressor genes, or checkpoint genes for DNA damage.

Our previous work utilizing these cell lines have shown that depending on which phase of the cell cycle DNA damage occurred, two different phenotypes were observed. Using centrifugal elutriation, we were able to fractionate cells at G1, S, or G2/M based on differences in cell volume at these distinct phases 34. These cell fractions were gamma irradiated and harvested 24 hrs later for cell cycle analysis by propidium iodide staining and FACS. We observed that the CTLL/WT cells exhibited a distinct phenotype; at G1, these cells were able to induce a G1/S checkpoint, while at S or G2/M phase, these cells apoptosed after gamma irradiation. Conversely, CTLL/703 cells continued to proliferate without any apoptosis. We believed these differences were due, at least in part, to the differing gene expression profile of these cells and the induction of DNA damage at a particular point in the cell cycle. Interestingly, it was also observed that unirradiated CTLL/WT cells displayed a higher prevalence of aneuploidy than CTLL/703. The appearance of aneuploidy occurred in the later fractions (G2/M phase) which represented the largest cell populations. Consistent with our findings, previous reports have also indicated that centrifugal elutriation was capable of separating mixed populations of diploid and aneuploid cells 3536.

To examine the chromosomal instability in CTLL/WT cells, we first performed a metaphase chromosome spread to determine the average number of chromosomes 637. Both CTLL/WT and CTLL/703 cells were processed as described in the Methods section. Thirty-five cells were analyzed for each cell type. As shown in Figure 1, CTLL/WT cells displayed higher numbers of chromosomes with an average number of 61 in contrast to CTLL/703 cells, where the average number of chromosomes were 44. The basic karyotype of the Mus musculus species is 2N = 40 38. These results support our earlier observation that wild-type Tax expressing cells (CTLL/WT), as compared to the CREB deficient Tax clones (CTLL/703), had a higher incidence of aneuploidy.

Figure 1

Aneuploidy in CTLL/WT cells

Aneuploidy in CTLL/WT cells. CTLL/WT and CTLL/703 cells were treated with 10 μg/ml colcemid, centrifuged, and resuspended in hypotonic solution to swell the cells. Cells were then fixed and dropped onto slides. After being stained with Giemsa and dried, slides were analyzed using the Olympus BX-60 microscope. A total of 35 metaphase spreads were counted. Representative chromosome spreads of CTLL/WT (panel A) and CTLL/703 (panel B) are displayed with 100 and 42 chromosomes, respectively. Panels C) and D) are graphical representations of the raw counts from these two cell types.

Since both of these cells are transformed, i.e. IL-2 independent 31, differences in transformation status cannot explain the presence of aneuploidy in one cell line and not the other. While it is possible that mutation of the Tax protein, resulting in a CREB transactivation deficient Tax (CTLL/703 cells), may disrupt interactions of Tax with late cell cycle checkpoint proteins whose dysregulation contributes to aneuploidy, however, this seems to be an unlikely event. For instance, Tax interaction with hMAD1 (also known as Txbp181) appears to be dependent on the zinc finger motif located within the N-terminus of Tax 22, and not the N-terminal domain as seen in the M47 mutant. Therefore, it appears probable that Tax may contribute to aneuploidy development, at least in part, through transcriptional activation of cellular genes critical for aneuploidy. This result would not be without precedence, since the involvement of Tax in immortalization has been shown to be mediated both at the transcriptional level and by direct protein:protein interactions 18. Furthermore, since Tax-dependent CREB transactivation was deficient in CTLL/703 and not in CTLL/WT cells (see below), it would appear that those genes involved in aneuploidy development may be CREB-dependent.

Gene expression profiling and promoter analysis

To begin to examine the contribution of Tax/CREB-dependent transcription in aneuploidy development, we compared the transcription profiles of CTLL/WT and CTLL/703 cells by microarray analysis. Our method for comparing the contribution of Tax/CREB to the aneuploidy phenotype is depicted in Figure 2. Through this analysis we obtained a global expression profile of the wild-type Tax-expressing cells as compared to Tax-703 expressing cells and subsequently narrowed our list by determining which genes contained CREB-response elements and potential aneuploidy associated genes. Microarray analysis was performed utilizing cytoplasmic mRNA from both cells and the Affymetrix Murine Genome U74A Array. Analysis was performed in duplicate to minimize inter-chip variability. After normalizing the fluorescence intensity of each probe and filtering for differentially expressed genes with a difference of at least 2-fold across experiments, a gene list was compiled. These genes were then functionally annotated utilizing the Database for Annotation, Visualization, and Integrated Discovery program (DAVID) 39 to generate a list of 439 genes differentially regulated in CTLL/WT cells [Additional files 1 and 2]. The majority of these differentially regulated genes were up-regulated in CTLL/WT cells (412 genes increased out of a total of 439 annotated genes, 94%) and many have been shown to be up-regulated in wild-type, and not in M47, Tax-expressing cells, including p21/waf1 40, cyclin D2 15, and Jun B 41, suggesting a correlation between our gene expression profile and previously published results.

Figure 2

Overview of microarray analysis, annotation, and promoter analysis

Overview of microarray analysis, annotation, and promoter analysis. A schematic depicting the workflow of the project. Gene expression analysis of CTLL/WT and CTLL/703 cells was performed utilizing the Affymetrix's Murine Genome U74A GeneChip. Genes that were either up-or down-regulated in CTLL/WT cells by a magnitude of at least two-fold were functionally annotated using NIH's DAVID bioinformatics program. Next, promoter sequences (2100 bp surrounding the predicted TSS) were retrieved from PromoSer. One third of the promoter sequences retrieved were checked for proper alignment against the mouse genome using Blastn and MapViewer tools through NCBI. CREB (TGACGT/C, A/GGGAGT) consensus sequences were obtained through TRANSFAC database and searched within the promoters obtained. Factors that contained the CREB sequences within their promoters were further probed for genes that contribute directly to mitosis, cytokinesis, and microtubule organization.

Additional File 1

wtminus_grtthan_703minus_annotated.

Click here for file

Additional File 2

wtminus_less_703minus_annotated.

Click here for file

Recent studies have utilized microarray analysis to determine whether gene sets were under the control of similar transcription factors by analyzing the promoter sequences for transcription factor binding sites 4243444546. We hypothesized that those genes differentially regulated in CTLL/WT cells would probably be CREB/ATF-dependent either directly or indirectly. This is based on the assumption that CTLL/703 cells contain mutated Tax, which is unable to transactivate viral or cellular transcription through the CREB/ATF pathway 31. To determine which promoters were in fact directly CREB/ATF-dependent, we extracted promoter sequences of the differentially expressed mouse genes utilizing PromoSer, a web-based promoter database extraction tool 4748. This retrieval database utilizes the mm4 draft release of the mouse genome. Promoter sequences encompassing 2000 bp upstream and 100 bp downstream of the predicted transcription start site (TSS) were retrieved. Of the 439 differentially regulated genes, only 341 promoter sequences could be extracted from the database (77% recovery) under the conditions used. One third of the promoters retrieved were initially compared against the mouse genome using NCBI's Blastn to verify that these sequences were properly aligned. Next, promoter sequences were searched for CREB consensus sequences (obtained through the Transfac database and literature searches 49505152) to determine which promoters might be CREB-responsive. Interestingly, 28% of these promoters (95 out of 341 extracted promoter sequences) were found to contain CREB binding sites. These annotated genes with corresponding fold change are depicted in Table 1. Most of the genes identified are involved in a number of pathways including transportation, signaling, cell cycle, transcription and RNA processing, metabolism, stress response, and cytoskeletal protein binding. To determine whether CREB-dependent promoters were preferentially activated, we also performed searches for NFκB and SRF response elements within the extracted promoters. Both of these transcription factors have been shown to be responsive to Tax 53. Of these promoters, only 2% (7 out of 341 promoters) were found to contain NFκB recognition sequences and less than 1% (2 out of 341 promoters) contained SRF sequences (data not shown). Therefore, there was a selective preference in Tax expressing cells to activate CREB-dependent promoters. Through a comprehensive PubMed and DAVID search, we determined which genes were involved in mitosis/cytokinesis and thus were likely candidates to contribute to the development of aneuploidy Table (1).

Table 1

Cellular genes containing CREB response element activated by Tax

Genbank Accession Number

Gene

Fold Change

Transcription/RNA Processing

Peptidase

Cytoskeletal Protein Binding

Transport

M etabolism

Transm em brane Receptor

Signaling

Cell Cycle

Stress Response

AF064071

Apaf1

a po pto tic protease activating factor 1

2.5

D87898

Arf1

ADP-ribosylation factor 1

2.6

BC010700

Atp5c1

ATP synthase, H+ transporting, m itochondrial F1 com plex, gam m a polypeptide 1

2.1

NM_009729

Atp6v0c

ATPase, H+ transporting, V0 subunit C

2.3

AF183960

Ccrn4l

CCR4 carbon catabolite repression 4-like (S. cerevisiae)

2.1

BC054097

Cetn3

centrin 3

2.2

AF016583

Chek1

checkpoint kinase 1 hom olog (S. pom be)

2.3

NM_009942

Cox5b

cytochrom e c oxidase, subunit Vb

2.3

BC010197

Cpe

carboxypeptidase E

17.3

U73445

Dld

dihydrolipoam ide dehydrogenase

2.4

NM_007897

Ebf1

early B-cell factor 1

2.9

D43689

Fdx1

ferredoxin 1

4.3

AJ534939

Sm c2l1

SM C2 structural m aintenance of chrom osom es 2-like 1 (yeast)

2.8

AB093214

Lpin1

lipin 1

2.4

NM_010239

Fth

ferritin heavy chain

2.4

AF024620

Gabrr1

gam m a-am inobutyric acid (GABA-C) receptor, subunit rho 1

2.3

M63801

Gja1

gap junction m em brane channel protein alpha 1

7.4

NM_010306

Gnai3

guanine nucleotide binding protein, alpha inhibiting 3

2.0

BC005683

Grcc10

gene rich cluster, C10 gene

2.2

NM_022310

Hspa5

heat shock 70kD protein 5 (glucose-regulated protein)

3.0

U53514

Guk1

guanylate kinase 1

4.2

M 21931

H2-Aa

histocom patibility 2, class II antigen A, alpha

3.8

BC010322

H2-Ab1

histocom patibility 2, class II antigen A, beta 1

3.1

M 58595

H2-D1

histocom patibility 2, D region locus 1

2.7

U35330

H2-DM b1

histocom patibility 2, class II, locus M b1

2.5

NM_013820

Hk2

hexokinase 2

2.5

BC052727

Hspa9a

heat shock protein, A

3.3

M 59821

Ier2

im m ediate early response 2

2.9

J03236

Junb

Jun-B oncogene

2.0

NM_010724

Psm b8

proteosom e (prosom e, m acropain) subunit, beta type 8 (large m ultifunctional protea

2.6

M16229

M dh2

m alate dehydrogenase 2, NAD (m itochondrial)

3.4

AY176058

Nfkbib

nuclear factor of kappa light chain gene enhancer in B-cells inhibitor, beta

2.6

AF026124

Pld3

phospholipase D3

2.4

BC046832

Ppp1cb

protein phosphatase 1, catalytic subunit, beta isoform

2.5

NM_011186

Psm b5

proteasom e (prosom e, m acropain) subunit, beta type 5

2.3

D87911

Psm e3

proteasom e (prosom e, m acropain) 28 subunit, 3

2.2

X61940

Dusp1

dual specificity phosphatase 1

8.1

NM_008989

Pura

purine rich elem ent binding protein A

2.4

BC026915

Rab6

RAB6, m em ber RAS oncogene fam ily

2.4

U67187

Rgs2

regulator of G-protein signaling 2

2.5

AF065924

Ccl1

chem okine (C-C m otif) ligand 1

0.3

X84037

Glg1

golgi apparatus protein 1

2.4

BC034674

Slc31a1

solute carrier fam ily 31, m em ber 1

2.9

BC021537

Srp14

signal recognition particle 14

2.5

BC028507

Tnfrsf9

tum or necrosis factor receptor superfam ily, m em ber 9

4.2

AF159593

Plscr1

phospholipid scram blase 1

5.2

AF033353

Ubl1

ubiquitin-like 1

2.3

AY162409

Vwf

Von W illebrand factor hom olog

2.3

AF077002

Ywhah

tyrosine 3-m onooxygenase/tryptophan 5-m onooxygenase activation protein, eta po

2.0

U77667

Zap70

zeta-chain (TCR) associated protein kinase

2.6

BC005589

Cfdp

craniofacial developm ent protein 1

2.1

BC008265

Psm b2

proteasom e (prosom e, m acropain) subunit, beta type 2

2.3

AF132726

Casp8ap2

caspase 8 associated protein 2

2.3

AF123312

H2afy

H2A histone fam ily, m em ber Y

2.1

BC012241

Atp5o

ATP synthase, H+ transporting, m itochondrial F1 com plex, O subunit

2.5

BC014798

Tax1bp1

Tax1 (hum an T-cell leukem ia virus type I) binding protein 1

2.3

AF406651

Hnrpa2b1

heterogeneous nuclear ribonucleoprotein A2/B1

2.5

AF098508

Dctn3

dynactin 3

2.4

AF133818

Zfp265

zinc finger protein 265

2.1

BC037732

Rragc

Ras-related GTP binding C

2.7

AF148447

Uchl5

ubiquitin carboxyl-term inal esterase L5

2.3

AB025406

Dstn

destrin

2.6

BC002126

Gabarap

gam m a-am inobutyric acid receptor associated protein

2.1

BC018430

Asna1

arsA (bacterial) arsenite transporter, ATP-binding, hom olog 1

2.2

AB015652

Park7

Parkinson disease (autosom al recessive, early onset) 7

2.1

BC022751

Isg20

interferon-stim ulated protein

4.8

BC005469

Vps35

vacuolar protein sorting 35

2.3

BC005620

Cyc1

cytochrom e c-1

2.2

BC024355

Usm g5

upregulated during skeletal m uscle growth 5

2.2

BC051934

Sdhb

succinate dehy