cancer genomics & proteomics

0 downloads 0 Views 723KB Size Report
1. Please read thoroughly, correct, and return the proofs to the Editorial. Office within 24 .... estimated by the K-nearest neighbor (KNN) approach. When ..... The charges are EURO € 150 per excess page, totalling EURO € 600. We ask you to ...
ISSN: 1109-6535

CANCER GENOMICS & PROTEOMICS Editorial Office: International Institute of Anticancer Research, 1st km Kapandritiou - Kalamou Rd., Kapandriti, P.O.Box 22, Attiki 19014, Greece Fax: 0030-22950-53389, Tel: 0030-2295052945, e-mail: [email protected]

Dear Sir/Madam: Enclosed are the galley proofs of your article for CANCER GENOMICS & PROTEOMICS. We would like to call your attention to the following: 1. Please read thoroughly, correct, and return the proofs to the Editorial Office within 24 hours. 2. Proofs should be returned by the fastest mail service available. Delays in the return of these proofs will necessitate the publication of your paper in a later issue of the journal. You may also send the proofs by fax or e-mail. 3. Please read the entire manuscript carefully to verify that no changes in meaning have been introduced into the text through language improvements or editorial corrections. 4. Corrections should be limited to typographical errors. 5. To promote rapid publication of articles in CANCER GENOMICS & PROTEOMICS, figures are not sent to the Author(s) unless alterations have been made through the editing process. 6. Should you require reprints, PDF files, issues or subscriptions, please fill the attached reprint order form. 7. Should you require information about your article (publication date, volume, page numbers, etc) please call: +30-22950-52945 or send an e-mail to [email protected]. 8. Please provide your complete address (not P.O.B.), telephone and fax numbers. 9. Please feel free to contact us with any queries that you may have (Tel./Fax: +30-22950-53389 or +30-22950-52945, e-mail: [email protected]). Thank you for taking the time to study these guidelines. Yours,

J.G. Delinassios Managing Editor Enclosures

CANCER GENOMICS & PROTEOMICS 7: xxx-xxx (2010)

No: 340-Z Please mark the appropriate section for this paper I Experimental I Clinical I Epidemiological

Effects of DNA Minor Groove Binding Agents on Global Gene Expression MALEK ZIHLIF1, DANIEL R. CATCHPOOLE2, BERNARD W. STEWART3 and LAURENCE P.G. WAKELIN4 1Department

of Pharmacology, Faculty of Medicine, University of Jordan, Amman, 11942, Jordan; Tumour Bank, The Children’s Hospital Westmead, Locked Bag 4001, Westmead, NSW, 2145, 3Cancer Control Program, South Eastern Sydney and Illawarra Public Health Unit, Locked Bag 88, Randwick, NSW 2031, Australia; 4School of Medical Sciences, University of New South Wales, Sydney, NSW 2052, Australia

2The

Abstract. The capacity of two minor groove binding agents that differ in their DNA sequence selectivity to modulate gene expression in human leukaemia cells was investigated. The chosen compounds were the chromomycin A3, a GC selective minor groove binder, and alkamin, an AT selective minor groove binder. As revealed by DNA microarray analysis of 6000 genes, at equitoxic doses, 5× IC50 values for growth inhibition, the two drugs disturbed transcription, resulting in both up- and down-regulation of many hundreds of genes, 24h after drug exposure. Direct comparisons between the most affected genes and also the cluster analysis indicated a relatively low degree of similarity between the tow expression profiles. Moreover, the ontological and the pathway responses also indicated a distinguished biological responses. Chromomycin treated was characterized by many negative impacts on the important cellular functions and by the activation for those functions that usually take the cells towards apoptosis. In the second biological profile the domination of many positive functions might indicate that the cells were attempting to overcome and repair the alkamin assault. Examples on those functions are positive regulation of gene expression, positive regulation of macromolecule biosynthetic processes, the cell cycle pathway and DNA repair. DNA minor groove binding agents can be divided into two groups. First, those compounds that bind to the wide minor groove of GC-rich DNA sequences and are best characterized by the aureolic acid group of antitumour antibiotics such as chromomycin, mithramycin and

Correspondence to: Dr. Malek Zihlif, Department of Pharmacology, Faculty of Medicine, University of Jordan, Amman 11942, Jordan. Tel: +962 65355000, Fax: +962 65356746, e-mail: [email protected] Key Words: Gene expression, DNA binding, minor groove binders.

1109-6535/2010 $2.00+.40

olivomycin (1-3). These agents have experimental antitumour activity and have found clinical use in the treatment of testicular tumours (4, 5) (Figure 1). Chromomycin A3 contains five sugar rings connected to an aureolic acid chromophore via O-glycosidic bonds: a disaccharide is attached at the 6 position and a trisaccharide at the 2 position (6, 3). It binds to DNA as a dimer, complexed through a Mg+2 ion, to GC-rich sequences, a particularly high affinity site being TGGCCA (3, 7). Unlike the complexes of ATselective minor groove binders (8) chromomycin-DNA complexes dissociate slowly, with half-lives similar to those of actinomycin that have a DNA occupation time of 300 sec (9, 10). The biological effects of chromomycin are clearly related to perturbation of DNA and RNA synthesis (11, 12). The second group of minor groove binders are those compounds that bind to the minor groove of AT-rich DNA sequences and are best characterized by the alkamin, which has been designed to maximise crosslink frequencies at sequences such as AATT, AAATTT, and AAAA, and AAAAA. Alkamin is a special aniline mustard minor groove alkylating agent that has been modified with dimethylaminomethyl groups to provide the necessary positive charge to promote reversible groove binding prior to alkylation (12, 13) (Figure 1). Molecular modelling experiments have shown that alkamin, the parent bifunctional polybenzamide mustard, binds to AT-tracts in a manner consistent with these design principles (12, 14). Alkamin is potently cytotoxic, with an IC50 value of 7 nM against P388 mouse leukaemia cells, has in vivo activity against this tumour in a stringent single dose assay and showed a positive hypersensitivity ratio of 15 for activity against a pair of Chinese hamster ovary cell lines designed to reveal whether the mode of drug action involves DNA inter-strand crosslinks (12, 13). The aim of this work was to investigate the capacity of two minor groove binders, chromomycin A3 and alkamin, to modulate gene expression in human leukaemia cells in culture. In particular the possible differential effects of the

3

CANCER GENOMICS & PROTEOMICS 7: xxx-xxx (2010) different DNA sequence selectivity of these agents on the global gene expression was examined. cDNA microarray technology was used by which it is also possible to analyze the gene expression on a genomic scale trying to find associations between characteristic gene expression patterns and molecular responses to drug therapy.

Materials and Methods Cell line. CCRF-CEM, CCRF-CEM, a human lymphoblastoid T-cell leukaemia cell line was obtained from Dr. Maria Kavallaris, Children’s Cancer Research Institute, Sydney, and maintained in complete medium consisting of RPMI medium 1624 (Gibco BRL, CA, USA) supplemented with 2 mM L-glutamine, 100 units/ml of penicillin/streptomycin and 10% foetal bovine serum (Trace scientific Ltd, Sydney, NSW, Australia) at 37˚C, in the presence of 5% CO2. The cultures were passaged twice weekly at which point they were at 70% of their maximum permissible cell density. The CCRF-CEM cells were seeded at a density of 1×105 cells/ml in T-75 tissue culture flasks (Corning, MA, USA) and incubated for 24 h before drug treatment with 5× IC50 concentrations for 24 h. Chromomycin was purchased from Sigma Co. Ltd., (St. Louis, USA). Alkamin was gifts from Professor William A Denny, Auckland Cancer Society Research Centre. The IC50 for alkamin is 16 nM while it is 3.6 nM for chromomycin. The treatment was performed on three occasions for each drug. Control cells were treated in the same way, except they did not received the drug treatments. Following the treatment periods the cells were harvested by centrifugation at 1200 rpm at 4˚C for 5 min. The culture medium was removed by aspiration, the cells washed with ice-cold PBS, and the pellets recovered again. RNA isolation and cDNA synthesis and probe labeling. These were performed as previously described in previous publication (15). Briefly, the total RNA was extracted using the acid guanidinium thiocyanate/phenol method. Using RNA from treated and untreated cells, the synthesis of cDNA was initiated by mixing 1.25 μg of oligo dT primers with 20 μg total RNA in a volume of 15.5 μL using diethylpyrocarbonate (DEPC)-treated water and incubated at 70˚C for 10 min. The incubation mixture was then chilled on ice for an additional 10 min, and mixed with a reaction cocktail consisting of 50 mM Tris-HCl (pH 8.3), 75 mM KCl, 3 mM MgCl2, 500 μM each of dATP, dCTP and dGTP, 300 μM dTTP, 200 μM 5-(3-aminoallyl)2’-deoxyuridine-5’triphosphate (aminoallyl-dUTP), 10 μM 1,4Dithiothritol (DTT), 150 U of Superscript II, and DEPC-treated water, all in a final volume of 30 μl. The mixture was incubated at 42˚C for 2 h, at which point the reaction was stopped by the addition of 10 μl of 1M NaOH followed by 10 μl of 0.5 M EDTA, which also served to hydrolyse the RNA, and then incubated for 15 min at 65˚C. The samples were then mixed with Cy5 (red fluorescence) or Cy3 (green fluorescence) dye and incubated for 1 h at room temperature in the dark, the coupling reaction being quenched by the addition of 4.5 μl of 4 M hydroxylamine followed by incubation for a further 15 min at room temperature. Un-incorporated dye was removed from the cDNA preparations by passage through a QIAquick PCR purification column according to the manufacturer’s instructions (Qiagen, Hilden, Germany). Microarray hybridization, washing and fluorescence imaging. The cDNA microarray consist of 5705 sequence-verified known human

4

genes and 420 control genes, some human, some yeast, spotted on to Telechem® slides (Genetix, New Milton, UK) in duplicate features (total of 12250 spots). The chips were purchased from the Ramaciotti Centre for Genome Function at the University of New South Wales. Following hybridization, the microarray was washed in a pre-warmed (50˚C) 1× saline sodium citrate (SSC) solution containing 0.03% sodium dodecyl sulfate (SDS) for 5 min followed by successive 5 min washes in 0.2×SSC and 0.05×SSC at room temperature. The microarrays were scanned on an Axon II Scanner with a multi-channel image generated which was subsequently analysed with Genepix software (Axon - Molecular Devices, Sunnyvale, CA, USA). Data analysis. Each full experiment, starting from drug treatment and culminating in hybridisation of the fluorescently-labelled cDNA to the microarray, was performed on three separate occasions, thereby generating three fully independent microarray data sets for all the compounds studied. For every spot on each array, the fluorescence intensity of the Cy5- (Red, R) and Cy3-(Green, G) labelled cDNA was determined from the scanned array images, after equalization of fluorescence emission intensity for each dye, and subtraction of background the fluorescence. The gene expression values for each spot were expressed as the logarithm base 2 of the ratio of Cy5 (R)- intensity to Cy3 (G)- intensity (log2 R/G). Using Bioconductor software (open software development for computational biology and bioinformatics, http://www.r-project.org), the log2 R/G was normalized across each microarray using a “printtip lowess” computation. This procedure produces a list of the relative intensities, and hence relative expression level, for each of the experimental genes detected in both channels. For each of the three normalized microarray data sets for each drug treatment, in which every gene had a duplicated spot per array, the median log2 R/G for all 6 expression measurements was determined for each gene. The distributions of the log2 R/G values were plotted as histograms and examined using standard descriptive statistics (mean, median and skewness), so allowing comparison of the global effects on gene expression of the drugs studied. The missing values were estimated by the K-nearest neighbor (KNN) approach. When investigating whether the drugs had similar or different effects on transcription, all the genes whose absolute variation in level of expression was by a factor of 3-fold, as revealed by those genes whose log2 R/G was ≥±1.5 were compared. All the genes on the array were subjected to unsupervised hierarchical clustering, the extent of the similarity and differences in expression profiles between the drug treatments being based on the correlation coefficient used within the software ‘Cluster’ and displayed as heat maps using ‘TreeView’ (EisenLabs, USA). The gene ontology and pathway database was interrogated via the web-based application DAVID (http://david.abcc.ncifcrf.gov) which provides for ontology and pathway mapping, annotation, and visualization of results. A level 5 search in the “biological process” category was carried out to provide the highest degree of specificity for the target function. For the pathway analysis Kyoto Encyclopedia of Genes and Genomes (KEGG) were used.

Results Impact of each drug on the global gene expression. Of the Log2 R/G values for each drug the standard descriptive statistics are presented in Table I. The histograms of the two

Zihlif et al: DNA Minor Groove Binding Agents on Global Gene Expression

Figure 1. Structure of alkanim and chromomycin.

drugs showed that the gene expression levels could remain unchanged, be elevated or be diminished. The two histograms shared negative means and medians indicating that the number of genes in the down-regulated region was greater than the number in the up-regulated region. Interestingly, alkamine exhibited lower mean, medium and skewness values, all of which indicated that alkamine had a greater inhibitory effect on global gene expression than chromomycin. This superiority in the inhibitory effect was clearly seen in the higher number of genes that were downregulated 2-fold or more. The number of up-regulation genes that scored more than 2-fold was sinilar for the two agents. Commonality between the chromomycin and alkamin. Surprisingly few similarities were found between the two treatments, especially in the down-regulated part, in which only 151 common genes scored more than 2-fold of inhibition. In the up-regulated genes, the commonalities were also relatively low reaching 128 shared genes with 2-fold or more. These findings prompted a broader analysis to see whether this behavior applied only to genes identified above or was applicable to the 5704 genes on the array (Figure 2). The broader analysis, an unsupervised hierarchal cluster analysis had the advantage of demonstrating all possible relationships between the examined agents at once. Clearly the results (Figure 2) revealed that each agent had a unique effect on the 5704 genes examined on the array. However, the heat map did demonstrate some interesting shared regions especially in the up-regulated genes. Ontological analysis. The ontological analysis for the upregulated genes identified 19 biological classes containing fifteen or more entries for at least one of the treatment (Table II). Although with both agents large proportion of the affected genes lay in the leading ontological groups such as cellular protein metabolic processes, intracellular signaling cascade and phosphorylation, unique behavior for each agent was clearly seen. Interestingly, the two profiles exhibited opposite behaviour. For example, alkamin treatment

Table I. Distribution statistics for global gene effects of alkamin and chromomycin.

Mean Median Standard deviation Maximum Minimum Skewness (a) No of genes with Log2 R/G >1 (b) No of genes with Log2 R/G