www.perkinelmer.com/lifesciences

Drug Discovery Research Clinical Screening

Imaging

CY

CL

ON

E

GE

NE

A

RR

AY

A

PP

LI

CA

TI

ON

N

OT

E

CY

C-

00

1

Cyclone® Gene Array System

Quantitative Analysis of DNA Arrays

By Loraine Upham, PerkinElmer Life Sciences, Inc.and Tony Fox,Glaxo SmithKline

IntroductionGene array methods are rapidly replacing other molecular biology methods for applications such as differentialgene expression, mutation screening, sequence analysis and drug target identification, validation, and secondaryscreening analysis. Miniaturization, accuracy, and speed are a few of the most attractive advantages of gene arrayanalysis as compared to previous methods involving microplates or electrophoresis1. Preprinted gene arrays onnylon membranes from Clontech, Incyte, Sigma Genosys and Research Genetics, among others, provide a conven-ient way to begin using DNA array methods. Using radiolabeled cDNA and the high resolution Cyclone® StoragePhosphor system, arrays provide sensitivity and quantitative accuracy required for large scale gene expressionstudies.

Radiolabeling has several advantages over non-isotopic methods for array analysis including no interference fromdust, no photobleaching effects or pollution of the sample emission signal2. The Cyclone Storage PhosphorSystem has been shown to provide high resolution images for in situ hybridization3, receptor autoradiography4

and a number of molecular biology applications5,6. With 42 µm pixel capability and highly sensitive storage phosphorscreens, the Cyclone scans images with over five logs of linear dynamic range for comparison of both weakly andstrongly expressed genes in one exposure. QuantArray® Analysis software has advanced microarray analysis toolsfor registration of up to four images, spot location, display and data export of spot to spot comparisons. Gene listsand templates for particular filters can be imported and used for sample identification and archiving. Scatterplotsand other display tools help to visualize and analyze results.

Through collaboration between Packard BioScience and Packard BioChip Technologies, the Cyclone is now avail-able with QuantArray software. The following report describes the use of Cyclone and QuantArray software forquantitative analysis of DNA macroarrays.

MethodsGene Expression Sample Preparation Fenofibrate is a ligand for the PPARα nuclear receptor7. PPARα is a member of the family of transcription factorsknown as Peroxisome Proliferator-Activated Receptors, which play a central role in regulating the storage andcatabolism of dietary fats8. PPARs form heterodimers with another nuclear receptor, known as the 9-cis-retinoicacid receptor (RXR), and bind to direct repeat DNA sequences known as DR-1 response elements in the targetgenes9,10. When agonist ligands bind to the PPAR/RXR heterodimer, recruitment of coactivator proteins leads totranscriptional activation11 (Figure 1). Fenofibrate induces liver tumors in rats at high doses, although at therapeu-tic doses in humans, there has been only clinical evidence of some adverse effects related to use of fenofibrate forhyperlipidemia. The study of gene expression changes associated with high doses of fenofibrate in rats may pro-vide clues as to the potential causes of adverse effects in humans (Fox, publication in progress).

2

Figure 1. Fenofibrate is a ligand for PPAR receptors, which form heterodimerswith RXR receptors and bind to DR-1 response elements and lead totranscriptional activation.

Atlas™ Rat Toxicology II arrays are filters containing rat liver total RNA obtained from Clontech (Cat. # 7732-1).These filters feature gene specific primers that are less complex and provide less cross-hybridization than arraysthat utilize probes generated using oligo(dT) or random primers (www.Clontech.com). They consist of 465 uniquecDNA fragments in duplicate. Rats were exposed to 200 mg/kg/day Fenofibrate treatment for ten days. RNAswere reverse transcribed into cDNA and radiolabeled with 33P-dATP. Gene expression profiles of the control ani-mals and the treated animals were analyzed for differences in gene expression in response to the high doses ofFenofibrate.

Image AcquisitionIn traditional film cassettes, the control samples were exposed for 18 hours and the Fenofibrate treated sampleswere exposed for 24 hours using the Cyclone SR, SuperResolution screen. Screens were scanned in the CycloneStorage Phosphor System using the highest resolution scanning mode which provides 42 µM pixel resolutionusing the unique, helical scanning mechanism and confocal optics. Images are saved automatically after the scanis complete in less than ten minutes (Figure 2a and 2b). Cyclone software includes image display and handlingfeatures such as cropping, magnifying, rotating and exporting for viewing, and archiving files. In addition, inten-sity of spots can be determined using the Cyclone grid analysis feature. However, for more advanced analysissuch as comparisons of control and treated samples, we recommend QuantArray software.

Figure 2a.

3

Figure 2b.

ResultsAnalysis with QuantArrayThe following is a step-by-step description of the use of QuantArray to analyze Cyclone images specifically.

1. File, Protocol Wizard.This is the best way to create the grid necessary for quantification and comparison of the spots. Protocol wizardtakes you through step by step to define the arrays and locate them properly on the registered images.

A. File, Open Images.Cyclone images may be opened directly into QuantArray software for analysis. Display contrast should be set toover 90% using the contrast display feature so that the lower activity spots can be visualized clearly.

B. Register Images.This feature enables one to overlay the control and treated images to obtain data relating to the differences.Although control arrows on the left side of the screen allow one to translate the images in vertical and horizontaldirections with respect to each other, there is no ability to rotate. Therefore, it is important filters are exposed toCyclone screens in a straight orientation, so that Cyclone images may be registered properly.

Cyclone images of rat liver total RNA hybridized with control (2a) and Fenofibrate treated (2b) rat liver total RNA reverse transcribed into cDNA and radiolabeled with 33P-dATP.

C. Create an Array Pattern with Protocol Wizard.It helps to know what the spot pattern is from the filter manufacturer and how many spots are there, so that youcan be sure that the grids are appropriate (Figure 4). Based on the grid created in QuantArray and the specifica-tions from Clontech, the 1176 genes are represented by spots in the image.

4

With both images open, align the spots using the arrow keys to get the best overlay. Contrasting color displayscreate color indicators of correct alignment (Figure 3).

Figure 3. Cyclone images displayed in QuantArray. Control image in green, treatmentimage in red and overlayed images shown in combined colors.

Figure 4. Protocol wizard helps to create an appropriate grid for quantitative analysisof data from filters.

3. Set Quantitation Parameters.Analyzing Cyclone images may be done using any of the three methods including Histogram, Fixed Circle, orAdaptive. Fixed circle is the one most like the OptiQuant™ grid quantitation feature and is used here for the purposes of manually defining the spot area and background area used.

Fixed Circle quantitation parameters may be set to mimic grid analysis in OptiQuant, so background inner andouter diameters were set to be close and outside the range of the activity of the spots. Signal was defined as 0-100and background was defined as 0-5. Total intensity was chosen as the quantitation output.

2. Locate Spots. Back in the main window, the location can be specified and edited for the best fit to the registered images usingthe cursor to place the small box in the uppermost left spot of the uppermost left array. Hit the start locate buttonto automatically locate the activity of each spot. When it is completed, each individual spot location can be edit-ed if necessary. Using the default parameters, the spots were accurately located with no corrections necessary onthis filter (Figure 5).

Figure 5. Spots located automatically with QuantArray.

5

4. View Reports. Reports can be viewed in a number of graphical ways including a scatter plot as shown in Figure 6a. The scatterplot is a graphical representation of the comparison of two images. Each point represents one spot on the image.The x-axis refers to the intensity in the control image and the y-axis refers to the intensity of the treated image.Below the scatter plot are details about whatever point is highlighted in the scatter plot (Figure 6a and 6b). Datamay be exported directly to Microsoft® Excel for archiving and further analysis. The scatter plot is a great visualway to check out immediately what genes are different between the two filters. Just click on the + of each spot inthe graph and get the relative intensities reported and the thumbnail image of the two spots.

6

Figure 6a. Scatterplot display of comparison of controland treated filters based on data fromCyclone.

Figure 6b. Close-up of representation of specific spots asselected in the scatterplot shown in Figure 6a.

5. Export Data. Data is exported to Excel for additional analysis. The gene database file was obtained from Clontech’s website atwww.Clontech.com and specified so that each specific gene may be identified.

Export filters may be invoked also at this time to create upper and lower thresholds. Normalization features arealso provided for post-processing in Excel, when filters are not exposed for the same exposure time, on the samescreens or when specific activity of probes is significantly different.

Discussion and ConclusionsThe control and fenofibrate treated filters were analyzed in QuantArray using the normalization to total macro inExcel from QuantArray. A total of 40 genes were expressed in the treated filter at a level that exceeded twice thecontrol filter expression level. Another 25 genes were expressed at less than 70% of the expression level shownby the control filter. The significant genes that showed increased expression include:

• copper-zinc-containing superoxide dismutase 1 (Cu-Zn SOD1)

• very long chain acyl-CoA dehydrogenase precursor (VLCAD)

• ribosomal protein L13A, long chain-specific acyl-CoA dehydrogenase precursor (LCAD; ACADL)

• 40S ribosomal protein S19 (RPS19)

• liver fatty acid-binding protein (L-FABP); Z-protein; squalene- & sterol-carrier protein (SCP)

• heat shock 60-kDa protein (HSP60); 60-kDa chaperonin (CPN60); GroEL homolog; mitochondrial matrix protein P1; p60 lymphocyte protein

• peroxisomal membrane protein 1 (PXMP1; PMP1); 70-kDa peroxisomal membrane protein (PMP70); ATP-binding cassette subfamily D member 3 (ABCD3)

• acyl-CoA oxidase

• malic enzyme

• cytochrome P-450 4A1

• hydroxyacyl CoA dehydrogenase

Among those genes with lower expression under the treatment condition are:

• cytochrome P450 IA2 (CYPIA2); P450-D; P448 + cytochrome P450 ISF/BNF-G

• apolipoprotein CIII precursor (APOC3)

• fibrinogen gamma chain

• alpha-1-antiproteinase precursor; alpha-1-proteinase inhibitor; alpha-1 antitrypsin

• phosphoenolpyruvate carboxykinase

• apolipoprotein AI precursor (APOA1)

These are being further studied individually to confirm their up or down regulation and to determine their possi-ble role in production of liver tumors.

Cyclone features a unique helical scanning mechanism with a particularly short distance between the screen andconfocal optics. This design provides quantitatively accurate, high-resolution images with sensitivity to detectchanges in gene expression. Shorter exposure times and the long linear dynamic range of the phosphor screensprovide quantitatively accurate data without the possibility of overexposure. QuantArray is a fast and easy-to-usesoftware package which enables one to overlay images, locate spots, visualize differences in expression, andexport data for further analysis. Compatibility between the Cyclone and QuantArray software creates a uniqueand affordable solution for analysis of radiolabled DNA arrays.

7

www.perkinelmer.com/lifesciences

References1. Schena, Mark. Microarray Biochip Technology. Eaton Publishing, Natick, Massachusetts. 2000.

2. L’Annunziata, M. F. Handbook of Radioactivity Analysis. Academic Press, New York. 1998.

3. McGreevey, L., and Seeley, R.J. “Phosphor Imaging System Used to Quantify Hypothalamic mRNA changesDuring Fasting.” BioMedical Products vol.23, No.11, p.42. Nov, 1998.

4. Lidow, M. S., and Solodkin, A. “Use of The Cyclone Storage Phosphor System for Rapid Development of aReceptor Autoradiographic Assay.” BioMedical Products vol.22, No. 10, p.4. Oct, 1997.

5. Devor, E. J., Manthey, A. J., Behlke, M. A., and Walder, J. A. “Use of the Cyclone Storage Phosphor System forQuantitative Image Analysis to Develop New DNA Sequencing Primers.” BioMedical Products vol.28, No.8,p.28. Aug.

6. Saban, M.B., Hellmich, H., Nguyen, N-B., Winston, J., Hammond, T.G., and Saban, R. “Time Course of LPS-induced gene Expression in mouse model of genitourinary inflammagion” Physiological Genomics 5:147-160, 2001.

7. Gaw, A., Packard, C. J., and Shepard, J. “Fibrates” Handbook of Experimental Pharmacology. 109: 325-348, 1994.

8. Mangelsdorf, D. J., Evans, R. M. “The RXR heterodimers and orphan receptors.” Cell. 83: 841-850, 1995.

9. “A unified nomenclature system for the nuclear receptor super family.” Cell 97: 161-163, 1999.

10. Kleiwer, S. A., Umesono, K., Noonan, D. J., Heman, R. A., and Evans, R. M. “Convergence of the 9-cis retinoicacid and peroxisome proliferator signaling pathways through heterodimer formation of their receptors.”Nature. 358: 771-774, 1992.

11. Willson, T.M., Brown, P. J., Sternbach, D.D., and Henke, B. R. “The PPARs: From Orphan Receptors to DrugDiscovery.” J. Medicinal Chemistry. 43, no 4. : 527-550, 2000.

Worldwide Headquarters: PerkinElmer Life Sciences, Inc., 549 Albany Street, Boston, MA 02118-2512 USA (800) 551-2121European Headquarters: PerkinElmer Life Sciences, Imperiastraat 8, B-1930 Zaventem Belgium

Technical Support: in Europe: techsupport.europe@perkinelmer.com in US and Rest of World: techsupport@perkinelmer.comBelgium: Tel: 0800 94 540 • France: Tel: 0800 90 77 62 • Netherlands: Tel: 0800 02 23 042 • Germany: Tel: 0800 1 81 00 32 • United Kingdom: Tel: 0800 89 60 46

Switzerland: Tel: 0800 55 50 27 • Italy: Tel: 800 79 03 10 • Sweden: Tel: 020 79 07 35 • Norway: Tel: 800 11 947 • Denmark: Tel: 80 88 3477 • Spain: Tel: 900 973 255

Atlas is a trademark of UCLA Laboratory of Neuro Imaging. Microsoft Excel is a registered trademark of Microsoft Corporation in the United States and/or other countries.All other trademarks or registered trademarks are the property of PerkinElmer Life Sciences, Inc.

P11358 © 2002 PerkinElmer Life Sciences, Inc.