Microarray Analysis of Sickle Cell Disease - Platelets - Lab Report Example

Running head: MICROARRAY ANALYSIS Microarray Analysis of Sickle cell disease: platelets Name Course Institution May 16, 2011 Abstract This report shows platelet transcriptome analysis from single donors. The analysis was done through validating and developing large gene lists that are statistically significant of genes differentially regulated in patients suffering from sickle cell. Studies in this report provide a view into molecular basis for platelets that are dysfunctional in the sickle cell disease. Introduction According to Schena (2003), hemoglobin S-containing erythrocytes are entrapped in microcirculation in patients suffering from sickle cell. This results in infarction and ischemia-reperfusion tissue injuries that are repetitive. Contribution to the above process is by endothelial activation, primary and secondary inflammation, adhesion molecule expression and oxidant stress. Arganise and erythrocyte are released into plasma due to premature hemolyzing of hemoglobin S- containing erythrocites. A reaction between hemoglobin and nitric oxide derived from endothelium inactivates the NO. Young, Gerson & High (2006) argue that arginine is converted by arginase to ornithine. The processes lead to an endothelial dysfunction which results in pulmonary hypertension development. Pulmonary hypertension is the current greatest risk factor of causing death in patients suffering from sickle cell. Ratner (2004) suggests that better understanding of intrinsic signaling pathways, which generally affect platelets transcriptome in patients suffering from sickle cell, would lead to identification of candidate gene environment interaction or novel therapeutic targets. Advancement in genomic technologies have enhanced the opportunities for large scale cell type specific molecular profiling through employment of microarray technology. This technology rapidly identifies cell systems molecular interactions that are associated with sickle cell disease. However, technical difficulties of gene transcript analyzing have hampered the studies and this is due to RNA’s low abundance in platelets. Technologies of High fidelity RNA amplification have recently been developed which enable small sample analyzing (Schena, Shalon, Davis & Brown, 1995). Alongside other methods like RNAi knockout and RT-PCR, Microarrays define the capability of generating data that is parallelized and of system wide high quantities (Ratner, 2004). Through microanalyses, researchers are capable of studying all mRNA of a tissue unlike in the other methods which only allowed study of an mRNA and a single gene at a time. With that capability, it is possible to do a comparison between the level of all gene expressions of tumor and normal cells. It also enables the researcher to do an evaluation of the expression of levels of organisms’ different tissues, carry out study on the variability of population, test resistance to drugs, genotyping and determine the effects of the environment on levels of gene expressions (Schena, Shalon, Davis & Brown, 1995). Microarray has a substrate with thousands of probes adhered to it. The probes include cDNA and oligonuclides. Micrometric distances separate the probes from each other. The sample targeted hybridizes to the probe in case they are complementary. Microarray experiments are interpreted by labeling the target using a radioactive element of a fluorescent dye. Spotted arrays on a glass slide and the GeneChip systems are the two main microarray technologies that are commonly used. The Serial Analysis of Gene Expression SAGE is the third method though it is not used commonly. Running of SAGE is done on an equipment of standard DNA sequencing and measurement of mRNA, is done by concentrated fragments sequencing of their cDNA. The number of cDNA fragments is equivalent to mRNA’s abundance in the sample. The Affymetrix’s GeneChip uses a process called photolithographic that is the same as production of computer silicon chip. Control of the oligo’s light driven synthesis on the chip surface is by masks. Every oligo is comparable to the length of 25 nucleotides and on each chip 40 oligos are present to help in detecting a gene product. Approximately half of oligos perfectly match regions that are characterized by genes (Schena, 2003). Relative abundances of two targeted sets of mRNA are provided by spotted arrays. The mRNA sets compete in order to hybridize with probes. A small sample is transferred by a robot to a glass plate coated with polylysine from microtiter plate. Reimers claims that the robot is used for spotting. Though Affymetrix chips are more uniform than spotted arrays, spotted arrays allows high flexibility because any probe can be spotted on the array through to designing. In this regard, spotted arrays are important especially in organisms that are not covered by Affymetrix. Schena (2003), views that the highly used spotted array experiment constitutes competitively hybridizing two samples. Using cyanine -5 (Cy-649nm) and ubiquitous fluorophore cyanine -3 (Cy3) every sample is tagged with fluorophore that is different. Reading of the slide is by a scanner that excites fluorophores using lasers. Because hybridization is competitive, a sample that has a higher mRNA expression attaches more to its probe while its detection is through the intensity of fluorophores emissions. Background of the Analysis Schena, Shalon, Davis & Brown, (1995) claim that argine metabolic pathways changes indicated by platelet transcriptone are a revelation of patients suffering from sickle cell disease. Homosapiens suffering from sickle cell diseases have vasculopathy and endothelial which are characterized by arginine bioavailability and reduced nitric acid produced by intravascular hemolysis and ischemia-reperfusion injuries. Recently carried out studies on the functions of platelets in patients suffering from sickle cell diseases, show a state that is basally activated (Ratner, 2004). This indicates that sickle cell disease vasculopathy are because of pathological platelet activation. An already existing microarray database reveals a subset of 72 relatively platelet specific genes and 220 abundant platelet genes. These are defined by >10-fold increase in expression in comparison to the medial of other types of cells in amplified transcripts database. In this regard, I carried out a study to determine transcriptional signaling pathways experienced in platelets that are disregulated in patients with sickle cell. In the study, I validated the feasibility of studies on comparative platelet transcriptone on patients through use of RNA amplification. After that a microarray analysis on probe set was done. Methods Data Processing and Analysis using Microarray. Percent present calls available on hybridized Affymetrix chip and signal intensity were calculated by a version 1.4 Affymetrix GeneChip operating software. Selection of differentially expressed genes between patients and subjects of health control was done by transforming signal intensity values by quantile normalizing transformation and adaptive variance- stabilizing. In order to identify relatively platelet specific genes and platelet abundant, I did a comparison between amplified platelet transcriptome’s gene expression profiles and other amplified profiles got from other cell types. The GeneChip software reported the log10 transformed and medium- normalized low signal intensities. In order to characterize each cell type’s expression I calculated each probe’s median Lmed and regarded the top probe sets as platelet abundant. I did a computation of relative platelet expression index for identification of genes that were specific to platelets. I also classified as platelet abundant probe sets that had relative platelet expression of >1.0. In validating the procedure of amplification, I compared platelet abundant genes with gene expressions previously published. Microarray analysis is by R, which is the de facto platform for carrying out the analysis. Bioconductor avails various packages and this forms R’s strongest point in analyzing genomic data. In the experiment, I used the Affymetrix GeneChip bovine arrays. In the expression array I used it consisted of 24,128 probe sets. 11,255 represented gene identities; annotated UniGene identities were 10,775 while control probes were 133. Every probe had 25-mer short oligos, perfect matches were 11 and mismatches were as well 11. A mismatch was in the base position 13. With the above set it was meant to target same gene which are in different regions. The data set consisted of 5 treatment slides and 5 control slides. Importation of data into R Since most of the data from platforms came in as text files using the function of read. Table, I imported them into R. The Affymetrix data was complicated because data files were in a binary format. In that regard I used the affy library’s function –readAffy- in reading the Cel files hence I came up with ten arrays that were bovine. The cdf-chip definition files helped me describe the oligos layout on the slides and the ones that formed a probe set. After importing data into R, I was able to do an evaluation of its quality. According to Ratner (2004) microarrays usually exhibit systematic and experimental variability of high levels which are not related to contrasts in the experiment. So as to come up with accurate I did an identification of the effects and handled them adequately. The bulk of the analysis was in the extensive steps of preprocessing for determining the slides quality as well as methods of calibration to eject variations that are spurious. Given that bad quality slides provide intensity measurements that are unreliable and affect the final outcome, I identified and removed them from the analysis. Results In consistence with the previously done studies on analysis of platelets in patients suffering from sickle cell disease revealed signs of platelet activation. Patients with sickle cell had their platelets expressing higher percentages of activated glycoprotein. Ways of evaluation data quality are numerous, but on choosing which to use I picked the ones that were uniform between and within the slides. Problematic slides were identified through quantitative and qualitative control measures. I generated image plots with the affyPLM package which fitted the Affymetrix data to the probe level. The affyPLM made it easy foe me to pick spatial effects from the data. Figure one showing bad quality slides, Apart from the slide on top left, the remaining are because of deriving from a PLM model that I fitted. In the top right slide, the light green color indicates regions of high weights while the low weights are indicated by dark green colors. The bottom left slide represents residuals; the blue color represents negative residuals while the red color represents positive residuals. Values of residuals are indicated by the intensity of coloring. White coloration reflects residuals with values close to zero. Sign residuals are represented by the bottom right slide; these were residuals without intensity scaling- red indicating positive and blue indicating negative. A probe level model was fitted with fitPLM after that four plots were made for slide number one. Using the raw data, I initially plotted log intensities and later weights that were used in regression with an aim of down-weighting outliers. I then plotted residuals before finally plotting the residual’s signs where I used two colors one for negative residuals and the other for positive residuals which easened spatial effect detection. Provan (2010), claims that relative log expression is commonly used as a QC measure. Using this measure, I plotted M values of every array in per the pseudo-median array using Function Mbox. To color coding every treatment, I used treatcol variable. I also made histograms and boxplots of raw log intensities. I plotted correlations between slides, which assisted me in identifying outlier arrays. In addition to that, I was able to detect whether the sample treatments grouped together. In regard to principles compared to slides that are not within treatment groups, those within are supposed to be more similar. In verifying that, I used the perfect match intensities as well as the matrix Plot. The data got had spatial effects and also so much variations between slides. I did a further analysis on the slides in order to compare to the existing results. Not all the slides underwent further analysis I used a combined QC measure to determine the slides for discarding. Preprocessing After discarding the bad quality slides I took those of good quality and treated them through the calibration steps. The steps consisted: background correction that aimed at removing intensity measures that were not important to target. The second was normalization which helped in cross array comparisons. This was attained by adjusting the intensity’s overall distribution hence making them the same across slides. Ratner (2004) puts it that incase new slides are added to a normalized experiment, renormalizing of the entire set is necessary. In the third step I did a summary that was specific to Affymetrix GeneChips. I did a summary of the probe set into one intensity value for every target on every array. I used various summarizing methods that included MAS 5.0, GCRMA and RMA that needed the library affy. Normalized data was very essential in stabilizing variances and means. I then plotted the data’s first two principle components on a graph which indicated similarity in samples of the same treatment. In approximately over a half of the samples, I filtered out absent and marginal probes and also the control probes. Control probes were extraneous, were spiked in concentrations that differed and I used them for intensity testing. I removed the control probes immediately I finished the QC stages. The average and absent probes were also removed to minimize the number of tests carried out. Linear model modeling that is from the package limma {55} was used to test for differential expression. In using this model I used a contrast matrix and design a matrix in defining the comparisons of interest. In the design matrix I defined samples as per their treatment and in the contrast matrix I defined that I wanted to test control x treatment. Discussion Sickle cell disease pathogenesis results from hypercoagulability and overt thromboembolism. Though platelets have always been regarded as phenotypically invariant, they are not. They have a characteristic of the ability of exhibiting gene expression patterns that vary. I therefore did a characterization of the global transcriptome from sickle cell patients. Platelets in people with sickle cell had mRNAs levels that had been altered. In those platelets, various enzymes were encoded in argeinine metabolism, which is a potential cause for polyamine level and arginine biodiversity reductions. Though platelets are anuclear, platelet gene transcript alterations in patients suffering from sickle cell are an indication that: an alteration of the stability of mRNA’s performance. Analysis from global transcriptone revealed that platelets exhibit different patterns of gene expression. The study found out that dysregulated arginine metabolism in combination with platelet transcriptones analysis suggest that in sickle cell disease platelet dysfunction involves a pretranslational level mechanism. Conclusion The study clearly revels that platelet arginase, polyamine metabolisms and altered arginine have a potential pathogenic role in sickle cell disease. A good framework for studying disease-specific biology is also given through this study. References Provan, D., (2010). Molecular Hematology, Edition3, John Wiley and Sons, 2010 Ratner, B. D., (2004). Biomaterials science: an introduction to materials in medicine, Academic Press Author Edition2, Reimers, M., An (Opinionated) Guide to Microarray Data Analysis National Cancer Institute, and Karolinska Instutitute, Dept. of Biosciences Schena, M., (2003). Microarray Analysis ISBN: 978-0-471-41443-8 Hardcover 664 pages Schena, M., Shalon, D., Davis, R. W., & Brown P. O., (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270, 368-371. Young, N. S., Gerson, S. L., & High, K. A., (2006). Clinical hematology, Elsevier Health Sciences. Read More