SEARCH

SEARCH BY CITATION

Historical perspective

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

Although microarray as a research tool seems to be new, the idea of detection of multiple gene expression during one experiment has been present for many years, perhaps starting with reprobing Northern blots with various probes. More recently, RNase protection assay with multiple probes, differential display and serial analysis of gene expression (SAGE) (1) were used. These techniques, although laborious, allowed expression detection of up to 20 or so genes in one RNA sample.

The current response to this demand was cDNA microarray developed by Patrick Brown and his group in 1994 (2, 3). That was based on cDNAs encoding genes of interest synthesized using standard reverse transcription technology and printed on glass slides using an in-house automatic printer. RNA was reverse transcribed using labeled nucleotides and hybridized to cDNA. The intensity of fluorescence was read by a computerized array reader. This system was able to detect expression of 1024 genes in a sample.

A commercial response was delivered early by Affymetrics, which developed a chip using 22–25 base oligonucleotides placed on a plastic surface and using 18–30 oligonucleotides per gene assayed. mRNA is reverse transcribed to obtain cDNA. The cDNA is then transcribed with biotin-labeled nucleotides to create biotin-labeled cRNA. Oligonucletide chips are hybridized with biotin-labeled cRNA and the abundance of specific transcript was obtained by reading the chip in a laser scanner. Total fluorescence obtained according to the complex algorithm from sense and missense probes is proportional to gene expression in the original sample.

These two technologies (oligonucleotide array and cDNA microarray) have been developed in parallel and may be used in parallel or sequential fashion. Although both have advantages and drawbacks, they set up a standard in whole genome expression assessment. The following paragraph summarizes the basic aspects of both technologies and delivers a comparison of them.

cDNA microarray

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

cDNA microarray (also known as a spotted array) was developed by groups led by Patrick Brown and Ronald Davis at Stanford University (Stanford, CA) (4, 5). The system is based on standard reverse transcription and PCR amplification followed by cloning of cDNA for each gene of interest. cDNA clones are then ‘printed’ on glass slides using a robotic printer. Several printing technologies were developed allowing, at least theoretically, to obtain an array with 100 000 spots. Each spot corresponds to one gene. Therefore, this system allows obtaining a customized array (6). mRNA obtained from biological samples using various methods is then reverse transcribed and the second strand of cDNA is also obtained in order to achieve double-stranded cDNA using primers end-labeled using fluorescent or radioactive tagged nucleotides and hybridized to the array. The fluorescence or radioactive emission from each ‘spot’ is proportional to the relative specific transcript abundance in the pool of cDNA hybridized to a cDNA spot. Abundance of a specific transcript in a sample is compared to the abundance of that transcript in a control sample. This technique, although laborious, delivers a high-density, completely customized system with high detection specificity. Materials to obtain a single set for spotted array are inexpensive, although robotic plasmid isolation systems, spot printer and scanner usually are parts of a core facility. The disadvantages comprise difficult clone handling and, in some cases, the system may not detect alternate splicing and does not correct for presence of SNPs in transcripts. Because the abundance of a transcript in a test sample is compared to that in a control sample, only relative abundance is determined.

Oligonucleotide microarray (GeneChip)

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

This technology was developed by Fodor in 1991 (7–9). It is patent protected by Affymetrics (Santa Clara, CA). In this system, oligonucleotide probes are directly synthesized on a solid surface using a chemical synthesis supported by a lithophotographic technique. Originally, this strategy was developed to detect DNA SNPs (9), later it was applied to measure mRNA expression. Each gene is represented by 20 sense 25-base-long oligonucleotide probes and as a control by 20 missense probes that have one base mismatch in the central part of the probe – usually a base is substituted with a complementary one. Probe sequences are obtained from publicly accessed data bases and in most cases cover 3′ parts of the coding sequence. This should increase the sensitivity of the assay and correct for single nucleotide polymorphisms. RNA obtained from biological samples is processed prior to hybridization. Double-stranded cDNA is obtained from total or mRNA (DNase treated) using reverse transcription. cRNA is then obtained using in vitro transcription with biotinylated nucleotides and after fragmentation is hybridized to an oligonucleotide array for 16 h at 65°C. Then, if biotinylated nucleotides were used to synthesize cRNA, the chip is visualized by staining with streptavidin–phycoerythrin conjugate. Finally, the chip is scanned using a laser confocal fluorescence scanner where a computer-assisted photomultiplayer detects fluorescence from all the molecules exposed to the laser light on the chip. The value representing expression of each gene is calculated based on fluorescence derived from each sense probe as compared to emission from each missense probe. If the difference between fluorescence emitted from a set of sense probes and missense probes is zero, the particular gene is not expressed in the sample. In early chips, sense and missense probes were situated on the same regions of the chip; recently, they are dispersed randomly on the whole surface of the matrix. The software collects data using an algorithm to find sense and missense probe sets. This would allow for minimizing the possibility of errors connected with uneven hybridization.

Although both spotted array and oligonucleotide arrays have some drawbacks and advantages, they currently seem to be (at least for some laboratories or core facilities) fairly complementary systems (10, 11). Gene chips may be used to screen for expression of a large number of characterized genes as well as ESTs. Spotted arrays, because of their relatively low cost are used to perform multiple arrays. A more comprehensive comparison between spotted and oligonucleotide arrays is presented in Table 1 and Fig. 1.

Table 1.  Comparison between cDNA and oligonucleotide arrays
cDNA arrayOligonucleotide array
Fully customizableBased on available sets of genes
Difficult clone handling and storageNo special storage requirements
Difficult comparison between arraysPossibility to normalize and compare data from different array and experiments
Relatively inexpensive if arrays are produced locally or in a core facilityHigh costs for single experiment
SNP and alternative splicing may influence detectionThe influence of SNPs and alternative splicing on gene expression detection avoided to some extent
image

Figure 1. Comparison between spotted array and oligonucleotide array.

Download figure to PowerPoint

Data analysis

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

Single experiments involving less then 20 gene chips or spotted arrays performed in a relatively short time delivers easily more data that one is able to analyze in months without good computational and statistical support (12). Over the last few years, the technology allowing more spots on a glass slide or more oligos on a gene chip has been developed. In addition, this process was also associated with an implementation of modern statistical and analysis specialized software.

Several software packages were developed to collect and analyze data including Microarray suite (Affymetrix), GeneSpring (Silicon Genetics) or Partek Pro (Partek Inc.). Our group has found JMP software (SAS Institute Inc.) useful, whereas others are using SAS (SAS Institute Inc.), SPSS (SPSS Inc.) or standard MS Excel packages. A few different approaches of analysis of gene expression detection by microarray have been developed so far.

Principal component analysis

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

This simple method allows one to answer fundamental questions – how different are the data based on numerous factors. More importantly – what factors account for the difference. This is basically a screening process that delivers information about whether experimental conditions such as array lots, operator, dates or sequence of data collection had a major influence on gene expression data. It also allows one to determine if different chips clustered based on the same experimental conditions. Principal component analysis might be a very useful tool to assess the quality of data collected.

False discovery ratio and the P value

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

Typically analysis of the difference between groups or experimental conditions uses a Student's t-test, a multifactor anova, time series analysis, or other relevant nonparametric statistical tests to compare data. In microarray analysis, we are dealing with the problem of multiple comparisons, where Bonferroni adjustment may have to be applied. This would result in a relatively high P value, making statistical difference practically not possible. Therefore, a more reasonable approach might be the use of the false discovery ratio (FDR), which controls the overall P value for a set of genes and is less conservative than simple Bonferroni adjustment. In this method, all P values are ranged and compared with the FDR threshold. This allows one to choose a group of genes with a predetermined FDR from all genes in a particular experiment.

Self-organizing maps (SOMs)

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

This approach, built into commercially available products such as the Genespring software, allows one to identify groups of genes related by similar changes in expression and then apply more sophisticated analysis strategy. SOMs are often used during screening for expression changes. Although, a SOM delivers a graphical approach it seems to be a reasonable screening tool.

Pseudogene strategy

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

Some gene expression analysis software (e.g. GeneSpring) allow drawing a gene expression pattern on graphical view. This pattern is frequently called a pseudogene. One can change the level of pseudogene expression at all experimental points and using the software search for a gene list including all genes with similar expression patterns with a specified correlation coefficient. The software delivers a list of genes that change in a manner similar to the pseudogene. This allows one to screen for genes correlated to a specific expression pattern.

Hierarchical clustering

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

This approach allows grouping of genes based on similar expression patterns across the experimental conditions. The output of clustering programs is usually a complex dendrogram showing distance (smaller distance, more similarities in expression; larger distances, less similarities in expression patterns) between genes.

Fold change

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

From the very beginning of the research application of microarrays, fold change has been the parameter that has been considered to be the most basic. Owing to an emerging number of publications using this particular measure and confirmatory experiments (often still using Northern blot or RT-PCR systems), the demands for significant fold change have decreased from double to single digits, probably a more realistic approach.

Data mining tools

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

The application of microarray experimental systems may appear to change our approach to experimental design. We traditionally explore possible biological pathways before designing an experiment. In the microarray era we are attempting to study biological pathways based on data obtained. Therefore, there are numerous software applications allowing for determination of which known pathways are involved in a particular experiment. GenMAPP is one of these programs (available at http://www.genmapp.org). This allows for putting microarray data into a net of existing pathways or predicting a new one. A different approach is presented by PubGene (http://www.pubgene.uio.no), which delivers a powerful tool to search and cluster available literature based on specific microarray data. Gene annotation tools such as a gene ontology database (Gene Ontology Consortium – AmiGO; http://www.godatabase.org/cgi-bin/go.cgi) are also available. This would allow adding some data interpretation and/or some explanatory information to the results of an experiment. Data analysis tools were extensively reviewed by Kaminsky et al. (13) and Kerr (14–16).

Confirmatory experiments and data validation

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

Gene expression at the mRNA level does not always correspond with protein expression. Although protein synthesis is a finely regulated process, alternative splicing, the half-life time of message and other aspects of post-transcriptional modification may influence the its expression. mRNA stability might be affected by 3′ UTR sequences forming docking elements for RNA-binding proteins and by influencing the efficiency of translation. Also, intron sequences may have an influence on gene expression – mostly through nuclear protein binding sites serving as enhancers or changes of chromatin structure (17, 18). Moreover, protein levels and activity is modified in many ways such as glycosylation, phosphorylation, receptor shedding, and sequestration to name a few (19, 20). Therefore, although microarray techniques are important tools in molecular biology and medicine, data obtained from ‘chips’ or ‘spots’ should be confirmed at the message level (using Northern blot, RNase protection assay, or real-time PCR), and also at the protein level – employing Western Blotting or two-dimensional electrophoresis supported by mass spectrometry. One should also remember that the final answer usually is delivered at the biological level as a change in a physiological measure, scale score or mediator levels in vivo.

Although, as stated above, the technology is relatively new, several interesting articles employing microarray data have already been published. The following pages consist of a review of a few microarray applications in allergy and immunology. Several diseases and models were studied, in many cases findings have yet to be confirmed by others using similar technology.

Application of functional genomics to studies in allergy

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

In order to locate a susceptibility locus for allergic asthma, Karp et al. (21) employed a well-characterized murine asthma model and whole lung RNA isolation followed by pulmonary gene expression profiling. Microarray studies have shown high correlation between expression of 21 filtered genes and bronchial hyperesponsivenes. The authors identified a high correlation between low expression of complement factor 5 and bronchial hyperactivity. C5 deficiency also led to a decrease in IL-12 production and TNF-α levels (21). The C5 locus in human 9q34 was identified also as an asthma candidate gene by others (22, 23). Suppression of IL-12 production may result in skewing the CD4 response towards the Th2 type (24).

Hamalainen et al., using an in vitro model polarizing naïve CD4 T cells to Th1 and Th2 through incubation with IL-12 or IL-4, respectively, employed microarray analysis of genes expressed by these two lymphocyte populations (25). A total of 20 genes were expressed differentially. IFN-γ, IL-8, GM-CSF, TNF-α, granzyme B, RANTES and several CC chemokine receptors (CCR1, 2 and 5) were highly expressed in Th1 cells as compared to Th2. Studies of Th2 lymphocytes keys higher levels of IL-4, IL-5, IL-13, IL-4R, FGFR, CCR4 and STAT4 and Smad2 expression. The aforementioned work nicely confirmed Th1/Th2 cytokine profiles (25). It also may have an impact on the search for the specific Th1 and Th2 transcription factors. A similar study was done in murine T cells showing an even broader spectrum of differences between Th1 and Th2 CD4 and CD8 cells – more than 100 genes were expressed differently (26). The authors obtained Th1, Th2, and Tc1 and Tc2 cells by incubating purified CD4+ or CD8+ lymphocytes with IL-12 or IL-4, respectively. Apart from the typical Th1/Th2 differences, the authors also described increased expression of c-fos and HLF1α-like factor in type 2 cells and an early B cell factor in type 1 cells. These transcription factors might play a role in the regulation of different cytokine patterns between both groups. Also, TRAF4 (TNF receptor associated factor) was preferentially expressed in type 1 cells and TRAF5 in type 2 cells.

The reported differences in expression of CCR1, CXCR4 (higher in Th2 cells) were also confirmed by others. Th2 cells tend to express higher levels of β7-integrin, whereas Th1 cells express more α4 integrin. Microarray technology also allowed for the identification of a difference in expression of granzymes between type 1 and type 2 cells. Granzyme C appears to be highly expressed in Th1 cells, Tc2 cells expressed elevated levels of granzyme D, E and F. While these differences need to be confirmed, they might be related to differences in migration properties or higher efficiency in clearing viral infections attributed to Tc1 cells (27). Chtanowa and colleagues described also a difference in expression for several ESTs between Th1 and Th2 cells. This group also confirmed that differences in gene expression in Th1 and Th2 cells and Tc1 and Tc2 cells are similar although there are several genes with different relative expression (26). Particularly, cytokine profiles Th1 and Tc1 vs Th2 and Tc2 might be similar. Actual secretion of cytokines by Th1/Th2 lymphocytes was studied using an ELISA microarray (28). A B-cell isotype control and IgE production study in atopic asthmatics and atopics without asthma revealed different expression of 23 (out of 78 studied) genes related to isotype control (29). Interleukin-4 receptor, IL-4, CD40 and TRAF3 were upregulated in both the atopic patients without asthma and the atopic asthmatic group. IL-10 was upregulated in the atopic group only. Interestingly, the EP2 receptor was preferentially upregulated in asthmatics. Expression of IL-6, lymphotoxin-α and TNF-α was increased in asthmatics and atopics as compared to controls. In addition, IL-2R seemed to be upregulated in atopic asthmatics. Interleukin-4, IL-13, IL-5, IL-9 and STAT6 expression were similar in all groups (30). Interestingly, the authors found differences in the TNF-α pathways between atopic asthmatics and atopic nonasthmatics subjects. The atopic nonasthmatic subjects had higher TNF-α expression with unchanged TNFR1 and 2, whereas atopic asthmatics had higher TNFR2 levels only (30). Asthma severity seems to be connected with an increased expression of CD86, TRAF3, ERK1 and MAPKAP, which may depict a higher level of inflammatory response driven through TNF receptors. As mentioned above, in a search for a diagnostic gene-expression profile based score for asthma and atopy – the same group (30) described a group of 10 genes with predictive properties for atopy including IL-1R, IL-6, CRAF1 and serine kinase. These genes are thought to have higher expression in atopics as compared to healthy subjects. Interestingly, integrin-α6 has lower expression in atopics as compared to healthy subjects.

At least in murine models most phenotypic asthma features depend on Th2-related cytokines (31). IL-13 secreted by Th2 cells might play a role in initiating events such as eosinophilic inflammation, mucus metaplasia and airway hyperesponsiveness (32). The effect of IL-13 on human bronchial epithelial cells, human airway smooth muscle cells and lung fibroblasts using an Affymetrix platform was studied by Lee et al. (33). Although IL-13 induces STAT6 phosphorylation in all three cell types – there are no common genes induced by this cytokine in all three cell groups. In airway epithelial cells, IL-13 induced mostly expression of extracellular matrix proteins, proteases and protease inhibitors and CCR3. In bronchial smooth muscle cells, expression patterns comprised PLA2, components of the mitogen-activated protein kinase cascade and CXCR2. Similarly, in human lung fibroblasts expression of MAP kinase cascade proteins was also observed. Moreover, IL-13 induces expression of MCP-1 and IL-6 and VCAM-1 in human lung fibroblasts. This interesting study suggests that IL-13 induced distinct transcriptional programs in different cell types although the signal transduction pathway appears to be related to activation of the same element (IL-4R and STAT6). This may suggest that the asthma phenotype is a result of a concert of effects on various tissues and cells driven by IL-13.

The influence of various cytokines (including TNF-α, IL-1β, IFN-γ) on gene expression in airway epithelial cells has also been studied (34). In general, these studies confirmed known facts, e.g. TNF-α and IL-1β have autocrine properties. These cytokines induce ICAM-1, MMPs, various types MAP kinases and increase heat shock protein expression.

A similar approach was applied to study the influence of TNF-α and IL-1β on gene expression in human airway smooth muscle by Hakonarson and colleagues (35). Combined IL-1β and TNF-α exposure resulted in an increase expression of 25 genes encoding cytokines, mostly playing a proinflammatory role such as Il-8, IL-6, RANTES, CSF-3, -2 and others. The autocrine effect was also observed in similar fashion as reported by Cooper et al. (34). A mixture of both cytokines also increased expression of several adhesion molecules – ICAM-1, VCAM-1 and extracellular matrix genes. Several transcription factor genes (e.g. STAT 5, p50-NF-kB, BCL-2A1) were also expressed. Dexamethasone repressed the IL-1β and TNF-α effect on most of the genes studied and stimulated (less than 50% in most cases) expression of a few genes, including MAP kinase subunits, CSF-3 and TNF-α IP3 and 6. Taken together, these data suggest that inflammatory cytokines play an important role in propagating inflammation through various components of the airways and many of these effects are steroid-sensitive. Moreover, these studies suggested that all three cytokines increased expression of several ESTs or proteins of unknown function.

We investigated the influence of IFN-γ and dexamethasone on gene expression in human bronchial epithelial cells (36). IFN-γ induces expression of several chemokines such as MCP-1, RANTES, and IL-8. This was inhibited by dexamethasone treatment. Also, IFN-γ increases expression of several enzymes including cyclooxygenase-2, toll receptor 2, IL-7 receptor and several transcription factors such as Ras, STAT1 and STAT2 (36). For all the aforementioned genes, the IFN-γ effects were inhibited by dexamethasone. Interestingly, dexamethasone alone increased expression of a limited number of genes.

Interestingly, Valenta's group developed an ‘allergen microarray’ allowing one to detect specific serum IgE in patients’ serum. Although this is not a microarray analysis of mRNA per se, it is analysis of gene expression at the functional (protein) level (37).

Applications of microarrays in clinical immunology

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

Chemini and colleagues presented an interesting microarray study on cytokine and chemokine production by peripheral blood mononuclear cells (PBMCs) in hyper-IgE syndrome (38). This study suggested that these PBMCs from patients with hyper-IgE syndrome express more IL-12, M-CSF, c-kit ligand and lower levels of IL-8, eotaxin and MCP-3 than controls. The authors were also unable to show dysregulation of the Th1/Th2 balance in hyper-IgE syndrome (38).

Microarray technology has also allowed for demonstration of an increased expression of several autoantigens in systemic sclerosis patients such as fibrillarin, centromeric protein B, centromeric antiantigen P27 and RNA polymerase II (39). Expression of all these transcripts was elevated in dermal fibroblasts in systemic sclerosis patients. This study underlines again an important role of connective tissue cells, which in many cases may serve as APC in the pathogenesis of immune-related diseases. Luzina et al. studied gene expression in cells from BAL in scleroderma patients (40). Several adhesion molecules, chemokines and chemokine receptors were upregulated in the patient group as compared to controls. These include ICAM-1, ALCAM, CXCR4, MCP-1, GRO-β, GRO-γ and PARC. These proteins may play a role in Th2 cell recruitment. Interestingly, IL-1Ra and TGF-β were also upregulated in systemic sclerosis patients as compared to the control group.

A proteomic approach may also be successfully employed in clinical diagnostics. The use of autoantigen and autoantibody arrays was recently reported by several authors (41, 42). An improvement in purifying proteins from a complex protein mixture will probably increase the application of these methods in clinical pathology.

Future directions in functional genomics

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

The future of functional genomics is complemented by automated applications in protein and/or enzyme activity analysis using two-dimensional protein electrophoresis followed by mass spectrometry. These technologies may for the first time allow connections between genetic polymorphism, subject-related factors, environmental conditions (e.g. allergen exposure, treatment), mRNA gene expression detection and protein expression. The net of new and existing tools to study gene expression is presented in Fig. 2. Functional genomics will play an important role in drug discovery allowing rapid search for new therapeutic targets and exploring unknown pathways that may be affected by treatment.

image

Figure 2. Relationship between various components of functional genomics studies.

Download figure to PowerPoint

Concluding remarks

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References

Microarray technology applied in research has fulfilled several important goals. The results of multiple studies have been confirmed in a few simple experiments. There is a trend towards discovery of new genes. Novel pathways have been identified and may play a role in pathogenesis of diseases. It is widely used as a probe to explore new aspects of known processes and models. Although expensive and demanding in terms of equipment and software, time consuming in experimental terms and in data analysis, it seems to be an important tool in current research. Finely tuned and precise analysis is a requirement, and the need to apply confirmatory experiments is obvious. Application of microarrays in the clinical diagnosis of asthma, allergy and immune diseases may be a next step. A similar approach is actually taking place in cancer diagnostics. Although this approach will not remove well-trained professionals from the clinic, it might actually help to determine proper diagnosis and treatment including therapies directed at novel targets.

References

  1. Top of page
  2. Historical perspective
  3. cDNA microarray
  4. Oligonucleotide microarray (GeneChip)
  5. Data analysis
  6. Principal component analysis
  7. False discovery ratio and the P value
  8. Self-organizing maps (SOMs)
  9. Pseudogene strategy
  10. Hierarchical clustering
  11. Fold change
  12. Data mining tools
  13. Confirmatory experiments and data validation
  14. Application of functional genomics to studies in allergy
  15. Applications of microarrays in clinical immunology
  16. Future directions in functional genomics
  17. Concluding remarks
  18. References
  • 1
    Velculescu VE, Zhang L, Vogelstein B, Kinzler KW. Serial analysis of gene expression. Science 1995;270: 484487.
  • 2
    Iyer VR, Eisen MB, Ross DT, Schuler G, Moore T, Lee JC et al. The transcriptional program in the response of human fibroblasts to serum. Science 1999;283: 8387.
  • 3
    Johannes G, Carter MS, Eisen MB, Brown PO, Sarnow P. Identification of eukaryotic mRNAs that are translated at reduced cap binding complex eIF4F concentrations using a cDNA microarray. Proc Natl Acad Sci USA 1999;96: 1311813123.
  • 4
    Schena M, Shalon D, Davis RW, Brown PO. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 1995;270: 467470.
  • 5
    Schena M, Shalon D, Heller R, Chai A, Brown PO, Davis RW. Parallel human genome analysis: microarray-based expression monitoring of 1000 genes. Proc Natl Acad Sci USA 1996;93: 1061410619.
  • 6
    Rose D. Microfluidic technologies and instrumentation for printing DNA microarrays. Natick: Eaton Publishing, 2000.
  • 7
    Fodor SP, Read JL, Pirrung MC, Stryer L, Lu AT, Solas D. Light-directed, spatially addressable parallel chemical synthesis. Science 1991;251: 767773.
  • 8
    Pease AC, Solas D, Sullivan EJ, Cronin MT, Holmes CP, Fodor SP. Light-generated oligonucleotide arrays for rapid DNA sequence analysis. Proc Natl Acad Sci USA 1994;91: 50225026.
  • 9
    Lipshutz RJ, Morris D, Chee M, Hubbell E, Kozal MJ, Shah N et al. Using oligonucleotide probe arrays to access genetic diversity. Biotechniques 1995;19: 442447.
  • 10
    Lipshutz RJ, Fodor SP, Gingeras TR, Lockhart DJ. High density synthetic oligonucleotide arrays. Nat Genet 1999;21: 2024.
  • 11
    Lipshutz RJ. Applications of high-density oligonucleotide arrays. Novartis Found Symp 2000;229: 8490.
  • 12
    Duggan DJ, Bittner M, Chen Y, Meltzer P, Trent JM. Expression profiling using cDNA microarrays. Nat Genet 1999;21: 1014.
  • 13
    Kaminski N, Friedman N. Practical approaches to analyzing results of microarray experiments. Am J Respir Cell Mol Biol 2002;27: 125132.
  • 14
    Kerr MK, Martin M, Churchill GA. Analysis of variance for gene expression microarray data. J Comput Biol 2000;7: 819837.
  • 15
    Kerr MK, Churchill GA. Bootstrapping cluster analysis: assessing the reliability of conclusions from microarray experiments. Proc Natl Acad Sci USA 2001;98: 89618965.
  • 16
    Kerr MK, Churchill GA. Statistical design and the analysis of gene expression microarray data. Genet Res 2001;77: 123128.
  • 17
    Agarwal S, Rao A. Modulation of chromatin structure regulates cytokine gene expression during T cell differentiation. Immunity 1998;9: 765775.
  • 18
    Hural JA, Kwan M, Henkel G, Hock MB, Brown MA. An intron transcriptional enhancer element regulates IL-4 gene locus accessibility in mast cells. J Immunol 2000;165: 32393249.
  • 19
    Hefner Y, Borsch-Haubold AG, Murakami M, Wilde JI, Pasquet S, Schieltz D et al. Serine 727 phosphorylation and activation of cytosolic phospholipase A2 by MNK1-related protein kinases. J Biol Chem 2000;275: 3754237551.
  • 20
    Zhao X, Bey EA, Wientjes FB, Cathcart MK. Cytosolic phospholipase A2 (cPLA2) regulation of human monocyte NADPH oxidase activity. cPLA2 affects translocation but not phosphorylation of p67(phox) and p47(phox). J Biol Chem 2002;277: 2538525392.
  • 21
    Karp CL, Grupe A, Schadt E, Ewart SL, Keane-Moore M, Cuomo PJ et al. Identification of complement factor 5 as a susceptibility locus for experimental allergic asthma. Nat Immunol 2000;1: 221226.
  • 22
    Ober C, Cox NJ, Abney M, Di Rienzo A, Lander ES, Changyaleket B et al. Genome-wide search for asthma susceptibility loci in a founder population. The Collaborative Study on the Genetics of Asthma. Hum Mol Genet 1998;7: 13931398.
  • 23
    Wjst M, Fischer G, Immervoll T, Jung M, Saar K, Rueschendorf F et al. A genome-wide search for linkage to asthma. German Asthma Genetics Group. Genomics 1999;58: 18.
  • 24
    van der Pouw Kraan TC, Boeije LC, de Groot ER, Stapel SO, Snijders A, Kapsenberg ML et al. Reduced production of IL-12 and IL-12-dependent IFN-gamma release in patients with allergic asthma. J Immunol 1997;158: 55605565.
  • 25
    Hamalainen H, Zhou H, Chou W, Hashizume H, Heller R, Lahesmaa R. Distinct gene expression profiles of human type 1 and type 2 T helper cells. Genome Biol 2001;2:0022.1–0022.11.
  • 26
    Chtanova T, Kemp RA, Sutherland AP, Ronchese F, Mackay CR. Gene microarrays reveal extensive differential gene expression in both CD4(+) and CD8(+) type 1 and type 2 T cells. J Immunol 2001;167: 30573063.
  • 27
    Wirth S, van den Broek M, Frossard CP, Hugin AW, Leblond I, Pircher H, Hauser C. CD8(+) T cells secreting type 2 lymphokines are defective in protection against viral infection. Cell Immunol 2000;202: 1322.
  • 28
    Tam SW, Wiese R, Lee S, Gilmore J, Kumble KD. Simultaneous analysis of eight human Th1/Th2 cytokines using microarrays. J Immunol Methods 2002;261: 157165.
  • 29
    Brutsche MH, Brutsche IC, Wood P, Mogulkoc N, Custovic A, Egan J, Woodcock A. B-cell isotype control in atopy and asthma assessed with cDNA array technology. Am J Physiol Lung Cell Mol Physiol 2001;280: L627L637.
  • 30
    Brutsche MH, Joos L, Carlen Brutsche IE, Bissinger R, Tamm M, Custovic A, Woodcock A. Array-based diagnostic gene-expression score for atopy and asthma. J Allergy Clin Immunol 2002;109: 271273.
  • 31
    Cohn L, Tepper JS, Bottomly K. IL-4-independent induction of airway hyperresponsiveness by Th2, but not Th1, cells. J Immunol 1998;161: 38133816.
  • 32
    Grunig G, Warnock M, Wakil AE, Venkayya R, Brombacher F, Rennick DM et al. Requirement for IL-13 independently of IL-4 in experimental asthma. Science 1998;282: 22612263.
  • 33
    Lee JH, Kaminski N, Dolganov G, Grunig G, Koth L, Solomon C et al. Interleukin-13 induces dramatically different transcriptional programs in three human airway cell types. Am J Respir Cell Mol Biol 2001;25: 474485.
  • 34
    Cooper P, Potter S, Mueck B, Yousefi S, Jarai G. Identification of genes induced by inflammatory cytokines in airway epithelium. Am J Physiol Lung Cell Mol Physiol 2001;280: L841L852.
  • 35
    Hakonarson H, Halapi E, Whelan R, Gulcher J, Stefansson K, Grunstein MM. Association between IL-1beta/TNF-alpha-induced glucocorticoid-sensitive changes in multiple gene expression and altered responsiveness in airway smooth muscle. Am J Respir Cell Mol Biol 2001;25: 761771.
  • 36
    Pawliczak R, Logun C, Munson P, Danner R, Shelhamer J. The influence of glucocorticosteroids and interferon gamma on gene expression in human lung epithelial cells. Am J Respir Crit Care Med 2002;165: A626.
  • 37
    Hiller R, Laffer S, Harwanegg C, Huber M, Schmidt WM, Twardosz A et al. Microarrayed allergen molecules: diagnostic gatekeepers for allergy treatment. FASEB J 2002;16: 414416.
  • 38
    Chehimi J, Elder M, Greene J, Noroski L, Stiehm ER, Winkelstein JA, Sullivan KE. Cytokine and chemokine dysregulation in hyper-IgE syndrome. Clin Immunol 2001;100: 4956.
  • 39
    Zhou X, Tan FK, Xiong M, Milewicz DM, Feghali CA, Fritzler MJ et al. Systemic sclerosis (scleroderma): specific autoantigen genes are selectively overexpressed in scleroderma fibroblasts. J Immunol 2001;167: 71267133.
  • 40
    Luzina IG, Atamas SP, Wise R, Wigley FM, Xiao HQ, White B. Gene expression in bronchoalveolar lavage cells from scleroderma patients. Am J Respir Cell Mol Biol 2002;26: 549557.
  • 41
    Haab BB, Dunham MJ, Brown PO. Protein microarrays for highly parallel detection and quantitation of specific proteins and antibodies in complex solutions. Genome Biol 2001;2: 4.14.13.
  • 42
    MacBeath G, Schreiber SL. Printing proteins as microarrays for high-throughput function determination. Science 2000;289: 17601763.