SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Genetic polymorphism of pathogens
  5. Host–pathogen interaction
  6. Gene expression profile of pathogens
  7. Identification of new targets for drug design and determination of their mechanism of action
  8. Conclusion
  9. Acknowledgements
  10. References

Microarrays are a promising technique for elucidating and interpreting the mechanistic roles of genes in the pathogenesis of infectious disease. Microarrays have been used to analyse the genetic polymorphisms of specific loci associated with resistance to antimicrobial agents, to explore the distribution of genes among isolates from the same and similar species, to understand the evolutionary relationship between closely related species and to integrate the clinical and genomic data. This technique has also been used to study host–pathogen interactions, mainly by identifying genes from pathogens that may be involved in pathogenicity and by surveying the scope of the host response to infection. The RNA expression profile of pathogens has been used to identify regulatory mechanisms that ensure gene expression in the appropriate environment, to hypothesize functions of hundreds of uncharacterized genes and to identify virulence genes that promote colonization or tissue damage. This information also has the potential to identify targets for drug design. Furthermore, microarrays have been used to investigate the mechanism of drug action and to delineate and predict adverse effects of new drugs. In this paper, we review the use of spotted and high-density oligonucleotide arrays to study the genetic polymorphisms of pathogens, host–pathogen interactions and whole-genome expression profiles of pathogens, as well as their use for drug discovery.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Genetic polymorphism of pathogens
  5. Host–pathogen interaction
  6. Gene expression profile of pathogens
  7. Identification of new targets for drug design and determination of their mechanism of action
  8. Conclusion
  9. Acknowledgements
  10. References

A main characteristic of pathogenic microorganisms is their ability to cause tissue damage and consequently disease in their host. For decades, scientists have used a wide variety of methods to study how microorganisms interact with the host to cause damage and the mechanisms by which the host protects itself from microorganisms. This information has been invaluable in the design of safe and effective diagnostics, therapeutics and vaccines. We are currently in the midst of an explosive increase in the availability of genomic sequence information, with 41 small genomes having been fully sequenced and published, and more than 120 more genomes in the process of being sequenced (http://www.tigr.org/tdb/). The availability of the complete sequence of both host and pathogen is promising in developing novel insights into the host–pathogen relationship. However, extracting biological knowledge from sequence data is the essential challenge of this post-genome era. DNA microarray-based approaches have gained rapid acceptance in a variety of fields for studying the roles of genes in the pathogenesis of infectious disease.

The theory and background of microarray technology, as well as the technology itself, have been described in detail elsewhere (Ferea and Brown, 1999; Lipshutz et al., 1999). In brief, a DNA microarray is a microscopic chequerboard representing thousands of different DNA sequences. There are several methods of producing microarrays, but the following review is restricted to the two most commonly used techniques: spotted glass slide microarray and high-density oligonucleotide array technology. In the spotted microarray, presynthesized single-stranded or double-stranded DNA is bound or ‘printed’ onto glass slides. The DNA can be generated from cloned, synthesized or polymerase chain reaction (PCR)-amplified material. Because of the technical simplicity of this approach, spotted microarrays can be produced in house (http://www.stanford.edu/pbrown/array.html) as well as purchased from commercial providers (Fig. 1). High-density oligonucleotide arrays are constructed by synthesizing short (≈ 25-mer) oligonucleotides in situ on glass wafers using a photolithographic manufacturing process and are thus available only from commercial vendors (Fig. 2) (Lipshutz et al., 1999). Both types of DNA microarrays are used to measure the relative abundance of DNA or RNA in order to compare genomes or gene expression profiles.

image

Figure 1. Schematic representation of the construction and use of the DNA spotted microarray. Single- or double-stranded DNA generated from direct synthesis of oligonucleotides or PCR amplification are bound or ‘printed’ onto glass slides. A mixture of test and control samples, each labelled with a different fluorescent dye, is hybridized to the microarray. The microarray is then scanned to measure the fluorescence intensity of every spot for each dye. The ratio of the fluorescent intensities reflects the relative abundance of the two samples.

Download figure to PowerPoint

image

Figure 2. Schematic representation of the construction and use of high-density oligonucleotide chip. Two types of oligonucleotides or probes are synthesized in situ on glass wafers using a photolithographic or light-directed synthesis. The perfect match (PM) probe is designed according to the sequence to be interrogated, and the mismatch (MM) probe is the same as PM but with a different basepair in the middle, which serves as a control for non-specific hybridization. In the experimental protocol of GeneChip, just one sample of labelled genomic DNA is hybridized. The GeneChip is stained and scanned with a confocal scanner. If a nucleic acid sequence is present, the PM probe will have a higher signal than the MM probe. If the nucleic acid is not present, the PM value will be similar to the MM value. When both PM and MM have a high signal, it may suggest non-specific hybridization.

Download figure to PowerPoint

Microarray-based approaches have several advantages over other systems that have been used previously to study pathogens and their interaction with hosts, such as those that measure the expression of a small number of genes in individual experiments. The major advantage of microarrays is their ability to measure simultaneously the presence of tens of thousands of different nucleic acid sequences. Thus, this technique permits the quantification of specific genes and their expression patterns in a comprehensive genome-wide framework. Although expensive relative to other quantitative hybridization and amplification methods, the high-throughput capacity makes it a cost-effective technique for a variety of applications. These analyses can be made within the same species and between different, but similar, species. In addition, arrays can be designed specifically to deduce simultaneously the sequence of multiple different genes.

There are important differences between the spotted microarrays and high-density oligonucleotide arrays that make each particularly well suited to a variety of applications. Because spotted microarrays can be designed and constructed in academic facilities, their content can easily be changed to accommodate evolving research needs. However, a disadvantage of spotted microarrays is the variation in the size of the spot and the amount of DNA contained in every spot. Also, variations in length and composition of sequence and variations in post-processing and hybridization result in less consistency and reproducibility within and across experiments than the high-density array. In high-density oligonucleotide arrays, several different oligonucleotides interrogate each gene, in part accounting for the improvement in the overall consistency and reproducibility of results. On the other hand, these chips are relatively inflexible because their construction requires the design and manufacture of specific sets of masks that will determine the nucleotides to be synthesized in situ, a process that must be repeated entirely if any changes are made.

It is not possible to cover all the work conducted to date using DNA microarrays. Thus, we will highlight selected studies that exemplify how this technology has been used to study the genetic polymorphism of pathogens, host–pathogen interactions, the gene expression profiles of pathogens and drug target identification.

Genetic polymorphism of pathogens

  1. Top of page
  2. Abstract
  3. Introduction
  4. Genetic polymorphism of pathogens
  5. Host–pathogen interaction
  6. Gene expression profile of pathogens
  7. Identification of new targets for drug design and determination of their mechanism of action
  8. Conclusion
  9. Acknowledgements
  10. References

The analysis of genetic polymorphisms among isolates from the same species is a way of distinguishing between different strains, studying the transmission dynamics of the microorganisms and exploring the relationship between genetic and phenotypic polymorphisms. In addition, genetic polymorphisms among isolates from the same or different species can be used to understand the population genetics and evolutionary relationship between closely related species. Both spotted microarray and high-density oligonucleotide array technologies have been used to study small-scale genomic deletions among clinical isolates (Salama et al., 2000; Kato-Maeda et al., 2001).

One of the earliest applications of microarrays for pathogen analysis was to determine which of a number of known alleles were present at a specific genomic locus, such as those conferring resistance to antimicrobial agents. For example, mutations in coding regions of the reverse transcriptase and protease genes of human immunodeficiency virus were detected using high-density oligonucleotide arrays in order to predict resistance to antiviral drugs (Merigan, 1995; Wilson et al., 2000) These arrays were designed in a ‘tiling’ fashion to interrogate the loci of interest for mutations. Four probes were designed to interrogate each nucleotide position of the target sequence. One probe was designed to be perfectly complementary, and the other three were designed with each potential nucleotide present at the interrogated position (Hacia, 1999). The data are interpreted assuming that hybridization will be highest to the perfectly homologous probe. This approach has been used to detect mutations in the rpoB and katG genes and the 16S rRNA of Mycobacterium tuberculosis (Gingeras et al., 1998). Interestingly, in addition to detecting mutations associated with drug resistance, the pattern of hybridization to these arrays discriminated between different mycobacterial species (Gingeras et al., 1998; Troesch et al., 1999; Salama et al.. 2000). These results suggested that array data specifically collected for one purpose (drug susceptibility testing) can be interpreted to reveal additional insights (speciation).

Microarray analysis is now a well-established technique for exploring the distribution of genes among collections of isolates from the same species. Using high-density oligonucleotide arrays, 58 genes that are present in a sequenced strain of Saccharomyces cerevisiae were found to be absent from an isolate from a different environment (Winzeler et al., 1999). In similar fashion, spotted microarrays were used to compare the genomes of 15 different Helicobacter pylori isolates, which showed that 22% of the genes were dispensable in one or more strains (Salama et al., 2000). These genes included restriction modification genes, transposases and genes encoding cell surface proteins. Although clinical data were not available for these strains, the authors postulated that some of these genes might be associated with virulence.

Microarrays can also be used to compare the genomic content of similar species. For example, spotted microarrays were used to compare M. tuberculosis, Mycobacterium bovis and the family of Bacille Calmette-Guérin (BCG), the antituberculosis vaccine produced by attenuating M. bovis (Behr et al., 1999). Eleven regions, including 91 genes, were absent from one or more virulent strains of M. bovis. Five additional regions encompassing 38 genes were present in M. tuberculosis and M. bovis but absent in some or all BCG strains. Based on these findings, the authors suggested that the different deletions detected in the BCG vaccines reflected a progressive adaptation of BCG to laboratory conditions that may have impaired their ability to stimulate a durable immune response in the host (Behr et al., 1999). It is postulated that the differences between M. tuberculosis and BCG could be exploited to develop diagnostics to distinguish between the immunity induced by infection and vaccination. Microarray-based comparative genomics may also be used to understand the evolutionary relationship between different species, such as the pathogen Streptococcus pneumoniae and the commensals Streptococccus mitis and Streptococccus oralis (Hakenbeck et al., 2001).

Integrating clinical data and genomic deletion data is another promising application of microarrays. In a recently published study, M tuberculosis cDNA high-density oligonucleotide arrays were used to identify segments of DNA that are present in the sequenced strain (M. tuberculosis, H37Rv) but missing from clinical isolates (Kato-Maeda et al., 2001). On average, 0.3% of the genome was deleted in each isolate, and a total of 1.7% of the genome was polymorphic among 19 clinical isolates studied from 16 different clones. The distribution of deletions varied between different clones, yet was conserved among clonally related organisms, suggesting that genomic deletion patterns may be a useful marker for molecular epidemiological studies. The most provocative finding in this study was that strains with a large amount of deleted DNA were less likely to cause cavitary pulmonary disease than strains with fewer deletions. Given that cavitary pulmonary disease is a particularly transmissible form of tuberculosis, the authors postulated that the accumulation of deletions may cause a decrease in bacterial fitness (Kline et al., 1995; Arber, 2000). This observation was concordant with Muller's (1964) prediction that the accumulations of mutations in clonal organisms will result in genetic deterioration. Alternatively, it can be postulated that the loss of genes can be a consequence of the adaptation to a new environment, as has been demonstrated in Rickettsia prowazekii and Chlamydia trachomatis (Andersson and Andersson, 1999).

Host–pathogen interaction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Genetic polymorphism of pathogens
  5. Host–pathogen interaction
  6. Gene expression profile of pathogens
  7. Identification of new targets for drug design and determination of their mechanism of action
  8. Conclusion
  9. Acknowledgements
  10. References

The establishment of an infection or disease in humans by a microorganism is an exceptional event. By understanding why certain strains of a particular organism are pathogenic or under which conditions they cause disease, it may be possible to develop new tools for disease prevention or management. The complex interaction between host and pathogen is now being explored using microarrays (Cummings and Relman, 2000; Manger and Relman, 2000). The basic model consists of an ex vivo measurement of gene expression of host cells before and after they are infected with a microorganism. By following the pattern of gene expression at different times, it is possible to elucidate which host genes are up- or downregulated over the course of infection. Identification of genes that are differentially regulated and the characterization of their functions provide a promising window on the understanding of pathogenicity.

The response of human fibroblasts to cytomegalovirus (CMV) infection has been explored using DNA microarrays (Zhu et al., 1998). A total of 258 out of 6600 genes from the human fibroblast changed by a factor of at least 4 after the onset of viral replication: 124 increased and 134 decreased. The specific genes involved can be examined to refine hypotheses about their role in the host–pathogen interaction. For example, it is hypothesized that CMV HLA-E mRNA genes are upregulated to avoid natural killer cell surveillance, whereas genes from the host prostaglandin E2 pathway are increased as part of the proinflammatory response.

The specific pattern of gene response also provides evidence of the scope of the host response to infection. Relatively few genes were induced when the gene expression profile of human colorectal and colonic epithelial cells was studied after infection with Salmonella dublin (Eckmann et al., 2000). These included cytokines, kinases, transcription factors and HLA class 1 genes. The authors interpret this as evidence that the epithelial mRNA response to infection is limited but specific. Similarly, few human bronchial epithelial genes were induced after being infected with Bordetella pertussis (Belcher et al., 2000). These included genes encoding proinflammatory cytokines with chemoattractant activities, which may explain the cellular infiltrate seen in patients with pertussis. In addition, the transcriptional analysis suggested that the physical properties of the respiratory tract mucus of the host might be altered by the activity of the pertussis toxin, creating a more advantageous microenvironment for the pathogen.

Directly comparing gene expression profiles of host cell infected by wild-type and mutant pathogen strains has been used to explore the host–pathogen interaction. Detweiler et al. (2001) used this approach to identify a role for the Salmonella response regulator phoP in human macrophage cell death. They directly compared gene expression profiles of the wild-type strain and phoP::Tn10 mutant-infected macrophages. Of 22 571 genes, only 34 genes were downregulated in phoP::Tn10-infected macrophages compared with wild type-infected macrophages. Six of those genes were known to play roles in cell death, the cell cycle or both. Therefore, the authors assumed that phoP might affect macrophage survival during Salmonella infection, which was confirmed by conventional cell biological methods (Detweiler et al., 2001)

Gene expression profile of pathogens

  1. Top of page
  2. Abstract
  3. Introduction
  4. Genetic polymorphism of pathogens
  5. Host–pathogen interaction
  6. Gene expression profile of pathogens
  7. Identification of new targets for drug design and determination of their mechanism of action
  8. Conclusion
  9. Acknowledgements
  10. References

A central challenge to all pathogens is adapting to changing environmental conditions. This adaptation involves regulating gene expression in response to different environmental signals. In general, it has been assumed that the expression of genes that are needed in a particular situation will be upregulated, whereas unnecessary functions in the same situation will be downregulated. The same principle can be applied to virulence genes, which are also subject to regulatory mechanisms that ensure expression in the appropriate host environment (Guiney, 1997). The power of microarrays is their ability to provide a whole-genome perspective on these responses. For example, the response of all 6200 genes of S. cerevisiae to growth in rich and minimal media has been explored using microarrays (Wodicka et al., 1997). The mRNA of 36 genes was more abundant in rich medium, whereas 140 genes were more abundant in minimal medium. Although some of the highly expressed genes encoded well-known enzymes, structural proteins and ribosomal proteins, the functions of many of the most highly expressed genes were unknown. The expression levels of all 4290 genes of Escherichia coli were studied in a similar fashion (Tao et al., 1999). One hundred and nineteen genes were strongly expressed on rich medium, whereas 225 genes were expressed at significantly higher levels on minimal medium. These genes were involved in a number of biosynthetic pathways and stress tolerance proteins.

Elucidating the repertoire of genes co-expressed at specific times or under certain environmental and physiologic conditions is a novel approach to identifying their functions. This approach is based on the assumption that genes with similar expression patterns are likely to have the same functions (‘guilt-by-association’). This method has been used to hypothesize the functions of hundreds of uncharacterized genes in yeast, Drosophila melanogaster, mice and humans. For example, in the early phase of sporulation in budding yeast, about 62 genes followed a definable pattern of expression (i.e. detectable within 0.5 h after transfer to sporulation medium and sustaining expression thereafter) (Chu et al., 1998). Because many of the genes had known functions in sporulation, it was inferred that the remainder of genes in this class had similar functions.

Microarrays have also been used to identify virulence genes that promote colonization or host damage (Eckmann et al., 2000). The approach is based on the assumption that virulence-associated genes are likely to be co-regulated. This entails measuring whole-genome expression profiles under a large number of conditions and then using ‘clustering’ algorithms to identify the subset of genes that are up- or downregulated together. Using this method, Ogawa et al. (2000) identified 22 PHO-regulated genes (genes involved in the scavenging and specific uptake of phosphate from extracellular sources) in S. cerevisiae. Eight of these genes had no previously known function and were characterized further by gene disruption. These genes were expressed in similar fashion to those known PHO genes.

Because genes with similar expression patterns may be regulated by the same regulatory genes, microarray technology can also be used to identify the cis- and trans-regulatory elements that control the expression network. de Saizieu et al. (2000) identified a group of regulatory genes controlled by an autoinduced peptide in S. pneumoniae using microarrays. They found that 16 genes were induced by Blpc peptide and belonged to eight different operons. An extended −10 promoter region was identified, and two direct 11 bp repeats spaced by exactly 10 bp were present in each promoter sequence except one. This suggested that there might be a constitutive binding site for specific regulatory protein. The authors also identified the trans-regulatory element (BlpC peptide) and its binding site (cis-regulatory element) in S. pneumoniae.

Identification of new targets for drug design and determination of their mechanism of action

  1. Top of page
  2. Abstract
  3. Introduction
  4. Genetic polymorphism of pathogens
  5. Host–pathogen interaction
  6. Gene expression profile of pathogens
  7. Identification of new targets for drug design and determination of their mechanism of action
  8. Conclusion
  9. Acknowledgements
  10. References

The aim of drug development is to design compounds targeting ‘disease-causing’ gene products to achieve disease modification or compounds targeting ‘disease-result’ gene products to alleviate symptoms (Debouck and Goodfellow, 1999). Many of the drugs sold today were designed in biochemistry laboratories. The drug discovery process started with the knowledge of biochemical pathways implicated in a pathophysiological process or by empirical observations. An enzyme, preferably catalysing a rate-limiting step in this pathway, was characterized and then tested with various potential inhibitory molecules. Once a good candidate molecule was found, medicinal chemists optimized the compound. Currently, the advent of molecular biology techniques and the availability of sequences from humans and several organisms have changed the field of drug discovery and development. Microarrays are also a useful tool for investigating the mechanisms of drug action. The basic approach is to compare the expression of normal and diseased cells, with or without drug exposure (Gray et al., 1998). The expression profile of a cell exposed to a specific drug has been called drug-specific gene expression signature. The gene expression signature for the immunosuppressive drug FK506 was studied in S. cerevisiae (Marton et al., 1998). The authors demonstrated that the drug altered the expression of 36 open reading frames (ORFs) from S. cerevisiae by more than twofold, similar to the expression observed in S. cerevisiae with mutations in the genes known to be inhibited by the drug. However, the same expression profile can be seen by the pharmacological inhibition of different target genes. Therefore, the authors proposed a ‘decoder’ strategy, in which the expression of a panel of genetic mutant strains was compared to evaluate the effect of a compound to inhibit pathways other that of its intended target. In addition, treatment of the null mutants with FK506 also revealed additional pathways distinct from the primary target. These newly discovered pathways are likely to be promising targets for new drugs.

In a similar fashion, spotted microarrays were used to determine the gene expression signature of M. tuberculosis exposed to isoniazid (INH) (Wilson et al., 1999). This drug is used in treatment against tuberculosis and blocks the mycolic acid biosynthetic pathway. The results confirmed biochemically derived observations that INH induces a set of genes that encodes polypeptide components of the FAS-II complex. Other INH-induced genes encoded proteins that have not been characterized biochemically but are homologous with proteins of known function. These include genes participating in the degradation of fatty acids, which may be part of the operon of the FAS-II complex. Another induced gene was an efflux protein, EfpA. The induction of this gene by INH suggests that the gene product transports molecules relevant to mycolic acid production. This enzyme is a particularly attractive drug target because the gene is present only within the pathogenic members of the mycobacterial genus (Wilson et al., 1999).

The analysis of the gene expression signatures includes data that may be useful to delineate and predict adverse events. If the expression of a drug candidate target closely matches the pattern generated by a known toxic agent, it can be assumed that the drug candidate would probably be toxic, a fact that can be confirmed by in vitro analysis (Braxton and Bedilion, 1998).

The whole-genome perspective of microarrays makes them promising tools for sifting through the genome to identify candidate genes as targets for drug development. However, expression data have limitations because mRNA levels may not reflect protein levels, and expression of a protein may not always have a pathological consequence (Gygi et al., 1999). Therefore, traditional pathology and toxicity studies will remain necessary.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Genetic polymorphism of pathogens
  5. Host–pathogen interaction
  6. Gene expression profile of pathogens
  7. Identification of new targets for drug design and determination of their mechanism of action
  8. Conclusion
  9. Acknowledgements
  10. References

Microarray technology is based on the well-established and long-exploited principle of nucleic acid hybridization. However, for the first time, it offers the possibility of simultaneously conducting tens or hundreds of thousands of simultaneous hybridizations. This increased experimental efficiency permits high throughput and whole-genome expression profiling of pathogens and hosts.

There has been an explosive increase in the amount of genomic sequence available over the past few years. However, in the absence of further analysis, these sequence data are of little value. Bioinformatic approaches hold great promise for predicting the role of specific genes in the interaction between host and pathogen. Functional data, such as that provided by microarray technology, are an exciting approach to generating additional insights from sequence information.

To date, much of these data have been generated in model systems, in which conditions can be controlled precisely and relatively large quantities of mRNA can be harvested. Because the results of pathogen gene expression and of host–pathogen interactions are influenced by the model system used, such results must be interpreted with caution. In the future, as the technology improves, it is anticipated that microarray analysis will be performed with much smaller quantities of mRNA. For example, improvements are being made in probe design, in the process by which these probes are bound to slides, in the labelling of nucleic acid and in slide scanning, as well as in data analysis. In the meantime, it is imperative to develop a standardized and more efficient system for easy microarray data processing, economic data storage and fast retrieval and facilitating data communication between different researchers.

Microarrays will only augment and not replace other established methods for investigating host–pathogen interactions. Most of the studies using this technology at this time have served only to refine hypotheses about genetic polymorphism or gene functions. These refined hypotheses subsequently need to be confirmed by other techniques. Lastly, because of the inconsistent correlation between mRNA transcript levels with protein expression and activity in prokaryotes, proteomic analysis is likely to provide a wealth of additional insights.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Genetic polymorphism of pathogens
  5. Host–pathogen interaction
  6. Gene expression profile of pathogens
  7. Identification of new targets for drug design and determination of their mechanism of action
  8. Conclusion
  9. Acknowledgements
  10. References
  • Andersson, J.O. & Andersson, S.G. (1999) Insights into the evolutionary process of genome degradation. Curr Opin Genet Dev 9: 664671.DOI: 10.1016/s0959-437x(99)00024-6
  • Arber, W. (2000) Genetic variation: molecular mechanisms and impact on microbial evolution. FEMS Microbiol Rev 24: 17.DOI: 10.1016/s0168-6445(99)00032-7
  • Behr, M.A., Wilson, M.A., Gill, W.P., Salamon, H., Schoolnik, G.K., Rane, S., Small, P.M. (1999) Comparative genomics of BCG vaccines by whole-genome DNA microarray (see comments). Science 284: 15201523.
  • Belcher, C.E., Drenkow, J., Kehoe, B., Gingeras, T.R., McNamara, N., Lemjabbar, H., et al. (2000) From the cover: the transcriptional responses of respiratory epithelial cells to Bordetella pertussis reveal host defensive and pathogen counter-defensive strategies. Proc Natl Acad Sci USA 97: 1384713852.DOI: 10.1073/pnas.230262797
  • Braxton, S. & Bedilion, T. (1998) The integration of microarray information in the drug development process. Curr Opin Biotechnol 9: 643649.
  • Chu, S., DeRisi, J., Eisen, M., Mulholland, J., Botstein, D., Brown, P.O., Herskowitz, I. (1998) The transcriptional program of sporulation in budding yeast. Science 282: 699705.
  • Cummings, C.A. & Relman, D.A. (2000) Using DNA microarrays to study host–microbe interactions. Emerg Infect Dis 6: 513525.
  • Debouck, C. & Goodfellow, P.N. (1999) DNA microarrays in drug discovery and development. Nature Genet 21: 4850.
  • Detweiler, C.S., Cunanan, D.B., Falkow, P. (2001) Host microarray analysis reveals a role for the Salmonella response regulator phoS in human macrophage cell death. Proc Natl Acad Sci USA 98: 58505855.DOI: 10.1073/pnas.091110098
  • Eckmann, L., Smith, J.R., Housley, M.P., Dwinell, M.B., Kagnoff, M.F. (2000) Analysis by high density cDNA arrays of altered gene expression in human intestinal epithelial cells in response to infection with the invasive enteric bacteria Salmonella. J Biol Chem 275: 1408414094.DOI: 10.1074/jbc.275.19.14084
  • Ferea, T.L. & Brown, P.O. (1999) Observing the living genome. Curr Opin Genet Dev 9: 715722.
  • Gingeras, T.R., Ghandour, G., Wang, E., Berno, A., Small, P.M., Drobniewski, F., et al. (1998) Simultaneous genotyping and species identification using hybridization pattern recognition analysis of generic Mycobacterium DNA arrays. Genome Res 8: 435448.
  • Gray, N.S., Wodicka, L., Thunnissen, A.M., Norman, T.C., Kwon, S., Espinoza, F.H., et al. (1998) Exploiting chemical libraries, structure, and genomics in the search for kinase inhibitors. Science 281: 533538.
  • Guiney, D.G. (1997) Regulation of bacterial virulence gene expression by the host environment. J Clin Invest 99: 565569.
  • Gygi, S.P., Rochon, Y., Franza, B.R., Aebersold, R. (1999) Correlation between protein and mRNA abundance in yeast. Mol Cell Biol 19: 17201730.
  • Hacia, J.G. (1999) Resequencing and mutational analysis using oligonucleotide microarrays. Nature Genet 21: 4247.
  • Hakenbeck, R., Balmelle, N., Weber, B., Gardes, C., Keck, W., De Saizieu A. (2001) Mosaic genes and mosaic chromosomes: intra- and interspecies genomic variation of Streptococcus pneumoniae. Infect Immun 69: 24772486.
  • Kato-Maeda, M., Rhee, J.T., Gingeras, T.R., Salamon, H., Drenkow, J., Smittipat, N., Small P.M. (2001) Comparing genomes within the species Mycobacterium tuberculosis. Genome Res 11: 547554.
  • Kline, S.E., Hedemark, L.L., Davies, S.F. (1995) Outbreak of tuberculosis among regular patrons of a neighborhood bar (see comments). N Engl J Med 333: 222227.
  • Lipshutz, R.J., Fodor, S.P., Gingeras, T.R., Lockhart, D.J. (1999) High density synthetic oligonucleotide arrays. Nature Genet 21: 2024.
  • Manger, I.D. & Relman, D.A. (2000) How the host ‘sees’ pathogens: global gene expression responses to infection. Curr Opin Immunol 12: 215218.DOI: 10.1016/s0952-7915(99)00077-1
  • Marton, M.J., DeRisi, J.L., Bennett, H.A., Iyer, V.R., Meyer, M.R., Roberts, C.J., et al. (1998) Drug target validation and identification of secondary drug target effects using DNA microarrays. Nature Med 4: 12931301.
  • Merigan, T. (1995) Individualization of therapy using viral markers. J Acquir Immune Defic Syndr Hum Retrovirol 10 (Suppl. 1): S41S46.
  • Muller, H.J. (1964) The relation of recombination to mutational advance. Mutat Res 1: 29.
  • Ogawa, N., DeRisi, J., Brown, P.O. (2000) New components of a system for phosphate accumulation and polyphosphate metabolism in Saccharomyces cerevisiae revealed by genomic expression analysis. Mol Biol Cell 11: 43094321.
  • De Saizieu, A., Gardes, C., Flint, N., Wagner, C., Kamber, M., Mitchell, T.J., et al. (2000) Microarray-based identification of a novel Streptococcus pneumoniae regulon controlled by an autoinduced peptide. J Bacteriol 182: 46964703.
  • Salama, N., Guillemin, K., McDaniel, T.K., Sherlock, G., Tompkins, L., Falkow, S. (2000) A whole-genome microarray reveals genetic diversity among helicobacter pylori strains. Proc Natl Acad Sci USA 97: 1466814673.
  • Tao, H., Bausch, C., Richmond, C., Blattner, F.R., Conway, T. (1999) Functional genomics: expression analysis of Escherichia coli growing on minimal and rich media. J Bacteriol 181: 64256440.
  • Troesch, A., Nguyen, H., Miyada, C.G., Desvarenne, S., Gingeras, T.R., Kaplan, P.M., et al. (1999) Mycobacterium species identification and rifampin resistance testing with high-density DNA probe arrays. J Clin Microbiol 37: 4955.
  • Wilson, J.W., Bean, P., Robins, T., Graziano, F., Persing, D.H. (2000) Comparative evaluation of three human immunodeficiency virus genotyping systems: the HIV-GenotypR method, the HIV PRT GeneChip assay, and the HIV-1 RT line probe assay. J Clin Microbiol 38: 30223028.
  • Wilson, M., DeRisi, J., Kristensen, H.H., Imboden, P., Rane, S., Brown, P.O., Schoolnik, G.K. (1999) Exploring drug-induced alterations in gene expression in Mycobacterium tuberculosis by microarray hybridization. Proc Natl Acad Sci USA 96: 1283312838.DOI: 10.1073/pnas.96.22.12833
  • Winzeler, E.A., Lee, B., McCusker, J.H., Davis, R.W. (1999) Whole genome genetic-typing in yeast using high-density oligonucleotide arrays. Parasitology 118: S73S80.
  • Wodicka, L., Dong, H., Mittmann, M., Ho, M.H., Lockhart, D.J. (1997) Genome-wide expression monitoring in Saccharomyces cerevisiae. Nature Biotechnol 15: 13591367.
  • Zhu, H., Cong, J.P., Mamtora, G., Gingeras, T., Shenk, T. (1998) Cellular gene expression altered by human cytomegalovirus: global monitoring with oligonucleotide arrays. Proc Natl Acad Sci USA 95: 1447014475.DOI: 10.1073/pnas.95.24.14470
Footnotes
  1. Present address: Department of Infectious Diseases, Instituto Nacional de Ciencias Médicas y Nutrición ‘Salvador Zubirán’, México City, México 14000.