Commentary Pushing the limits of cellular microbiology: Microarrays to study bacteria–host cell intimate contacts Rino Rappuoli* IRIS, Chiron S.p.A., Via Fiorentina 1, 53100 Siena, Italy or more than a century, microbiologists then prepared from each of the RNAs by nome of a pathogen usually become Fstudied pathogens in pure culture, reverse transcription, usually by using available as soon as the genomic se- whereas cell biologists studied mammalian oligo(dT) to probe host-cell genes and quence of the bacterium is published; cells in tissue culture. Both fields used rich random oligonucleotide primers to an- sometimes they are available even before media and conditions optimized for in vitro neal to bacterial genes. A balanced mix- publication. Usually chips are printed by growth. A few years ago, researchers real- ture of oligos specific for each gene using each of the bacterial genes, which ized that these laboratory methods of grow- present in the microarray may also be have been previously amplified by PCR. ing pathogens and their hosts were quite used. (This is particularly valuable for Microarrays containing all genes of the artificial and had very little in common with preparing probes of bacterial RNA.) To sequenced bacterial genomes are either real life, where pathogens and hosts coexist, probe printed microarrays, RNA is con- available or are being prepared. These interact, and compete in conditions that are verted into a fluorescent cDNA probe by include Helicobacter pylori, Neisseria often far from optimal. To better mimic incorporation of different fluoro- meningitidis, Mycobacterium tuberculosis, what happens in real life, the study of the chromes during the reverse-transcription and Escherichia coli. Chips containing interaction between microbes and host cells reaction (typically by using Cy3- and the genes of many other bacteria will be was proposed, taking advantage of techno- Cy5-dCTP). The probes are then used to available within the next few months. logical progress that allowed cocultures of hybridize microarrays containing bacte- In the experiment shown in Fig. 1e, the microbes and cells to be handled together. A rial genes (Fig. 1d) or eukaryotic genes RNA prepared in b and labeled with Cy3 new discipline (cellular microbiology) was (e). High-density genome-wide microar- (green) is mixed with the RNA prepared born (1). This discipline took over quite rays of eukaryotic cells are commercially in c and labeled with Cy5 (red). The two COMMENTARY rapidly, a new journal named after this new available. The Hu6800 Array (Af- fluorescent cDNAs are then mixed to- discipline was started (Cellular Microbiology, fymetrix , Santa Clara, CA), which in a gether and constitute the probe, which is Blackwell, Oxford), and textbooks were chip of 1.28 ϫ 1.28 cm contains probes hybridized to the chip. Sometimes RNA published (2, 3) describing in detail the for approximately 6,800 human genes, is prepared from a pool of control condi- techniques to be used in cellular microbiol- the one used in most published studies; tions͞cells is used to prepare a Cy3- ogy and the scientific problems that could be however, today a set of 5 chips (Human labeled reproducible control, which is addressed. These textbooks are just a few Genome U95 set), each containing ap- then mixed with Cy5-labeled cDNA pre- months old, when a powerful new approach proximately 12,000 genes͞ESTs per chip pared from the experimental condition, to cellular microbiology is described (4). for a total of more than 60,000 human including uninfected cells. This allows an This novel approach pushes the limits that genes and ESTs, is available. Many other invariant comparator, so that any experi- we had previously and suggests it is probably systems are also available [UniGem V mental profile can be compared with any already time for a new edition of these (Incyte Pharmaceuticals, Palo Alto, other, even months apart, by normalizing textbooks. The new technique, reported in CA), MassArray Spectrochip (Seque- all data to the standard Cy3 control. The part by Belcher et al. in this issue of PNAS nom, Hamburg, Germany and San ability to compare different studies from (4) and by other recent studies, describes Diego, California), Universal Arrays the same laboratory and from different how, instead of studying one parameter at a (Genometrix, The Woodlands, TX), Hy- laboratories is one of the major problems time, we can use microchips to study, within chip (Hyseq, Sunnyvale, CA), and the we face today. Therefore, particular care a single experiment, all of the host genes and Atlas mouse cDNA expression array should be used to standardize conditions. those of the bacteria whose expression is (CLONTECH)]. Homemade chips can After hybridization, laser scanning, image modified during host–pathogen interaction: also be printed when the genes are avail- detection, and analysis of each chip, the the global picture of the dialogue be- able. Pat Brown (Stanford University, ratio of the probes in Fig. 1 b and c tween the pathogen and the host in one Stanford, CA) is the leader in printing provides a precise indication of the rela- experiment! homemade high-density arrays. In col- tive expression of the genes of the mam- laboration with Lou Staudt of the Na- malian cells grown in the conditions in Fig. Microarrays to Study Host–Pathogen Interac- tional Cancer Institute, he has created 1 b and c. More specifically, for the label- tions. A typical microarray experiment is the ‘‘Lymphochip,’’ which contains shown in Fig. 1. RNA is prepared from about 18,432 human gene elements (5), See companion article on page 13847. bacteria grown in standard laboratory and he is now collaborating with David *E-mail: rino࿝[email protected]. conditions (Fig. 1a), from host cells Botstein (Stanford University) on a sep- Ͼ Article published online before print: Proc. Natl. Acad. Sci. grown under optimal conditions in tissue arate array with 25,000 elements USA, 10.1073͞pnas.010505497. Article and publica- culture (b), and from cells that have been (genes) on a single glass microscope tion date are at www.pnas.org͞cgi͞doi͞10.1073͞ infected with bacteria (c). A probe is slide. Chips representing the whole ge- pnas.010505497 PNAS ͉ December 5, 2000 ͉ vol. 97 ͉ no. 25 ͉ 13467–13469 Downloaded by guest on September 30, 2021 reaction of mammalian cells to infection. The neutrophil chemoattractant chemo- kines IL8 and GRO␤, which are responsible for chemotaxis and activation of effector inflammatory cells, are activated in most systems. Similarly, chemokines and cyto- kines, which are directly responsible for inflammations such as IL1, IL6, IFN␤, MCP1, and tumor necrosis factor ␣, and transcription factors involved in the in- flammatory response such as NF␬B, are affected. Pathology, Pathogenesis, and Pharmaco- genomics. In addition to the simple obser- vation of host-gene up- and down- regulation, microarrays can also be used to ask very specific questions about the clin- ical manifestation of a disease and the role in pathogenesis of individual virulence factors and to predict the clinical outcome of specific drugs. This is the novel concept introduced by Belcher et al. (4). Links to pathology have already been made in this work. Infiltrates of monocytes, neutrophils, and lymphocytes observed in the lungs of B. pertussis-infected animals are explained by the activation of the chemo- tactic and proinflammatory response in- duced in host cells. Diffuse bronchopneu- monia with increased secretion of mucus observed in autopsies of pertussis victims can be explained by the increased secretion of mucins induced by B. pertussis infection. Fig. 1. Schematic description of how microarrays can be used to study host-cell–pathogen interactions. The hypothesis generated by the microarray experiment that increased mucin expression could play a role in pathogenesis was then ing reaction [RNA in Fig. 1b (Cy3)͞RNA response of macrophages and epithelial cells experimentally confirmed by showing that in Fig. 1c (Cy5)], a red dot indicates gene respectively to Salmonella infection and B. pertussis binds mucin in vitro. et al. activation after bacterial addition to the Ichikawa (9) reported the response of B. pertussis is an ideal bacterial pathogen cells, and a green dot means down- the A549 pneumocyte cell line exposed to to study pathogenesis. The bacterium pro- regulation of the gene, whereas a yellow Pseudomonas aeruginosa infection. In this duces many virulence factors that are well dot indicates no change in gene expres- issue of PNAS, Belcher et al. (4) report the characterized from the biochemical and ge- sion. The final results of the analysis are change in gene expression of the simian ͞ netic point of view: toxins such as pertussis the expression ratios of Fig. 1 b c obtained virus 40-transformed human bronchial epi- toxin (PT), adenylate cyclase, dermon- from the mean value of two labeling con- thelial cell line BEAS-2B after infection ecrotic toxin, tracheal cytotoxin, and ad- ditions (direct and reverse labeling). with Bordetella pertussis. All studies report hesins such as filamentous hemagglutinin, Similarly, genome-wide changes in bacte- genes that are up-regulated and those that pertactin, fimbriae, etc. Furthermore, rial gene expression after contact with eu- are down-regulated after incubation with knockout isogenic mutants are available for karyotic cells can be obtained (Fig. 1d)by bacterial pathogens [for instance, 74 were these factors. Comparing the host-cell re- hybridizing bacterial chips with a mixture of up and 23 down in Cohen et al. (6) and 33 sponse to wild-type B. pertussis and each of theprobesinFig.1a and c. It should be kept up and 65 down in Belcher et al. (4)]. these mutants will help understand how in mind that DNA arrays provide a measure Up-regulated and down-regulated genes be- each contributes to the clinical outcome. of the average mRNA present (steady-state long to different functional clusters (inflam- Belcher et al.