HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 26, 24223-24231, June 29, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/26/24223    most recent M100951200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Blader, I. J. Articles by Boothroyd, J. C. Search for Related Content PubMed PubMed Citation Articles by Blader, I. J. Articles by Boothroyd, J. C. Social Bookmarking                      What's this?

Microarray Analysis Reveals Previously Unknown Changes in Toxoplasma gondii-infected Human Cells^*

Ira J. Blader, Ian D. Manger^§, and John C. Boothroyd^¶

From the Department of Microbiology and Immunology, Stanford University, Stanford, California 94305-5124

Received for publication, February 1, 2001, and in revised form, April 4, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cells infected with the intracellular protozoan parasite Toxoplasma gondii undergo up-regulation of pro-inflammatory cytokines, organelle redistribution, and protection from apoptosis. To examine the molecular basis of these and other changes, gene expression profiles of human foreskin fibroblasts infected with Toxoplasma were studied using human cDNA microarrays consisting of ~22,000 known genes and uncharacterized expressed sequence tags. Early during infection (1-2 h), <1% of all genes show a significant change in the abundance of their transcripts. Of the 63 known genes in this group, 27 encode proteins associated with the immune response. These genes are also up-regulated by secreted, soluble factors from extracellular parasites indicating that the early response does not require parasite invasion. Later during infection, genes involved in numerous host cell processes, including glucose and mevalonate metabolism, are modulated. Many of these late genes are dependent on the direct presence of the parasite; i.e. secreted products from either the parasite or infected cells are insufficient to induce these changes. These results reveal several previously unknown effects on the host cell and lay the foundation for detailed analysis of their role in the host-pathogen interaction.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Toxoplasma gondii, which is an obligate intracellular apicomplexan parasite that can infect most nucleated cells, causes devastating disease in humans and is related to Plasmodium, the causative agent of malaria (1). Immediately following invasion, T. gondii resides and replicates within a parasitophorous vacuole that, although free of host membrane proteins, is surrounded by host mitochondria and endoplasmic reticulum (2). Once the parasite begins growing and dividing within the parasitophorous vacuole, it must acquire nutrients such as purine nucleosides and cholesterol from the host cell (3). The mechanisms by which the morphological and metabolic changes to the host cell occur are unknown.

T. gondii infection also causes the induction of numerous immune modulators that activate both the innate and adaptive immune responses (4-7). Both live parasites and parasite extracts stimulate IFN-¹ synthesis, which is required for protection, primarily via dendritic cell activation (6, 8, 9). However, the mechanism by which infected cells activate and recruit dendritic cells remain unclear but presumably requires synthesis of C-C chemokines that bind and activate the C-C chemokine receptor, CCR5 (10). Although previous data demonstrated that T. gondii-infected cells produce these chemokines, the molecular mechanisms that regulate their expression are still unknown (4, 5, 7). Moreover, the full repertoire of host genes whose expression is modulated by T. gondii as well as the signal transduction cascades underlying these changes are also unknown.

Recently, the study of host-parasite interactions has been greatly aided by large-scale gene expression analysis using DNA microarrays (11). Spotted with grids of highly dense and organized spots of DNA, cDNA microarrays are hybridized to two different cDNA samples each labeled with a different fluorophore thus allowing the simultaneous monitoring of differential gene expression for thousands of genes (12). Recent studies using microarrays have enhanced our knowledge of the host response to viruses (e.g. human immunodeficiency virus, cytomegalovirus, and coxsackievirus) and bacteria (e.g. Salmonella typhimurium, Pseudomonas aeruginosa, and Listeria monocytogenes) (13-18). Microarrays demonstrated that the transcriptional response of macrophages to lipopolysaccharide and Salmonella are largely overlapping, suggesting that lipopolysaccharide may be largely responsible for the inflammatory response to Salmonella (14).

The goal of our study was to analyze the global changes that occur when a protozoan intracellular parasite infects a mammalian host cell. Thus, we examined the transcriptional profile of HFFs in response to T. gondii infection. Intact parasites, as well as soluble parasite-derived factors, rapidly increase the abundance of transcripts associated with the immune response. Later following infection, numerous host genes were modulated that are involved in diverse cellular processes, including metabolism, transcription, protein targeting, and apoptosis. Finally, we were able to discriminate between genes that are modulated due to either parasite- or host cell-derived secreted factors and those that are modulated only in the presence of intracellular parasites.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cells, Parasites, and Reagents-- HFFs (passage 10-16 post-isolation) were used for all assays. The PDS strain of T. gondii, cloned from ME49 (19), was routinely passaged in HFFs using standard T. gondii culture conditions (19) with DMEM supplemented with 10% fetal bovine serum, 2 mM glutamine, and gentamicin (Life Technologies; Rockville, MD). HFF and parasite strains were regularly inspected for mycoplasma, and found to be negative, using a Mycoplasma PCR ELISA detection kit (Roche Molecular Biochemicals; Indianapolis, IN).

Parasite Preparation and Infection Assays-- Parasites were harvested from infected HFF monolayers when lysis by infection was almost completed and few, if any, intact host cells remained. The flasks were then scraped, and the entire material harvested and passed twice through a 27-gauge needle to rupture any remaining HFFs and release the parasites within. The resulting material was pelleted at 600 × g, washed three times in serum-free DMEM, and parasites counted by light microscopy. No intact host cells could be detected in this preparation. Parasites were added to HFF monolayers (4-7 days old) at an multiplicity of infection of 5-10. Infection assays were routinely performed with HFF monolayers in 75-cm² flasks grown in 15 ml of media or a proportionate volume for different size flasks for the indicated times. For the Transwell experiments, HFFs were plated in the lower chamber of a 10-cm Transwell tissue culture insert (final volume 10 ml; Corning-Costar, Corning, NY) with 5 ml of media in the upper chamber. Parasites, in serum free media, were added to the upper chamber, and the plates were incubated for 4 h.

RNA Preparation-- Total RNA was prepared using the RNAeasy Midi kit (Qiagen, Valencia, CA). The amount of T. gondii total RNA co-purified 24 hpi was estimated to be ~20% by comparing ribosomal RNA band intensities of total RNA separated by agarose gel electrophoresis (not shown).

Microarray Hybridization-- cDNA microarrays were synthesized at Stanford University using standard protocols (12). The microarrays were spotted with 18,000-27,000 sequence-verified clones that are available from Research Genetics (Huntsville, AL). cDNA probe preparation was performed essentially as described previously (12). Briefly, total RNA (20-25 µg) was converted to first-strand cDNA using Superscript II (Life Technologies, Rockville, MD) with oligo-dT₁₈ (New England BioLabs, Bedford, MA). The cDNA was labeled with Cy3-dUTP (infected) or Cy5-dUTP (uninfected) (Amersham Pharmacia Biotech, Uppsala, Sweden) via a random prime reaction using random nanomer. We found that labeling cDNA with random nanomer versus direct dye incorporation during the first-stand cDNA synthesis reaction (12) did not indicate any significant difference between the two protocols (not shown). Microarrays were hybridized overnight at 65 °C, washed as described (12) and scanned with a GenePix 4000 microarray scanner (Axon Instruments, South San Francisco, CA). Each time point and condition was repeated typically three-five times.

Data Analysis-- Microarrays were analyzed with the Scanalyze program (written by Mike Eisen and available on the Web) to determine the fluorescent intensities of the two dyes for each spot. The fluorescence intensities were normalized by applying a scaling factor so that the median fluorescence ratio of all spots with detectable signals above background on each microarray was 1.0 (20). The data were then filtered such that only spots with intensities that were three times greater than background in either channel were used in the analysis. Only those spots that displayed a 2-fold or greater difference in fluorescence intensities between the two dyes were used to generate gene clusters. For all such clusters, each spot was manually examined to assess its quality and those that exhibited poor quality throughout the analysis were removed. Spots of poor quality in individual microarrays were discarded from the dataset and are represented in the gene clusters as gray bars. Poor quality spots were removed if they were 1) very small, 2) irregularly shaped, or 3) with pixels that were not uniformly distributed throughout the spot. In the gene clusters, black bars represent genes that displayed -fold changes of 2 or less. Data clustering was performed using the Cluster program and figures generated using Tree View (both available on the Web) (21).

Northern Blot-- Total RNA (5 µg) was separated by agarose gel electrophoresis using standard molecular biological techniques. GRO1 cDNA was prepared by RT-PCR from total RNA from T. gondii-infected HFFs using the following oligonucleotides: GRO1-F 5'-CGGAAAGCTTGCCTCAATCCT-3' and GRO1-R 5'-GATCCGCCAGCCTCTATCACA-3'. -actin (IMAGE: 867606) and DKK1 (IMAGE: 669375) cDNAs were purchased from Research Genetics (Huntsville, AL) and were sequenced-verified. Probes were labeled with [-³²P]dGTP with the Random Primed DNA labeling kit (Roche Molecular Biochemicals, Germany) and hybridized with Express-Hyb (CLONTECH, Palo Alto, CA) according to the manufacturer's instructions. Blots were exposed to film, and the autoradiographs were scanned and analyzed using the ImageQuaNT analysis program (Molecular Dynamics, Sunnyvale, CA)

ELISA-- Secreted GRO1 protein was detected by ELISA using the GRO1 ELISA kit (R&D Systems, Minneapolis, MN) according to the manufacturer's protocol.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Human Microarrays to Study Host Response to T. gondii-- To characterize the host cell response to T. gondii infection, we employed cDNA microarray analysis, which is a powerful tool for analyzing global changes in gene expression. Because this technique can be subject to variability due to a combination of RNA preparation and handling, spot quality, and data analysis (15), it is necessary first to determine the experimental error for the conditions being used. Once determined, these variables allow thresholds to be set below which changes are not regarded as significant. Thus, HFF cultures were mock-infected or infected with T. gondii tachyzoites for 14 h in duplicate, and RNA was isolated from infected and mock-infected cells. HFFs were used in this study, because they are a well characterized T. gondii cell culture model. Moreover, they are non-transformed and do not require pretreatment with differentiation factors, such as growth factors or phorbol esters, which limit the extent of variability between cell preparations. Microarrays were then hybridized with cDNA probes synthesized using RNA from the infected and the mock-infected samples. We compared the difference between infected and uninfected cells for each spot whose signal intensity in either channel was at least 500 fluorescence units, which corresponds to three times greater than background. Using this filter, 11,976 spots (45% of the total number of spots) were included in the analysis. To assess assay reproducibility, the fold difference for each spot was plotted for the two duplicate experiments (Fig. 1A). The data showed a linear regression of 0.88, indicating a generally high level of correlation. Next, the 11,976 spots were filtered to include only spots of high quality that displayed a fluorescence ratio between infected and uninfected cells of >2 in either microarray. Using these criteria, a total of 1503 spots (~6% of the original 11,976 spots) were analyzed with a regression correlation of r = 0.95 (Fig. 1B). Similar analysis demonstrated that no significant bias was observed when the Cy3-dUTP and Cy5-dUTP labels were reversed (not shown).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Microarray assay reproducibility. A, cDNA, prepared from mock- or parasite-infected HFFs for 14 h in duplicate, were hybridized to microarrays. Scatter plot depicting the Cy5-dUTP/Cy3-dUTP ratio for corresponding spots between the two microarrays. Spots were included that had a signal strength of three times over background in both microarrays (n = 11,976, 45% of total). The black line indicates the regression correlation (r = 0.88), and the red line represents a perfect correlation of 1.0. B, spots from A were filtered for those of high quality and displayed a -fold change of two or greater in at least one of the microarrays (n = 1503, 6% of total). The black line indicates the regression correlation (r = 0.95). C, gene cluster of the data from B. Each row represents a spot on the microarray and each column a separate microarray. Note the high degree of correlation (represented by color intensities) between the two experiments.

Because T. gondii is an intracellular parasite, the RNA preparation protocol used resulted in the purification of not only HFF RNA but also parasite RNA. Thus, the apparent gene modulation observed could have conceivably been an artifact due to cross-hybridization between T. gondii cDNA and spots on the human microarray. To control for this, a human cDNA microarray was hybridized with Cy5-dUTP-labeled cDNA probe made from 20 µg of uninfected HFF RNA and Cy3-dUTP-labeled cDNA probe made from 16 µg of uninfected HFF RNA to which 4 µg (20%) total parasite RNA, prepared from washed and filtered tachyzoites, was added. Data analysis indicated that T. gondii cDNA does not cross-hybridize to any measurable extent with any spot on the human microarray (not shown).

Because the parasites were cultured in vitro in HFF monolayers and purified by syringe-lysis, host cell debris remaining in the parasite preparation could have modulated gene expression in uninfected monolayers. Thus, the response of HFFs infected for 2 h with T. gondii were compared with HFFs treated with an equivalent preparation of uninfected HFFs, which were scraped and prepared using conditions identical to those used to prepare the parasites. Two different microarrays were hybridized with either cDNA from the parasite-infected or debris-treated cells against cDNA from untreated HFFs. The microarray from the cell debris-treated sample showed no significant changes in gene expression whereas the infected cells displayed the usual profile of gene induction (not shown). These data indicate that the changes in gene expression detected by microarrays are real, reproducible, and the result of T. gondii infection.

T. gondii Stimulates Rapid Induction of Pro-inflammatory Genes-- To dissect the temporal response of HFFs infected with T. gondii, a time course was performed during which RNA was purified from HFFs that were mock- or parasite-infected for 1, 2, 4, 6, and 24 hpi with T. gondii tachyzoites. To minimize host cell lysis following parasite egress, the time course was limited to 24 h and we used the PDS strain, a ME49 clonal derivative, that grows more slowly than the highly virulent Type 1 RH strain (doubling time of ~12 h (PDS) versus ~6-8 h (RH)) (19). For each time point, a microarray was used to compare cDNA prepared from mock-infected versus parasite-infected cells. Following hybridization, the data were filtered based on the criteria described above: spots were three times greater than background in at least one channel, of high quality, and displayed a 2-fold change or greater (relative to uninfected) in at least one time point during the experiment. The filtered data were organized into temporal gene expression clusters. This approach was previously used to classify cell-cycle-regulated genes in Saccharomyces cerevisiae and identify serum responsive genes in HFFs (21-23). Clustering of the time course data indicated that two distinct gene expression patterns were present: genes that are modulated rapidly after infection (1-2 hpi) and those that are modulated later (>6 hpi) (Fig. 2). Due to the large number of genes modulated during the time course, only data related to genes that have some functional information are shown and discussed. The full group, including ESTs and genes not discussed here, can be downloaded from J.C.B.'s website at Stanford University.

View larger version (51K): [in this window] [in a new window]   Fig. 2.   Transcriptional response of HFFs infected with T. gondii for 1-24 h. A, gene cluster of spots of high quality representing unique, known genes that change >2-fold at least once during the time course. B, enlarged view of the cluster from A showing the genes that are modulated 1-2 hpi. *, genes known to be associated with the immune response.

During the first 2 h of infection, a total of 63 unique, known genes exhibited a 2-fold or greater increase in their abundance, whereas 15 unique, known genes exhibited a 2-fold or greater decrease. A high proportion of the induced genes (27 genes, 43%), are genes previously known to be associated with the immune response. These immunomodulatory molecules include chemokines (GRO1, GRO2, LIF, and MCP1), cytokines (IL-1, IL-6), cell matrix and adhesion proteins (ICAM1 and matrix metalloproteinase 3), apoptotic (superoxide dismutase 2) and transcriptional regulatory factors (REL-B, NF-B p105, I-B). Several additional cytokines previously shown to be important in T. gondii infection, such as -interferon, IL-10, and IL-12 (24, 25), were not among the genes spotted on the microarrays and therefore were not included in this analysis. Although TNF- and IFN-, which is not normally expressed in HFFs, were not spotted on the microarrays used for this time course, subsequent microarray experiments, using microarrays that included TNF- and IFN-, showed that they were not induced even by 24 hpi (not shown). Although the early induced genes are predominantly associated with the immune response, these represent a relatively small number of immune response genes spotted on the microarrays. For example, only 4 out of 24 genes encoding chemokines and 2 out of 41 genes encoding interleukins or their receptors were up-regulated by T. gondii 2 hpi.

Northern Blot Confirmation of Microarray Data-- Although microarrays are a powerful tool to study global transcriptional responses to infection, it is important to confirm microarray data using independent methods such as Northern blot analysis, ribonuclease protection assay, or quantitative reverse transcription-polymerase chain reaction. Northern blots of representative genes from among those induced, repressed, or unchanged by infection were performed to verify the microaray data. RNA prepared from uninfected and infected cells (2 and 24 hpi) were used to make cDNA for hybridization to the microarrays and for analysis by Northern blotting using probes for an up-regulated gene (GRO1), a down-regulated gene (DKK1), and an unmodulated gene (-actin). The Northern blot data were consistent with the microarrays providing an independent validation of the microarray data (Fig. 3). Although -actin remained unchanged by both microarray and Northern blot analysis, it was observed that GRO1 was up-regulated and DKK1 was down-regulated with both techniques. Importantly, the Northern blot data corroborated the microarray GRO1 induction trend, i.e. 8- and 20-fold induction (2 hpi) and 38- and 95-fold induction (24 hpi) (microarray and Northern, respectively).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Northern blot analysis of microarray expression data. RNA was purified from mock- and parasite-infected HFFs 2 hpi (right panel) and 24 hpi (left panel) and hybridized to microarrays. The same total RNA was separated by agarose gel electrophoresis and Northern-blotted with probes against GRO1 (induced), DKK1 (repressed), and -actin (unmodulated).

To determine whether the change in GRO1 mRNA levels reflects a change in the GRO1 protein produced, we used an ELISA-based assay to determine the amount of this chemokine in the medium of uninfected versus T. gondii-infected HFF cultures. The results indicated that 4 hpi of T. gondii stimulated an increase in the GRO1 concentration from ~600 to ~9000 pg/ml (data not shown). These results are in good agreement with the 8- to 20-fold induction of GRO1 mRNA indicated by the microarray and Northern blot data.

T. gondii Up-regulates the Abundance of Host Glycolytic and Mevalonate Metabolic Transcripts-- To examine the late pattern of gene expression in greater detail, two independent 24-h infections were analyzed with microarrays. The data were filtered to identify genes whose spots were of high quality and in both experiments were up- or down-modulated at least 2-fold. Using these criteria, 567 unique genes were identified of which 376 corresponded to previously characterized genes with some functional information while 191 were ESTs or genes of unknown function. Categorizing the known genes by function indicated that numerous cellular processes are modulated during T. gondii infection. These included genes involved in carbohydrate and lipid metabolism (13.7%); transcriptional regulation (13.2%); protein synthesis, targeting, and degradation (12.3%); cell signaling (11.4%); inflammation (8.2%); cell adhesion and cytoskeleton (8.2%); nucleotide and amino acid metabolism (4.7%); cell cycle (4.1%); and apoptosis (3%). It is important to note that this type of functional annotation may be misleading, because some genes may be involved in more than one process but have been arbitrarily grouped into a single class. For example, although ICAM1 is an adhesion molecule, its expression is often associated with the inflammatory response (26). Because the list of modulated known genes and unknown ESTs is too large to be presented in this report, the list can be downloaded from J.C.B.'s website at Stanford University. Examination of those down-regulated genes revealed no clear trends for any group of genes involved in a particular function or pathway. Examination of the up-regulated genes, however, showed some clear functional groupings and is discussed below.

Similar to other intracellular parasites, T. gondii must scavenge from its host nutrients such as glucose and cholesterol (27). Interestingly, our results show that many enzymes associated with these pathways are up-regulated during infection.

Glycolysis-- Glycolysis is a 10-step metabolic pathway that converts glucose to 2 molar equivalents of pyruvate with a net yield of 2 molar equivalents of ATP. Genes encoding each of the host glycolytic enzymes were present on these microarrays. Of these genes, hexokinase 2, phosphofructokinase 1, triose phosphate isomerase 1, phosphoglycerate kinase, phosphoglycerate mutase 1, and enolase 2 were significantly up-regulated 2-fold (Fig. 4). However, glucose phosphate isomerase, aldolase A, and pyruvate kinase 1 were not significantly changed. Glyceraldehyde-3-phosphate dehydrogenase was up-regulated (3.7-fold) in one experiment, and in the second, it was up-regulated ~1.8-fold, which is below the minimal threshold of significance.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Transcriptional changes of genes involved in glycolysis and squalene biosynthesis. Shown are the -fold changes for genes 24 hpi encoding enzymes involved in glycolysis (left) and squalene metabolism (right).  (Induced); NC (no change); PS (poor spot); NP (not printed on the microarrays).

Depending on the aerobic status of the host, pyruvate is further metabolized to either lactate or acetyl-CoA. Microarray data indicated that lactate dehydrogenase-A (muscle isoform) was up-regulated 24 hpi. Besides lactate dehydrogenase-A, two other lactate dehydrogenase isoforms have been previously characterized (lactate dehydrogenase-B (heart isoform) and lactate dehydrogenase-C (sperm isoform)) (28). Of these, lactate dehydrogenase-B was not spotted on the microarrays, and the spot for lactate dehydrogenase-C was of insufficient intensity to be included in our analysis, suggesting that it is not expressed in HFFs. Acetyl-CoA, which is generated from pyruvate by the pyruvate dehydrogenase complex, can be utilized by the TCA cycle that is coupled to oxidative phosphorylation. We found that expression of pyruvate dehydrogenase complex genes was not significantly changed during T. gondii infection. Moreover, pyruvate dehydrogenase kinase isoenzyme 1, which is a negative regulator of pyruvate dehydrogenase, was up-regulated ~2.5-fold. Fructose-bisphosphatase, which catalyzes the rate-limiting reaction in gluconeogenesis, was not significantly altered 24 hpi. Genes encoding components of the nine TCA cycle holoenzymes, except succinyl-CoA synthase, were spotted on the microarrays, and none were significantly induced or repressed during infection. Similarly, no significant changes were observed in the expression of genes encoding enzymes involved in either amino acid catabolism, which yield either pyruvate or acetyl-CoA, or in the pentose phosphate pathway. Together, these data suggest that T. gondii stimulates the specific up-regulation of transcripts involved in anaerobic glycolysis but not oxidative phosphorylation.

Cholesterol Biosynthesis-- T. gondii cannot synthesize sterols via the mevalonate pathway and, therefore, must obtain them from the host cell (29). There are three mechanisms that cells employ to replenish depleted cholesterol stores: 1) stimulation of de novo cholesterol biosynthesis via the mevalonate pathway; 2) enhanced LDL uptake; or 3) mobilization of intracellular cholesterol esters. Although T. gondii appears to acquire cholesterol by subverting the LDL trafficking pathway (29), it is not clear what effect this has on the host. De novo cholesterol biosynthesis is dependent on the mevalonate pathway that metabolizes HMG-CoA to squalene. Examination of the genes on the microarrays encoding enzymes involved in the mevalonate pathway indicated that several were induced 24 hpi. These include the rate-limiting enzyme HMG-CoA reductase, diphosphomevalonate decarboxylase, and farnesyl pyrophosphate synthase. Mevalonate 5-phosphotransferase and phosphomevalonate kinase showed no significant change while the spot for isopentylpyrophosphate isomerase was of insufficient quality in all replicas and thus was not included in the analysis. Moreover, squalene epoxidase, which is the second enzyme in the committed cholesterol biosynthetic pathway, producing 2,3-oxidosqualene, was up-regulated 5-fold following infection, suggesting that induction of mevalonate biosynthetic enzymes may be necessary to increase levels of cellular squalene. Farnesyl pyrophosphate, the squalene precursor, is also a substrate for several other metabolic pathways, including dolichol and terpenoid biosynthesis. Two key enzymes in dolichol metabolism, dolichyl-phosphate N-acetylglucosamine phosphotransferase 1 and dolichyl-phosphate mannosyltransferase, which were spotted onto the microarrays, were not significantly changed 24 hpi. Regulation of these genes may be very complex, because the induction of mevalonate metabolic genes did not always occur at 24 hpi. The microarray experiments were repeated six times, and the genes were significantly induced in five of six experiments, including the two used for this analysis. These data indicate that infection with T. gondii stimulates up-regulation of several key genes involved in mevalonate metabolism.

Infected Conditioned Medium Reveal a Role for Soluble Factors in Modulating the Transcriptional Response-- One drawback of these assays is that it is not possible to discriminate between those genes modulated by intracellular parasites and those modulated by factors secreted from either infected host cells or from the parasites themselves. For example, the microarray data indicate that T. gondii-infected HFFs synthesize IL-1. The release of these proteins may activate neighboring cells regardless of whether they are infected or not. Moreover, T. gondii secretes proteins from intracellular organelles and sheds cell surface proteins that, similar to intact parasites, bind cells and stimulate chemokine and cytokine expression (30-34). Thus, it is necessary to determine which genes are modulated directly by infection and which are modulated by soluble factors derived from either the infected host or the parasite itself.

First, filtered conditioned media from mock- and parasite-infected cells was used to discriminate between genes regulated only in the presence of T. gondii and genes modulated by secreted factors. Media from 24-hpi mock- and parasite-infected HFFs were collected, filtered, and added to uninfected HFF monolayers. RNA was harvested at 4, 14, and 24 h from each conditioned media-treated monolayer as well as from the 24-hpi mock- and parasite-infected cells from which the conditioned media was obtained. Each microarray was hybridized with labeled cDNA prepared from each time point and analyzed and clustered as in Fig. 1. Gene clustering indicated that two classes of genes were modulated during infection: 1) those regulated by both conditioned media and T. gondii and 2) those regulated only in the presence of T. gondii (Fig. 5 and the dataset can be downloaded from Stanford University on the Web).

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Effect of infected HFF-conditioned media on gene expression in uninfected HFFs. Media from 24-hpi mock- and T. gondii-infected cultures were collected and filtered through a 0.2-µm filter. The filtered conditioned media were added to uninfected HFF monolayers for 4, 14, and 24 h. cDNA, prepared from RNA harvested from the treated cultures as well as the original mock- and T. gondii-infected cultures used to generate the conditioned media, was hybridized to microarrays. Shown is the gene cluster for spots of high quality representing unique genes that change >2-fold. Five major classes of genes are seen in the gene cluster: 1) genes up-regulated by conditioned media and infection, 2) genes up-regulated only during infection, 3) genes modulated only by the conditioned media, 4) genes down-regulated only during infection, and 5) genes down-regulated by infection and conditioned media.

The early response genes that remained up-regulated 24 hpi (such as GRO1, IL-1, IL-6, GTP cyclohydrolase, ICAM 1, and IB alpha) were also up-regulated by conditioned media from T. gondii-infected HFFs. Besides the early response genes, other genes up-regulated 24 hpi were also up-regulated by the conditioned media and include IL-13 receptor alpha 2, IL-7 receptor, and BCL2/adenovirus E1B 19kDa-interacting protein 3. These data indicate that these genes were modulated by secreted factors derived either from the parasite or the infected host cell.

Genes modulated by infection but not by the infected conditioned media suggested that their transcriptional regulation was a direct consequence of intracellular infection and growth. Examination of the mevalonate metabolic genes, which were up-regulated 24 hpi, indicated that they were not up-regulated by the conditioned media. These include HMG-CoA reductase, farnesyl pyrophosphate synthase, and squalene epoxidase. The spot corresponding to the other induced mevalonate metabolic gene, mevalonate decarboxylase, was of poor quality in the microarrays used for these experiments and therefore, not included in this analysis.

Of the glycolytic genes induced by T. gondii 24 hpi and represented by spots of high quality in these experiments, triose phosphate isomerase 1, phosphoglycerate kinase 1, phosphoglycerate mutase 1, and lactate dehydrogenase-A were not induced by the conditioned media. Hexokinase 2 was not spotted on these microarrays, and the spot representing phosphofructokinase 1 was of poor quality. Because glyceraldehyde-3-phosphate dehydrogenase 2 and enolase were induced by the conditioned media, these data indicate that transcriptional regulation of glycolytic genes may be complex and requires further investigation to determine whether their modulation is regulated directly by T. gondii infection or by secreted factors.

Next, HFFs were treated with cycloheximide to prevent new protein synthesis following infection and thus, prevent secondary effects from the first wave of gene expression. HFFs were incubated with or without cycloheximide (100 µg/ml) 30 min prior to infection and then mock- or parasite-infected for 2 h in the continued presence of cycloheximide. The toxic effect of cycloheximide to the HFFs was minimized by determining, using ³⁵S incorporation, that this concentration and time was the minimal amount needed to block >95% de novo protein synthesis (not shown). The expression levels of the early response genes were compared by hybridizing two different microarrays with cDNA prepared from mock- and parasite-infected cultures grown with or without cycloheximide. The majority (>82%) of the early response genes (whose spots were of good quality) were induced in the presence of cycloheximide (Table I). These included GRO1, IL-6, IL-1, c-REL NF-B (p105), ICAM1, interferon regulatory factor 1, c-Jun, LIF, pre-B-cell colony-enhancing factor, immediate early response 3, TNF--induced protein, and TNF--induced protein 3. Thus, induction of early response genes is largely independent on the synthesis of host-derived factors.

Host Gene Regulation by Soluble Toxoplasma-derived Factors-- Previously, it has been demonstrated that intact extracellular T. gondii tachyzoites secrete factors that stimulate expression of some immune response genes (32-34). Therefore, we hypothesized that induction of early response genes in the conditioned media and the cycloheximide experiments may be largely mediated by factors secreted from the parasites. Direct host-parasite interaction was blocked by adding parasites to the upper compartment of a Transwell chamber in which HFFs were plated in the lower compartment. More tachyzoites (5-fold more) were added to the Transwell chamber than normally used to infect a monolayer to allow for the dilution and distance separating parasites from the host cells. After 4 h, RNA was extracted from the HFFs and compared with RNA prepared from mock-treated HFFs in which only media was added to the upper Transwell compartment. In parallel, RNA was prepared from HFF monolayers that were either mock- or parasite-infected for 4 h in the absence of a Transwell chamber. The cDNAs were compared by microarray analysis, and the data indicated that all but one of the early response genes represented by spots of high quality, plasma membrane Ca²⁺ ATPase, were similarly induced by parasites in contact with the host cell as by parasites in the Transwell chamber (Fig. 6). These data indicate that exposure to secreted, soluble parasite-derived factors can stimulate transcription of the early response genes.

View larger version (28K): [in this window] [in a new window]   Fig. 6.   T. gondii-derived soluble, secreted factors modulate gene expression in HFFs. Serum-free DMEM and 1 × 107 or 1 × 108 intact parasites were added to the upper compartment of a 10-cm transwell chamber in which HFFs were plated in the lower compartment. In parallel, HFFs were mock- or parasite-infected in the absence of a transwell. RNA was harvested from HFFs and cDNA hybridized to microarrays. Shown is a gene cluster of the of the early response genes from Fig. 2.

To exclude the possibility that a small number of host cells contaminate the parasite preparation and that in response to the parasite suspension these host cells are responsible for production of the soluble factor, we performed the following experiments. First, we incubated our purified parasites in the same medium and for the same time as used in the transwell experiments and then removed the medium and assayed it for the presence of GRO1. (As discussed above, GRO1 is abundantly produced by infected cells and thus serves as an easily assayed marker for the presence of soluble factors released by such cells.) No GRO1 could be detected in the medium using an ELISA-based assay that is sensitive down to 25 pg/ml (data not shown). These results show that, to the limit of sensitivity of this assay, there are no contaminating host cells in our purified parasite preparation. Next, we determined how much host cell-derived factors (again monitored through GRO1) would be needed to elicit a response. To do this, various dilutions of conditioned media from mock- or T. gondii-infected (4 hpi) cultures were added to uninfected HFFs for 4 h. The net amount of GRO1 produced was determined by subtracting the amount of GRO1 initially added to the cells from the amount of GRO1 produced after the 4-h incubation. The data showed that even a 1:10 dilution of T. gondii-infected conditioned media (which contains 50 pg/ml GRO1) was not able to stimulate a substantial net amount of GRO1 (not shown). Moreover, the amount of GRO1 produced by the cells, to which the 1:10 and lower dilutions of T. gondii-infected conditioned media was added, was comparable to the amount of GRO1 constitutively produced by cells to which only undiluted mock-infected conditioned media was added. Thus, if there are host cells contaminating the parasite suspension, they are too few to produce the amount of host-derived material (e.g. <25 pg/ml GRO1) to activate an autocrine response and, therefore, we conclude that the soluble factor is parasite-derived.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Large-scale expression analysis is rapidly becoming an important tool to study parasite-host interactions. In this study, the response of HFFs to T. gondii was characterized and found that within 1-2 hpi infected cells rapidly respond to infection by up-regulating the expression of numerous genes associated with the immune response. Much of this response does not require direct contact between the parasite and host cell, but rather, is regulated via secreted, parasite-derived factors. In addition, two metabolic pathways were identified that are transcriptionally up-regulated by infection. Finally, experiments were performed to identify candidate host genes that are modulated only in the presence of the parasite.

Although HFFs are not considered a major target for T. gondii in vivo, they do apparently reflect many features seen in other cell types as well as the infected animal. For example, many of the cytokine and chemokine genes induced in HFFs are the same as those induced in monocytes and in vivo during mouse infection (4, 7). Chemokines function to recruit immune cells, such as macrophages, dendritic cells, and neutrophils, to sites of infection, and their synthesis has been proposed as an important component in cell-mediated immunity to T. gondii (35). The cellular source of these chemokines is unknown but has been proposed to originate from both infected cells and other cells recruited to the site of infection (35).

To address whether the pro-inflammatory genes induced early during infection could be a nonspecific response of HFFs to intracellular parasitism (i.e. independent of the pathogen), we compared the response of the same HFFs to another intracellular protozoan parasite, Trypanosoma cruzi, using replicate microarrays and the same data analysis methods as for the T. gondii experiments. We found that only ~4% (23 of 614) of the genes, modulated by T. gondii 24 hpi, were also modulated by T. cruzi 24 hpi.² Of these 23, only three (MCP1, pre-B-cell colony-enhancing factor, and bradykinin receptor B2) are up-regulated 2 hpi with T. gondii, indicating that these genes do not represent a general HFF response to intracellular protozoan parasites.

Although we, like others, are interested in the function of the genes modulated during infection, microarrays also allow the identification of candidate intracellular signaling pathways that are activated by the parasite and lead to host gene regulation. One candidate pathway involves the NF-B family of transcription factors, which is important in both innate and cell-mediated immune responses to T. gondii (36). Stimulation of NF-B-dependent transcription leads to the induction of many of the early response genes that are stimulated within 2 hpi. These include the NF-B subunits REL-B, NF-B p105, as well as I-B, GRO1, GRO2, ICAM1, and c-myc (37, 38). Moreover, MCP1 induction by T. gondii requires NF-B activation (5). In vivo, protection against T. gondii infection requires NF-B activation in immune effector cells, but a role for NF-B activation in the primary infected cell remains unclear (36, 39). The finding here that soluble T. gondii-derived factors, as well as intact parasites, induce a similar repertoire of cytokines and chemokines in HFFs as they do in other cell types (4, 7) suggests that the signaling pathways activated are similar between the different cell types.

Changes in host cell metabolism as a consequence of nutrient scavenging by intracellular parasites is difficult to study biochemically because of the inability to effectively separate parasite-derived activities from the host functions. Thus, the microarray data presented here provide the first description of the response of host glycolytic and mevalonate pathways, which were not previously suspected of being transcriptionally regulated during T. gondii infection. Compared with most other genes involved in carbohydrate or energy metabolism that remained unchanged during infection, glycolytic genes were significantly up-regulated, suggesting that intracellular T. gondii growth may induce an anaerobic environment that may result from a general starvation or stress response. Moreover, these data may represent a metabolic basis for the finding that bradyzoite cysts are predominantly found in tissues (brain and muscle) that are major users of glycolysis (40).

A second explanation for the induction of glycolytic enzymes could be that parasites may use pyruvate or acetyl-CoA as primary carbon sources. This possibility seems unlikely, because transcript levels of host genes encoding enzymes involved in amino acid catabolism to pyruvate or acetyl-CoA remained unchanged during infection. Moreover, many of the glycolytic enzymes are abundantly expressed in T. gondii, suggesting that parasites can utilize glucose as a carbon source (41). A third explanation is that these genes are induced nonspecifically as an incidental by-product of a response needed for other purposes and is discussed further below.

Up-regulation of mevalonate metabolic enzyme transcripts may be due to the fact that T. gondii appears to scavenge cholesterol by subverting LDL trafficking. Thus, the infected cell may sense loss of LDL cellular pools and respond by up-regulating expression of these enzymes. Interestingly, Coppens et al. (29) did not observe any increase in bulk HMG-CoA reductase enzymatic activity in Chinese hamster ovary cells infected with a rapidly growing, highly virulent type I T. gondii strain (RH). The reason for this discrepancy is unclear and may indicate a difference in the flux rate of HMG-CoA through the pathway between uninfected and infected HFFs. In addition, these data may reflect differences between different cell types (HFFs versus Chinese hamster ovary cells) and/or parasite strains (RH versus PDS) used in the two studies.

Host genes modulated during infection represent three functionally distinct classes: 1) genes required for host defense, 2) genes required for parasite growth, and 3) genes incidentally regulated as a consequence of modulating the first two classes. How the genes identified by microarray analysis in this study fit into these three groups and what parasite factors are involved cannot be deduced from the work presented here. However, such assignments will be possible by combining microarrays with biochemical and cell biological analysis of mutant host cells and parasite strains.

	ACKNOWLEDGEMENTS

We thank Drs. Barbara Burleigh, Karen Guilliman, Ed Mocarski, Silvia Vaena, and Tina Wasson-Blader for critical reading of the manuscript and members of the Boothroyd laboratory for helpful discussion. Also, we thank Dr. Jason Lih, Jeannette Lam, and Kimberly Chong for microarray preparation and Dr. Gavin Sherlock and the Stanford Microarray Data base for bioinformatic support and guidance.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported by National Institutes of Health Grant 1F32AI10478-01.

^§ Current address: Mycometrix, 213 E. Grand Ave., South San Francisco, CA 94080.

^¶ Supported by National Institutes of Health Grants AI21423 and AI45057. To whom correspondence should be addressed: Dept. of Microbiology and Immunology, Stanford University, 299 Campus Dr., Stanford, CA 94305-5124. Tel.: 650-723-7984; Fax: 650-723-6853; E-mail: john.boothroyd@Stanford.edu.

Published, JBC Papers in Press, April 9, 2001, DOI 10.1074/jbc.M100951200

² S. Vaena, I. J. Blader, J. C. Boothroyd, and B. Burleigh, manuscript in preparation.

	ABBREVIATIONS

The abbreviations used are: IFN-, interferon ; HFFs, human foreskin fibroblasts; hpi, hours post-infection; TCA, tricarboxylic acid; DMEM, Dulbecco's modified Eagle's medium; ELISA, enzyme-linked immunosorbent assay; IL-1, interleukin-1; TNF-, tumor necrosis factor ; EST, expressed sequence tag; LDL, low density lipoprotein; HMG-CoA, 3-hydroxy-3-methylglutaryl-Coenzyme A.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				E. D. Phelps, K. R. Sweeney, and I. J. Blader Toxoplasma gondii Rhoptry Discharge Correlates with Activation of the Early Growth Response 2 Host Cell Transcription Factor Infect. Immun., October 1, 2008; 76(10): 4703 - 4712. [Abstract] [Full Text] [PDF]

				M. M. Nelson, A. R. Jones, J. C. Carmen, A. P. Sinai, R. Burchmore, and J. M. Wastling Modulation of the Host Cell Proteome by the Intracellular Apicomplexan Parasite Toxoplasma gondii Infect. Immun., February 1, 2008; 76(2): 828 - 844. [Abstract] [Full Text] [PDF]

				V. B. Carruthers and Y. Suzuki Effects of Toxoplasma gondii Infection on the Brain Schizophr Bull, May 1, 2007; 33(3): 745 - 751. [Abstract] [Full Text] [PDF]

				A. E. Fouts and J. C. Boothroyd Infection with Toxoplasma gondii Bradyzoites Has a Diminished Impact on Host Transcript Levels Relative to Tachyzoite Infection Infect. Immun., February 1, 2007; 75(2): 634 - 642. [Abstract] [Full Text] [PDF]

				S. Taylor, A. Barragan, C. Su, B. Fux, S. J. Fentress, K. Tang, W. L. Beatty, H. E. Hajj, M. Jerome, M. S. Behnke, et al. A Secreted Serine-Threonine Kinase Determines Virulence in the Eukaryotic Pathogen Toxoplasma gondii Science, December 15, 2006; 314(5806): 1776 - 1780. [Abstract] [Full Text] [PDF]

				H. Andersson, B. Hartmanova, P. Ryden, L. Noppa, L. Naslund, and A. Sjostedt A microarray analysis of the murine macrophage response to infection with Francisella tularensis LVS. J. Med. Microbiol., August 1, 2006; 55(Pt 8): 1023 - 1033. [Abstract] [Full Text] [PDF]

				M. Okomo-Adhiambo, C. Beattie, and A. Rink cDNA Microarray Analysis of Host-Pathogen Interactions in a Porcine In Vitro Model for Toxoplasma gondii Infection Infect. Immun., July 1, 2006; 74(7): 4254 - 4265. [Abstract] [Full Text] [PDF]

				M. Fahling, R. Mrowka, A. Steege, P. Martinka, P. B. Persson, and B. J. Thiele Heterogeneous Nuclear Ribonucleoprotein-A2/B1 Modulate Collagen Prolyl 4-Hydroxylase, {alpha} (I) mRNA Stability J. Biol. Chem., April 7, 2006; 281(14): 9279 - 9286. [Abstract] [Full Text] [PDF]

				R. E. Molestina and A. P. Sinai Host and parasite-derived IKK activities direct distinct temporal phases of NF-{kappa}B activation and target gene expression following Toxoplasma gondii infection J. Cell Sci., December 15, 2005; 118(24): 5785 - 5796. [Abstract] [Full Text] [PDF]

				X. W. Zhou, B. F. C. Kafsack, R. N. Cole, P. Beckett, R. F. Shen, and V. B. Carruthers The Opportunistic Pathogen Toxoplasma gondii Deploys a Diverse Legion of Invasion and Survival Proteins J. Biol. Chem., October 7, 2005; 280(40): 34233 - 34244. [Abstract] [Full Text] [PDF]

				M. UNNIKRISHNAN and B. A. BURLEIGH Inhibition of host connective tissue growth factor expression: a novel Trypanosoma cruzi-mediated response FASEB J, November 1, 2004; 18(14): 1625 - 1635. [Abstract] [Full Text] [PDF]

				T. N. Ledger, P. Pinton, D. Bourges, P. Roumi, H. Salmon, and I. P. Oswald Development of a Macroarray To Specifically Analyze Immunological Gene Expression in Swine Clin. Vaccine Immunol., July 1, 2004; 11(4): 691 - 698. [Abstract] [Full Text]

				L. R. Portugal, L. R. Fernandes, G. C. Cesar, H. C. Santiago, D. R. Oliveira, N. M. Silva, A. A. Silva, J. Lannes-Vieira, R. M. E. Arantes, R. T. Gazzinelli, et al. Infection with Toxoplasma gondii Increases Atherosclerotic Lesion in ApoE-Deficient Mice Infect. Immun., June 1, 2004; 72(6): 3571 - 3576. [Abstract] [Full Text] [PDF]

				H. Zhao, M. L. Whitfield, T. Xu, D. Botstein, and J. D. Brooks Diverse Effects of Methylseleninic Acid on the Transcriptional Program of Human Prostate Cancer Cells Mol. Biol. Cell, February 1, 2004; 15(2): 506 - 519. [Abstract] [Full Text] [PDF]

				R. E. Molestina, T. M. Payne, I. Coppens, and A. P. Sinai Activation of NF-{kappa}B by Toxoplasma gondii correlates with increased expression of antiapoptotic genes and localization of phosphorylated I{kappa}B to the parasitophorous vacuole membrane J. Cell Sci., November 1, 2003; 116(21): 4359 - 4371. [Abstract] [Full Text] [PDF]

				T. M. Payne, R. E. Molestina, and A. P. Sinai Inhibition of caspase activation and a requirement for NF-{kappa}B function in the Toxoplasma gondii-mediated blockade of host apoptosis J. Cell Sci., November 1, 2003; 116(21): 4345 - 4358. [Abstract] [Full Text] [PDF]

				B. Lin, M. T. Vahey, D. Thach, D. A. Stenger, and J. J. Pancrazio Biological Threat Detection via Host Gene Expression Profiling Clin. Chem., July 1, 2003; 49(7): 1045 - 1049. [Abstract] [Full Text] [PDF]

				K. Natarajan, M. S. Rajala, and J. Chodosh Corneal IL-8 Expression Following Adenovirus Infection Is Mediated by c-Src Activation in Human Corneal Fibroblasts J. Immunol., June 15, 2003; 170(12): 6234 - 6243. [Abstract] [Full Text] [PDF]

				A. B. van 't Wout, G. K. Lehrman, S. A. Mikheeva, G. C. O'Keeffe, M. G. Katze, R. E. Bumgarner, G. K. Geiss, and J. I. Mullins Cellular Gene Expression upon Human Immunodeficiency Virus Type 1 Infection of CD4+-T-Cell Lines J. Virol., December 20, 2002; 77(2): 1392 - 1402. [Abstract] [Full Text] [PDF]

				E. S. Mocarski Jr. Virus self-improvement through inflammation: No pain, no gain PNAS, March 19, 2002; 99(6): 3362 - 3364. [Full Text] [PDF]

				S. V. de Avalos, I. J. Blader, M. Fisher, J. C. Boothroyd, and B. A. Burleigh Immediate/Early Response to Trypanosoma cruzi Infection Involves Minimal Modulation of Host Cell Transcription J. Biol. Chem., January 4, 2002; 277(1): 639 - 644. [Abstract] [Full Text]

This Article Abstract Full Text (PDF) All Versions of this Article: 276/26/24223    most recent M100951200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Blader, I. J. Articles by Boothroyd, J. C. Search for Related Content PubMed PubMed Citation Articles by Blader, I. J. Articles by Boothroyd, J. C. Social Bookmarking                      What's this?