Genome-wide transcriptional changes induced by phagocytosis or growth on bacteria in Dictyostelium

Abstract

Background: Phagocytosis plays a major role in the defense of higher organisms against microbial infection and provides also the basis for antigen processing in the immune response. Cells of the model organism Dictyostelium are professional phagocytes that exploit phagocytosis of bacteria as the preferred way to ingest food, besides killing pathogens. We have investigated Dictyostelium differential gene expression during phagocytosis of non-pathogenic bacteria, using DNA microarrays, in order to identify molecular functions and novel genes involved in phagocytosis.

Results: The gene expression profiles of cells incubated for a brief time with bacteria were compared with cells either incubated in axenic medium or growing on bacteria. Transcriptional changes during exponential growth in axenic medium or on bacteria were also compared. We recognized 443 and 59 genes that are differentially regulated by phagocytosis or by the different growth conditions (growth on bacteria vs. axenic medium), respectively, and 102 genes regulated by both processes. Roughly one third of the genes are up-regulated compared to macropinocytosis and axenic growth. Functional annotation of differentially regulated genes with different tools revealed that phagocytosis induces profound changes in carbohydrate, amino acid and lipid metabolism, and in cytoskeletal components. Genes regulating translation and mitochondrial biogenesis are mostly up-regulated. Genes involved in sterol biosynthesis are selectively up-regulated, suggesting a shift in membrane lipid composition linked to phagocytosis. Very few changes were detected in genes required for vesicle fission/fusion, indicating that the intracellular traffic machinery is mostly in common between phagocytosis and macropinocytosis. A few putative receptors, including GPCR family 3 proteins, scaffolding and adhesion proteins, components of signal transduction and transcription factors have been identified, which could be part of a signalling complex regulating phagocytosis and adaptational downstream responses.

Conclusion: The results highlight differences between phagocytosis and macropinocytosis, and provide the basis for targeted functional analysis of new candidate genes and for comparison studies with transcriptomes during infection with pathogenic bacteria.

Figure 1

Time course of gene expression in cells incubated in axenic medium or with E. coli. Exponentially growing AX2 cells were washed from axenic medium, resuspended at a concentration of 5 × 10⁶ per ml and incubated either in axenic medium or with a 1000-fold excess of E. coli B/r in Soerensen phosphate buffer. At the time indicated, cells were washed and the RNA extracted with TRIzol as described in Methods. After agarose electrophoresis, the Northern blots were hybridized with the DNA probes indicated on the left. (Right) The intensity of the bands in Northern blots was calculated using ImageJ and normalized to the value of the histone H1 gene as internal control. The normalized values at each time point were divided for the values at time 0 and expressed as fold increase or decrease, setting the 0-point to 1.

Figure 2

Venn diagram of differentially regulated genes. The total number of up- and down-regulated genes in each comparison and the numbers of differentially regulated genes in the various clusters of the three comparisons are shown in black. Arrows indicate up- or down-regulated genes. The seven possible clusters of the diagram (grey numbers) can be grouped in three larger clusters (table below): cluster 1/5/7 includes genes regulated by phagocytosis, cluster 4 genes regulated by phagocytosis and growth on bacteria, and cluster 2/6 genes regulated by growth on bacteria. "Phagocytosis" means all stimuli that may arise from bacterial binding to the membrane, particle uptake, phagosome intracellular traffic; "growth on bacteria" means stimuli arising from metabolic adaptation to the different nutrients in bacteria vs. axenic medium. See text for details.

Figure 3

Verification of gene expression profiles by Northern blots. Eleven genes were selected randomly to compare their expression profiles obtained with DNA microarray with expression in Northern blots. The intensity of the bands in Northern blots was calculated using ImageJ and normalized to the value of the histone H1 gene as internal control. The log ratio of expression values in comparisons A, B and C is shown in grey for microarray and black for Northern blots. The correlation coefficient was calculated using the Excel function. Numbers on top indicate the probed genes: 1, act15 (DDB0185015); 2, sevA (DDB0188380); 3, arcB (DDB0204976); 4, coaA (DDB0192207); 5, talA (DDB0219577); 6, eIF6 (DDB0203843); 7, gadB (DDB0188068); 8, aclyB (DDB0205386); 9, (DDB0219898); 10, rabR (DDB0168590); 11, nramp1 (DDB0202615).

Figure 4

Selection of enriched GO terms in cluster 1/5/7 (genes regulated by phagocytic stimuli). The GOAT programme [33] was used to calculate the enrichment (abscissa) and statistical significance (P-value) of every GO term (GO-level in the ordinate and GO-annotation on the right). The observed number of genes in a specific category (List) and the number of genes that might appear in the same category in case of random selection from the same reference list (Total) are shown. Due to redundancy in GO-terms, only a selection of enriched terms is shown.

Figure 5

Selections of enriched GO terms in cluster 4 and cluster 2/6. In cluster 4 (genes regulated by phagocytosis and growth on bacteria) and cluster 2/6 (genes regulated by growth on bacteria) the GO terms resulted enriched only for down-regulated genes. See legend to Figure 4 for experimental conditions and explanations.

Figure 6

Manual annotation of categories for cluster 1/5/7 (genes regulated by phagocytic stimuli). The percentage of manually annotated genes was calculated for up- or down-regulated genes (A and B, respectively). The annotated genes were categorized and the percentage of genes in each category was calculated, taking as 100% the total number of annotated genes in the cluster. Using the total number of genes as reference shows more clearly the differences in percentage within up- and down-regulated categories.

Figure 7

Manual annotation of categories for cluster 4 (genes regulated by phagocytosis and growth on bacteria). See legend to Figure 6.

Figure 8

Manual annotation of categories for cluster 2/6 (genes regulated by growth on bacteria). See legend to Figure 6.

Figure 9

Regulation of glutamate metabolic pathway by phagocytic stimuli inferred by the gene expression profile. Twenty genes involved in aminoacid metabolism are detected in the DNA microarray. Of these, 14 are directly or indirectly linked to glutamate metabolism, as shown in the Figure. Over- and under-expressed genes are shown in yellow and grey, respectively. GadA (DDB0206436) and gadB (DDB0188068), encode the glutamate decarboxylase A and B [51], respectively and are strongly over-expressed. Up-regulated are also a putative proline dehydrogenase (prodhA, DDB0167252), which catalyzes the formation of L-glutamic acid γ-semialdehyde, the histidine ammonia lyase (hisl, DDB0187742) and the glutamic acid formiiminotransferase (ftcdA, DDB0187716), which are both involved in the major pathway for converting histidine to glutamic acid. Up-regulated is also the gene for the 10-formyltetrahydrofolate synthetase (fths, DDB0188868), which generates a potential substrate for the formiimino group released by the formiiminotransferase. Slightly up-regulated is also the gene for the glutamate dehydrogenase (gdhA, DDB0187484), which leads to α-ketoglutarate formation from L-glutamic acid. Downregulated are seven genes encoding putative enzymes involved in the conversion of L-glutamic acid to, respectively, L-ornithine (argC, DDB0186887, argD, DDB0190334 and argE, DDB0189404), serine (serA, DDB0203991 and serC, DDB0190692) or lysine (aass, DDB0218638 and sdh, DDB0186419). In the assumption that the regulation of all these genes is reflected at enzymatic level, a preferential accumulation of glutamate and its conversion to GABA and α-ketoglutarate could be inferred.

Figure 10

Regulation of genes encoding actin cytoskeleton proteins by phagocytosis and growth on bacteria. The figure shows the protein products of genes for actin cytoskeletal proteins and their small G protein regulators, which are detected in the microarray, with their interactions as known from literature. Over- and under-expressed genes are shown in yellow or grey, respectively. In italics are indicated genes, which are detected in the comparison V12M2 cells vs. AX2 growing on bacteria. The vast majority of these genes is down-regulated. See text for comments. For the DDB ID numbers of the genes see Additional Files 1, 2, 3.

Figure 11

Simplified model of regulation of differential gene expression in phagocytosis and growth on bacteria. The model depicts some differentially regulated genes in phagocytosis and growth on bacteria, with emphasis on genes encoding putative membrane receptors, signal transducers, transcription factors and selected metabolic pathways. Over- or under-expressed genes are indicated in yellow and grey, respectively. Genes detected in the comparison V12M2 vs. AX2 growing on bacteria are indicated in italics. Phagocytosis leads to overexpression of putative membrane receptors, tetraspanins and metabotropic receptors. Receptor clustering due to bacterial binding, and to the scaffold activity of tetraspanin, may generate signalling complexes containing also GPCR's. Activation of the heterotrimeric G protein [5], putatively via autocrine signals or bacterial metabolites binding to GPCR's, will regulate actin reshaping in the phagocytic cup, favouring phagocytosis. We further suggest that the same signal complex at the membrane of the phagocytic cup or the engulfed phagosome may generate signals regulating gene expression, which can be mediated by a TIR-domain containing protein acting on transcription factors. Up-regulation of genes involved in sterol metabolism leads to increased production of sterols, which we suggest to be preferentially incorporated in phagocytic cups and phagosomal membrane. GABA may accumulate as catabolic product of glutamate, due to the high activity of GadA and GadB glutamate decarboxylase. Whether it is released and acts as autocrine signal is open. Regulation of phosphoinositides may regulate phagosomal maturation and fusion with lysosomes. See Additional Files 1, 2, 3 for DDB ID numbers of genes encoding the protein products indicated. The DDB ID numbers in the lower box, without protein product indication, identify genes, which are strongly stimulated by phagocytosis and encode hypothetical membrane proteins with no identifiable domains.