|Home | About | Journals | Submit | Contact Us | Français|
Conceived and designed the experiments: MCC KS DSS. Performed the experiments: MCC KS. Analyzed the data: MCC KS DSS. Contributed reagents/materials/analysis tools: MCC KS DSS. Wrote the paper: MCC KS DSS.
Immunity and metabolism are intimately linked; manipulating metabolism, either through diet or genetics, has the power to alter survival during infection. However, despite metabolism's powerful ability to alter the course of infections, little is known about what being “sick” means metabolically. Here we describe the metabolic changes occurring in a model system when Listeria monocytogenes causes a lethal infection in Drosophila melanogaster. L. monocytogenes infection alters energy metabolism; the flies gradually lose both of their energy stores, triglycerides and glycogen, and show decreases in both intermediate metabolites and enzyme message for the two main energy pathways, beta-oxidation and glycolysis. L. monocytogenes infection also causes enzymatic reduction in the anti-oxidant uric acid, and knocking out the enzyme uric oxidase has a complicated effect on immunity. Free amino acid levels also change during infection, including a drop in tyrosine levels which may be due to robust L. monocytogenes induced melanization.
We need to be careful with the way we draw lines around biological systems. For example, survival during infection is affected by most physiologies in the body and thus it seems dangerous to gather one group of proteins that are responsible for killing microbes and say “these define immunology.” It matters little if we can explain “the immune system” but don't understand all that helps us survive infections. Drosophila are a useful innate immunity model, but are also useful for studying microbial pathogenesis to find these “non-immune” factors. Metabolism is an example of such a system that influences immunity and, in Drosophila, the only explanations we have for how microbes cause disease involve changes in energy stores or gut function –. Though we don't have a broad understanding of what defines a “sick” metabolic state is during an infection several experiments in Drosophila focus on select groups of metabolites.
Research in Drosophila melanogaster has repeatedly ascribed metabolism a role in immunity. Mycobacterium marinum is a model pathogen that causes “consumption” in flies; infected flies lose all of their energy stores and waste to death. Reduction of FoxO, a negative regulatory transcription factor in the insulin signaling pathway, slows the infection induced wasting and increases the mean time to death . FoxO is also important in modulating the transcript levels of some immune effectors, anti-microbial peptides (AMPs). In the absence of an infection, starvation causes a FoxO-dependent upregulation of AMPs, effective even in the absence of Toll and Imd, the most notable regulators of AMP transcription , . Additionally, FoxO binds directly to the promoter of Drosomycin, suggesting that these are direct effects of the protein. Taken together, these data support an intimate connection between insulin and immune signaling.
The relationship between metabolism and immunity is complex and the results depend upon the pathogen tested and measured immune output. Ayres and colleagues found that Listeria monocytogenes and Salmonella typhimurium cause infection-induced anorexia. In return, diet restriction alters a fly's tolerance and resistance to infection in a pathogen specific manner, making it worse at resisting L. monocytogenes but better able to tolerate S. typhimurium . Drosophila larvae fed Pseudomonas entomophila have a melanized proventriculus and consume less food, which is potentially one cause of death from infection , . Pathology also may occur because P. entomophila causes gut damage using a pore-forming protein, monalysin .
Work in other insects supports the link between diet composition and immunity; in caterpillars of the African spotted leaf worm (Spodoptera littoralis), Cotter and colleagues showed that the protein and carbohydrate content of food impacted the animals' immune responses in a pathway dependent manner and that there were potential trade-offs and no “perfect” immune diet . Adamo and colleagues showed that infected crickets (Gryllus texensis) were less resistant to infection when fed high fat diets and would self-select lower fat diets when infected and given a choice . These studies not only show that a beneficial diet is infection dependent but that animals possess the ability to distinguish and choose more beneficial diets when infected.
Drosophila C Virus infection of the Drosophila S2 cell line revealed a specific role for a single metabolic pathway: fatty acid biosynthesis . RNAi knockdown of components of the synthetic pathway inhibited viral replication. More recently this has been extended into mammalian systems showing that regulation of fatty acid biosynthesis by AMPK is important for a number of viruses . This does not appear to be solely a viral pathogenesis phenomenon as branched chain fatty acids promote intracellular growth of Listeria monocytogenes in a mammalian cell line . The ability to obtain energy is important to both host and bacterium and changes in metabolism and nutrition likely impacts both. While each of these papers highlights a narrow aspect of the metabolism-immunity relationship, a more global approach to metabolite changes has the benefit of revealing system-wide changes and unanticipated roles for metabolites.
Here, we take a broad look at the changes in metabolism seen during L. monocytogenes infection of Drosophila. We chose L. monocytogenes as a pathogen because it causes infection-induced anorexia and a previously published microarray analysis from the lab showed infection-induced transcriptional changes in a number of metabolic pathways , . By combining both a metabolic focused analysis of our microarray to look at metabolism transcripts and broad spectrum chemistry approach to look at the metabolites themselves, a picture emerges of a shifting energy landscape in infected fruit flies.
L. monocytogenes infection reduces metabolite concentrations in the two primary energy generation pathways: beta-oxidation and glycolysis. The messages encoding enzymes in these pathways are also down regulated. As the infection proceeds, triglyceride and glycogen stores fall. In addition, two shifts in metabolites were related to reactive oxygen species (ROS) generation. First, there was enzymatically driven decrease in the anti-oxidant, uric acid, which has a complicated effect on immunity. Second, there was a decrease in tyrosine, an important precursor for L-3,4-dihydroxyphenylalanine (L-DOPA), the substrate for phenol oxidase, which generates microbe killing ROS and helps resolve infections.
We measured changes in metabolite concentrations at 0, 6, 24, and 48 hours post-infection and found that 169 metabolites, out of the 221 identified, were significantly different during at least one time-point after infection (Figure 1 and Supplementary Figures 1–5). These metabolites fell into the following broad categories: amino acid metabolism (Supplementary Figure 1), carbohydrate metabolism (Supplementary Figure 2), lipid metabolism (Supplementary Figure 3), nucleotide metabolism (Supplementary Figure 4), and additional miscellaneous categories (Supplementary Figure 5).
Past work from our group demonstrated that flies infected with M. marinum suffered from a wasting disease that reduced the two main energy stores, fat (triglycerides) and glycogen . Transcript levels data from our L. monocytogenes microarray suggested that transcripts for enzymes in the beta-oxidation and glycolysis pathways were similarly down regulated, indicating that energy metabolism might be altered during this infection as well.
We first examined how fat metabolism shifted during L. monocytogenes infection. Triglycerides are composed of three fatty acids joined by ester bonds connecting their carboxyl groups to the hydroxyl groups of glycerol. Triglycerides are broken down by lipases which release free fatty acids and these are then transported into the mitochondria where they are metabolized through beta-oxidation. The basal levels of free fatty acids in uninfected flies varied depending on both length of the carbon chain and saturation state (Figure 2A). Upon infection, all free fatty acids with twelve or more carbons decreased regardless of initial levels or saturation state (Figure 2B). We confirmed the L. monocytogenes induced decrease in fatty acids by specifically isolating free fatty acids and running in-house gas chromatography–mass spectrometry (GC-MS) (Figure 2C). We also confirmed the reduction in message for three genes (CG8732 – long chain fatty acid-CoA ligase, CG4600 – acetyl-CoA acyltransferase, CG3902 – short-branched-chain-acyl-CoA dehydrogenase) implicated by our previously published microarray by qRT-PCR (Figure 3A). Seven out of the 17 transcripts assayed by microarray decreased, although this does not include the rate limiting enzyme acyl-carnitine transferase (CPT-1: CG12891). The changes in beta-oxidation are summarized in Figure 3B.
Glycogen and sugar metabolism during L. monocytogenes infection presents a slightly more complicated picture. Glycogen is broken down by glycogen phosphorylase into Glucose-1-P, which is a substrate for glycolysis. Glucose and other more complex sugars can also enter into glycolysis. During infection with L. monocytogenes, the sugars can be broken into three groups, those with no significant change, those with an intermittent shift and those that drop and stay low. The primary circulatory sugar in flies, trehalose, drops for the first days of infection but rises back to baseline by 48 hours post-infection (Figure 4B). Glucose, fructose, sucrose and maltotetraose, a complex sugar composed of four glucose molecules, all remain low during infection (Figure 4B). The dimers of glucose, maltose and trehalose, along with maltotriose, a glucose trimer, decrease intermittently during infection (Figure 4B). The compounds that do not significantly shift are the longer glucose complexes, maltopentaose and maltohexaose, which are present at relatively low levels in the fly and show higher variation (Figure 4B). Eleven out of thirty eight transcripts assayed by microarray decreased; and although the rate limiting enzyme for glycolysis phopho-fructokinase (Pfk: CG4001) does not decrease significantly, the rate limiting enzyme for gluconeogenesis, fructose-1,6-bisphosphatase (fbp: CG31692), does. Figure 5 summarizes the transcriptional and metabolite effects of L. monocytogenes on glycolysis.
To assess the long-term effect these rapid changes in beta-oxidation and glycolysis had on the energy storage in infected flies, we measured triglyceride and glycogen levels during infection. We found that flies infected with L. monocytogenes have significantly reduced levels of both triglycerides (Figure 6A) and glycogen (Figure 6B). This indicates an overall decrease in energy stores.
Uric acid is a potent anti-oxidant that is present in human blood at near saturating concentrations and is considered to provide over half of the blood's anti-oxidant power . In the fly, uric acid is produced as a waste product of purine elimination, and accumulates in malpighian tubules, filling them with uric acid crystals, which are then dumped into the hind gut where they are excreted in the frass . Flies encode the enzyme uricase, which converts uric acid into the more soluble allantoin. This reaction is highly conserved in the animal kingdom although it is notably missing in higher apes (including humans) and Dalmatians , . Uric acid can also be converted to allantoin via a reactive oxygen driven chemical reaction, which can happen in all animals. This is why, in humans, allantoin is a biomarker for reactive oxygen (ROS) stress as it can only be made as the product of reactive oxygen species and uric acid .
During L. monocytogenes infection of D. melanogaster the levels of uric acid decrease and allantoin its metabolic product increase (Figure 7A). Why does this happen? Is it the result of enzymatic breakdown or an indication that ROS is being produced? Uricase is upregulated 3-fold during L. monocytogenes infection supporting a role for the enzymatic pathway (Chambers MC, 2012). To test the role of uricase in modulating uric acid levels we looked at the effect of infection on uric acid levels in a uricase mutant. We expected that if the uric acid decrease was primarily driven by the ROS and uric acid reaction then the mutant would still have an infection induced decrease in uric acid. However, a uricase mutant, which has a 10-fold knockdown of uricase message (Figure 7B), had no change in uric acid levels during infection. This supports the idea that uricase is the primary effector of the infection-induced decreases in uric acid levels.
We next wanted to assay how uric acid regulation would affect the host during infection. Why would the host alter a potential antioxidant when it was going to produce ROS? The immune implications of uricase regulation are complicated. Uricase mutants do not have a consistent significant change in survival or bacterial load during L. monocytogenes, S. typhimurium, and Burkholderia cepacia infection (data not shown). Uricase mutants are, however, more resistant to Francisella novicida (Figure 8A,C) and Enterococcus faecalis (Figure 8B,D) infection. The uricase mutants also better survive wounding as seen in Figure 9. Supplementary Figure 2 shows the summary of all media controls and shows a significant increase in median time to death at both 29°C and 25°C (p<0.0001 and p=0.0381 respectively). The lifespan of unmanipulated flies, however, is unchanged relative to the parental line (data not shown).
Melanization is an important part of the insect immune response and is required for resistance to both L. monocytogenes and S. typhimurium which both cause robust disseminated melanization in the fly. The primary enzyme responsible for melanization, phenoloxidase (PO) uses L-3,4-dihydroxyphenylalanine (L-DOPA) and this is synthesized by a pathway which converts phenylalanine to tyrosine to L-DOPA. We find that both phenylalanine and tyrosine are depleted through infection and that L-DOPA levels are also down at 6 hours post-infection (Figure 10, Supplementary Figure 1). This is in contrast to many other amino acids, which remain constant or even increase during infection.
Immunity is tightly linked to metabolism, and this paper discusses a range of connections, from anti-oxidants to energy stores, at both the transcript and metabolite level. The changes are likely a product of integrating host energy needs, changes in feeding during infection, energy requirements of the bacteria, and virulence driven changes in host biology.
Most prominent in our data was the impact that L. monocytogenes infection has on fruit fly energy metabolism. Similar to what was seen in M. marinum infected flies, these L. monocytogenes infected flies also suffered a consumption-like disease with decreases in both triglyceride and glycogen levels . Preceding these decreases in energy stores, the flies showed a marked decrease in the immediate substrates for mitochondrial beta-oxidation and glycolysis. Concomitantly, the flies also had decreased investment in the transcripts for both of these pathways. The entire energy landscape of these flies appears to be in flux. There are many potential functional causes and consequences of these changes. These flies are diet restricted and potentially increasing energy expenditure to mount an immune response; these are both host factors which could drain energy reserves. On the bacterial side, L. monocytogenes is replicating during infection potentially consuming resources as well. One could imagine the host shifting its metabolism to reduce available nutrient for the bacteria, further impacting these pathways. There is likely a complicated interplay between all of these factors, and manipulating any of these components has the potential to impact the survival of the host. It could be a multi-faceted trade-off between host nutrition, bacterial nutrition and investment in immunity.
Another potential trade-off is the enzymatic reduction in the anti-oxidant uric acid and this also has a compelling potential host-bacteria interaction story. Theoretically, keeping the level of an anti-oxidant high during infection could impede the effectiveness of the ROS immune effector response because it could protect the infecting bacteria as well as the host. Antioxidants would thus allow the bacteria to grow to higher levels, but also improve tolerance to infection. However, we found that stabilization of uric acid levels had no effect during a number of infections and the flies better resisted the bacteria during F. novicida and E. faecalis infection. This is the opposite of what we expected. We have two potential explanations.
First, changes in uric acid metabolism could lead to compensating changes in the levels of other anti-oxidants. Second, uric acid could have an additional role in immune or metabolic signaling that is yet unappreciated. This story is further complicated by effect of wounding on uricase mutants; loss of uricase increases survival during wounding and this makes it difficult to deconstruct the immune phenotype.
During infection, we also saw that amino acid precursors for melanization, a core immune response against L. monocytogenes decreased. This decrease could be due to the robust L. monocytogenes induced melanization response; however, bacteria can import amino acids through permeases and transporters, and shifts in the free amino acid composition could also be due to bacterial use.
These metabolic data provide a stepping stone from which we can launch additional and more targeted studies on metabolism. Studies of metabolic-immune interactions could potentiate development of cheaper treatments for infection for if we can modulate our resistance and tolerance through diet shift instead of expensive pharmaceuticals, we have the potential to bring treatment to a much larger swath of the global community.
The wild-type parental stain used in all experiments is white1118 (Bloomington 6326). Uricase experiments were done using a piggybac mutant (Bloomington 18814).
Listeria monocytogenes (strain 10403S) cultures were grown in 4 ml brain heart infusion (BHI) broth at 37°C without shaking after inoculation from L. monocytogenes grown overnight on a Luria Bertani (LB) agar plate. L. monocytogenes stocks were stored at −80°C in BHI broth containing 15% glycerol.
Francisella novicida (strain U1112) cultures were grown in 4 ml tryptic soy broth (TSB) supplemented with 0.2% L-cysteine at 37°C on a rotating wheel after inoculation from F. novicida stocks stored at −80°C in TSB broth containing 15% glycerol.
Enterococcus faecalis (strain V583) cultures were grown in 4 ml LB broth at 37°C with shaking after inoculation from E. faecalis grown overnight on a LB agar plate. E. faecalis stocks were stored at −80°C in BHI broth containing 15% glycerol.
For infection, 50 nL of the bacterial cultures were injected at the following optical densities (OD600): L. monocytogenes, 0.01 (approx. 1,000 CFU/fly); F. novicida, 0.001 (approx. 1,000 CFU/fly); E. faecalis, 0.6 (approx. 5,000 CFU/fly).
Five to seven day post-eclosion male flies were used for injection. The flies were raised at 25°C, 65% humidity on yeasted dextrose food in a light cycling incubator (12 hours dark, 12 hours light). Flies were anesthetized with CO2. A picospritzer (Parker Hanninfin, http://www.parker.com) was used to inject 50 nL of liquid into each fly with pulled glass capillary needles that were individually calibrated by measuring the size of the expelled drop under oil. About 20 flies were placed per vial and then experiments were kept at 29°C, 65% humidity in a light cycling incubator.
Flies were injected with 50 nL of L. monocytogenes culture (about 1000 CFU/fly) or left unmanipulated. Following injection, the flies were placed in dextrose vials and incubated at 29°C. At 0, 24, 48 and 72 hours post-injection triplicates of 100 flies were flash frozen and then homogenized in methanol. These preparations were then sent to Metabolon (http://www.metabolon.com), which performed a combination of Gas and Liquid Chromotography techniques combined with Mass Spectrometry (GC/LC-MS). A table of metabolites that changed in our samples was returned both with the raw normalized area for each metabolic peak and with calculated amounts relative to zero hour w1118 flies. Statistical significant differences were determined using Welch's two tailed t-test.
Isolation and quantification protocols closely followed Nandakumar M et al. . Briefly, 10 flies were homogenized in 300 uL of 3% H2SO4. The solution and accompanying fly debris were then transferred to glass tubes and the total volume brought to 4 mL (with 3% H2SO4). Tubes were capped and incubated at 80°C for two hours. After allowing samples to cool, 1.5 mL water and 750 uL of hexane plus standard (I-Br decane) were added. Each sample was vortexed and then spun at 550 rpm for about one minute without brake. Take 200–400 uL of the top layer and apply to GC-MS.
Flies were injected with 50 nL of L. monocytogenes OD600=0.01 (about 1000 CFU/fly) or kept unmanipulated. Following injection, the flies were placed in dextrose vials and incubated at 29°C for six hours. Groups of 12 flies were homogenized in TriZOL and stored at −80°C until processed. RNA was isolated using a standard TriZOL preparation, and the samples were treated with DNase (Promega, http://www.promega.com). Quantitative RT-PCR was performed as described previously by Schneider et al. using a Bio-Rad icycler and the following primer sets: CG8732, CG4600, CG3902, and uricase (Supplementary Table 1) . Briefly, relative levels of mRNA were determined using a standard curve and then normalized to 15a expression levels. Samples were then normalized to the unmanipulated w1118 expression level to yield fold induction.
Quantification of triglyceride and glycogen levels was assayed using kits from BioVision. The protocol provided by the manufacturer was followed. Samples were prepared by homogenizing 5 flies in 125 uL of the assay buffer Tris-EDTA (TE; 10 mM Tris, 1 mM EDTA, pH 8) with 0.1% Triton X-100. All samples were normalized to total protein content, values are represented as a percent relative to unmanipulated flies.
Quantification of uric acid levels was done using a kit from BioVision. The protocol provided by the manufacturer was followed. Samples were prepared by homogenizing 5 flies in 200 uL of the assay buffer.
Mutant flies and the parental control were injected with 50 nL of the bacterial culture or medium at the following optical densities (OD600): L. monocytogenes, 0.01 (approx. 1,000 CFU/fly); F. novicida, 0.001 (approx. 1,000 CFU/fly); E. faecalis, 0.6 (approx. 5,000 CFU/fly).. About sixty flies were assayed for each condition and placed in three vials of 20 flies each. Death was recorded daily. Survival curves are plotted as Kaplan-Meier plots and statistical significance is tested using log-rank analysis using Prism software (http://www.prism-software.com). All experiments were performed at least three times and yielded similar results.
Colony forming units (CFUs) were determined using both spot-plating and an autoplate spiral plater (Spiral Biotech http://www.aicompanies.com). For spot-plating, eight individual flies were collected at each time point. These flies were homogenized, diluted serially and plated onto the appropriate media (LB agar for L. monocytogenes/E. faecalis and Mueller-Hinton agar supplemented with ferric pyrophosphate (0.05%) and L-cysteine (0.1%) for F. novicida) and grown overnight at 37°C. Some L. monocytogenes experiments were completed using the Spiral Biotech plater and for these six individual flies were homogenized and diluted. 50 µL of liquid was plated exponentially on a LB plate, grown overnight at 37°C and then counted using QCount, which calculates the original number of CFU per fly. For statistical analysis, if CFU/fly did not approximate a Gaussian distribution we analyzed the log(10) transform of the data. CFU experiments were assessed for sources of variation using a two-way ANOVA and followed with Bonferroni post-tests for specific comparisons of interest.
Amino Acid Pathway Metabolites. Metabolites assayed by a combination of GC/LC-MS. All values are presented as a fold change relative to uninfected flies. Significantly increased metabolites, as determined by a Welch's two-tailed T-Test, are in red cells, and significantly decreased metabolites are in blue cells.
Carbohydrate Pathway Metabolites. Metabolites assayed by a combination of GC/LC-MS. All values are presented as a fold change relative to uninfected flies. Significantly increased metabolites, as determined by a Welch's two-tailed T-Test, are in red cells, and significantly decreased metabolites are in blue cells.
Lipid Pathway Metabolites. Metabolites assayed by a combination of GC/LC-MS. All values are presented as a fold change relative to uninfected flies. Significantly increased metabolites, as determined by a Welch's two-tailed T-Test, are in red cells, and significantly decreased metabolites are in blue cells.
Nucleotide Pathway Metabolites. Metabolites assayed by a combination of GC/LC-MS. All values are presented as a fold change relative to uninfected flies. Significantly increased metabolites, as determined by a Welch's two-tailed T-Test, are in red cells, and significantly decreased metabolites are in blue cells.
Miscellaneous Pathway Metabolites. Metabolites assayed by a combination of GC/LC-MS. All values are presented as a fold change relative to uninfected flies. Significantly increased metabolites, as determined by a Welch's two-tailed T-Test, are in red cells, and significantly decreased metabolites are in blue cells.
qRT-PCR Primers. Sequences of both the forward and reverse primers used for the qRT-PCR experiments.
We thank Mary Beth Mudgett, David Relman, Denise Monack and all the members of the Schneider lab for thoughtful discussion. We also greatly appreciated the Stanford Mass Spectrometry facility's help with refining the fatty acid GC-MS assay.