|Home | About | Journals | Submit | Contact Us | Français|
Aspergillus fumigatus is a saprophytic fungus commonly found in soil and compost piles. In immunocompromised patients it takes on a sinister form as a potentially lethal opportunistic human pathogen. We currently have a limited understanding of the in vivo growth mechanisms used by A. fumigatus during invasive pulmonary aspergillosis (IPA). The ability of A. fumigatus to adapt to various microenvironments encountered during growth in the human host may explain why A. fumigatus is the most frequently occurring opportunistic pathogenic mold. The transcriptional and metabolic responses to changing microenvironments found in the mammalian lung require the activation of pathways implicated in resistance to unique stresses. Thus, the production of primary metabolites in vivo may give clues to the critical pathways used by A. fumigatus to cause disease in human hosts. We recently have identified primary metabolites in the mammalian lung typically associated with fungal growth under hypoxic environments suggesting that A. fumigatus may encounter low oxygen tensions during IPA. These and other studies on A. fumigatus metabolism are the focus of this review.
Aspergillus fumigatus is a saprophytic, asexual reproducing fungus commonly found in soil and compost piles. Its primary ecological function is to recycle carbon and nitrogen through the environment [1-3]. However with the rising number of immunocompromised patients, A. fumigatus has become a leading cause of mortality in this patient population [4, 5]. While A. fumigatus is responsible for a number of clinically relevant diseases, invasive pulmonary aspergillosis (IPA) is the most lethal with mortality rates ranging from 60-90% [4, 6]. A. fumigatus is the most prevalent causal organism of IPA, but we still do not understand how the fungus survives and thrives in the harsh microenvironments of the mammalian lung.
To cause disease, A. fumigatus must face and overcome a number of in vivo challenges once it is inhaled in the mammalian host (Fig. 1). It has been proposed that A. fumigatus possess unique virulence factors, which allow it to cause disease, but arguably no unique A. fumigatus virulence factor has been discovered. It seems most likely then, that A. fumigatus possess robust adaptive attributes that allow it to thrive in immunocompromised patients [2, 7]. From an evolutionary perspective, these attributes are not true virulence factors (i.e. factors that have arisen from selective pressures to overcome and evade a host defense response); however, they allow A. fumigatus to cause disease in mammalian hosts and likely differentiate it from other saprophytic moulds not associated with invasive mycoses.
Mammalian organisms present a broad variety of microenvironments in which A. fumigatus must survive to cause disease. These microenvironments, from the initial engulfment of conidia by alveolar macrophages and neutrophils to the hyphal invasion of alveoli in the lung, place unique stresses on the fungal organism. How does the fungus obtain the necessary nutrients, and what are these nutrient sources, to survive in vivo? How does the fungus overcome the low oxygen tensions typically found at sites of inflammation and infection to persist and cause disease? The focus of this short review is on recent findings that are beginning to elucidate the in vivo requirements for fungal growth in mammalian hosts. We conclude the review with a brief discussion of a recent finding by our laboratory on the production of primary metabolites associated with fungal growth in hypoxic environments in a mouse model of IPA.
During invasive growth, A. fumigatus has to deal with unique microenvironments found in the mammalian lung, and further, these environments can rapidly change depending on the current stage of the infection. Thus, the metabolic and developmental response to changing environments in vivo requires the transcription of specific genes and the subsequent translation of the mRNA into proteins. For example, amino acids are required for translation and necessary for fungal growth and development. Fungi get amino acids from their environment either by transport processes, from precursors, which are derivatives of carbon and nitrogen primary metabolism, or from degradation of proteins which are no longer required under specific conditions. All three processes have to be coordinated and are carefully regulated in fungi (reviewed in Braus et al. ).
In numerous yeasts as well as filamentous fungi, amino acid starvation activates a complex genetic network, the ‘cross-pathway control’ (CPC) system (also known as General Control [GC] of amino acid biosynthesis) [9-11]. This conserved system consists of a sensor kinase (eIF2α kinase CpcC/Gcn2p) and a transcriptional activator protein (CpcA/Gcn4p). The sensor kinase detects the accumulation of uncharged tRNA molecules, which is a sign for amino acid depletion. During amino acid starvation, the sensor kinase is activated and leads to the down-regulation of general translation. At the same time, the expression of the transcriptional activator protein (CpcA/Gcn4p) is not repressed . This activator induces the expression of many genes, for example, genes involved in the biosynthesis of amino acids, purines, pyrimidines, and vitamins, as well as genes involved in peroxisomes, amino acid transport, and mitochondrial carriers .
In A. nidulans the activation of the cross-pathway system results in impaired sexual fruit body formation , whereas in the human pathogenic dimorphic yeast Candida albicans Gcn4p affects filament formation . The influence of this system on fungal pathogenicity has been studied in A. fumigatus. Krappmann et al. (2004)  observed that A. fumigatus requires the transcription factor CpcA during periods of amino acid starvation and for full disease development. These results suggest that A. fumigatus needs the cross-pathway control system to survive in specific microenvironments within the mammalian host. In addition, this result suggests that not all nutrients can be obtained from the host by the fungus during pathogenesis.
During pathogenesis, it appears that the biosynthesis of uracil/uridine and lysine is required by the fungus for in vivo growth. In A. fumigatus, two auxotrophic markers based on the biosynthesis of uracil/uridine and lysine are well characterized and used in generating genetically modified strains of the fungus. The strain lacking the pyrG gene encoding orotidine-5′-phosphate decarboxylase (and therefore auxotrophic for uridine or uracil)  and the strain lacking the lysF gene encoding for homoaconitase (and therefore auxotrophic for lysine) , however, are not capable of establishing disease in experimental murine models of IPA. The pyrG mutant conidia remained ungerminated in alveolar macrophages ; whereas conidia from the lysF mutant strains were able to persist and germinate likely due to lysine stored in conidia . However, lysine appears to be limited in the lung, because the ΔlysF strain was not able establish invasive hyphae and cause disease . In both cases, full virulence could be restored by complementing the mutant gene or by supplementing the drinking water of the animals with the missing amino acid. These findings demonstrate the importance for the fungus to synthesize amino acids on its own with specific metabolic pathways or by uptake from the environment. However, as shown for lysine and uracil, not all amino acids are readily available in the mammalian host during infection.
It is also likely that A. fumigatus obtains important nutrients in vivo from destruction of host tissue. There are hints that A. fumigatus metabolizes proteins in vivo to satisfy the need for nutrients during infection . The genome of A. fumigatus contains several genes most likely coding for proteases  and it has been shown that A. fumigatus produces several proteases during lung infections . However, different proteases may play unique or overlapping roles during pathogenesis, and thus, this is a difficult hypothesis to obtain evidence for with traditional gene replacement technologies. Still, the advantage of in vivo protein degradation to the fungus is clear and can result in the allocation of all building blocks necessary for energy and biomass production.
Evidence that the degradation of amino acids is important in A. fumigatus pathogenesis was recently obtained from generation of a gene replacement mutant in a key enzyme in the methylcitrate cycle, methylcitrate synthase. Methylcitrate synthase is essential for the degradation of propionyl-CoA, which is a degradation product of valine, methionine and isoleucine [22-24] in fungi and some bacteria. Ibrahim-Granet et al. (2008)  observed that methylcitrate synthase has a strong impact on the pathogenicity of A. fumigatus which supports their hypothesis that proteins are indeed metabolized by the fungus during infection.
However, much less is known about potential carbon sources in vivo for the fungus. Based on in vitro data, the production of gliotoxin in vivo suggests access to simple, fermentable sugars such as glucose that stimulate production of this toxin. An additional potential source of carbon is via the glyoxylate cycle, a metabolic pathway occurring in plants and several microorganisms but not mammals. This pathway allows organisms to use fats for the synthesis of carbohydrates, which most vertebrates, including humans, cannot perform. Studies on the biological function of the glyoxylate cycle correlated its activity with the virulence of the bacterium Mycobacterium tuberculosis , the plant-pathogenic fungi Leptosphaeria maculans  and Magnaporthe grisea , and the human pathogenic yeast C. albicans [28, 29]. Results from these studies and others concluded that fatty acids and lipids account for the major carbon sources in infected tissue . It was therefore hypothesized that the glyoxylate cycle plays an important role during IPA.
Recent studies on A. fumigatus and A. nidulans have shown that the glyoxylate cycle is essential for growth when fatty acids or C2 compounds are the sole carbon source [31, 32]. In contrast to other pathogens, however, this pathway does not appear to be important for the development of IPA in a murine infection model as isocitrate lyase mutants are still capable of causing disease [33, 34]. This is similar to the finding that isocitrate lyase was not required for pathogenesis by the human pathogenic yeast Cryptococcus neoformans . These results imply that A. fumigatus has access to other carbon sources in vivo, and is not dependent upon fatty acids or C2 compounds to cause disease.
Metabolism, however, cannot proceed without energy, and thus the focus of our next discussion is on how A. fumigatus, an obligate aerobe, may generate energy in vivo during fungal pathogenesis to drive the metabolic pathways required for fungal growth and disease development.
During infection, A. fumigatus causes significant damage to host tissue, and when present, recruitment of effector cells to sites of inflammation (Fig. 2). While direct measurements have not been taken during A. fumigatus infections, sites of inflammation are known to contain significantly low levels of oxygen. Most eukaryotic cells obligatorily use oxygen to carry out many of their biochemical reactions. Oxygen is a key component of energy production where it functions as a terminal electron acceptor in the formation of ATP from glucose during aerobic respiration. Thus, during an infection, pathogens are often exposed to extremely low levels of oxygen, and oxygen levels can rapidly change depending upon the tissue infected and current status of the immune response.
Pathogens, therefore, must have alternative respiration pathways to generate energy in low oxygen environments found in vivo in mammalian hosts. In fact, respiratory flexibility, switching from aerobic respiration to various forms of anaerobic respiration has been implicated as an important virulence attribute in prokaryotic pathogens [36, 37]. For example, the prokaryote M. tuberculosis, Shi et al.  used transcriptional profiling and specific gene deletions to show that respiratory flexibility was an essential component of M. tuberculosis adaptation to the host immune response. How pathogenic fungi respond to low oxygen tensions found in vivo at sites of infection is largely unknown.
In the model organism Saccharomyces cerevisiae, a mechanism of adaptation to hypoxic conditions has been described. In aerobic conditions, heme biosynthesis activates the transcriptional regulator Hap1p , which induces genes involved in respiration and oxidative stress-responses. In addition, the transcriptional repressors Rox1p and Mot3p are activated and repress transcription of genes required for hypoxic adaptation . In hypoxic conditions, Rox1p and Mot3p are transcriptionally induced and this leads to transcription of genes involved in the adaptation to hypoxic conditions . To allow expression of hypoxic genes, transcription factors are required that utilize Rox1p-binding sequences, low oxygen-response elements (LORE) or other regulatory elements within promoters [42, 43]. Bioinformatic analyses of the A. fumigatus genome in our laboratory, however, have so far failed to identify a potential Rox1p homolog in this infectious mould. Given the facultative anaerobic life-style of S. cerevisiae and the obligate aerobic life-style of A. fumigatus, it is not surprising that mechanisms to deal with hypoxia are likely different in these fungi.
The human pathogenic yeast C. albicans also responds differently to hypoxic conditions than S. cerevisiae. Setiadi et. al  have determined the transcriptional response of to hypoxia and have shown that the C. albicans ScRox1 homolog, Rfg1, is not involved in the repression of hypoxic genes but has a function in filamentous growth. In addition, a homolog of ScHap1 could not be found in C. albicans. In this transcriptional screen, no global regulator could be identified, although Upc2, which was characterized as the C. albicans homolog of Sre1, a sterol regulatory element binding protein , was 1.6 fold upregulated .
The recent discovery of a putative oxygen sensor, Sre1, in the fission yeast Schizosaccharomyces pombe, however, has opened the door to discovering the potential role of hypoxia tolerance in eukaryotic fungal pathogenesis. In mammals, lipid homeostasis is regulated by a family of membrane-bound transcription factors designated sterol regulatory element-binding proteins (SREBPs). SREBPs directly activate the expression of more than 30 genes dedicated to the synthesis and uptake of cholesterol, fatty acids, triglycerides and phospholipids . Recently, a SREBP homolog in fungi, Sre1, was discovered in the yeast S. pombe . As in mammalian cells, Sre1 activates sterol biosynthesis enzymes in fungi, which require oxygen, but apparently also acts as an indirect sensor of oxygen levels by regulating the transcription of genes required for adaptation to hypoxic environments in this yeast.
In the human fungal pathogen C. neoformans, a Sre1 ortholog has been identified and characterized [48, 49]. A transcriptional profile of C. neoformans cells grown under normoxic conditions compared to cells grown under hypoxic conditions revealed that 347 transcripts showed changes in expression. Transcriptionally induced genes were involved in stress regulation, carbohydrate metabolism and respiration, and transcriptionally repressed genes included those involved in translation, vesicle trafficking and cell wall synthesis . The authors also discovered that Sre1 is required for hypoxic induction of genes encoding for oxygen-dependent enzymes involved in ergosterol synthesis. Importantly, mutants in the SREBP pathway displayed defects in their ability to proliferate in host tissues and cause disease in a murine model of cryptococcocis . Thus, the SREBP pathway has great potential to yield insights into how fungi adapt to and generate energy in hypoxic conditions in vivo.
The mechanism(s) of hypoxic adaptation in A. fumigatus are currently unknown. It seems likely that this saprophytic mould has evolved mechanisms to deal with low oxygen tensions commonly found in the soil and in compost piles. Preliminary transcriptional profiling experiments of A. fumigatus grown in hypoxic conditions has begun to reveal how this filamentous fungus adapts to hypoxia (Grahl et al. unpublished data). In these studies, we identified a putative Sre1 homolog in A. fumigatus, SrbA (SreA is already in use in A. nidulans for an unrelated protein). We hypothesize that this putative SREBP homolog plays an important role in adaptation to hypoxic microenvironments found in vivo during IPA. To address our hypothesis, we have generated an SrbA null mutant, which is no longer capable of growth in hypoxic conditions (Willger et al., unpublished data). Studies are underway to examine whether this SREBP signaling pathway is required for A. fumigatus pathogenesis, and how this signaling pathway may regulate genes involved in generating energy in hypoxic conditions.
Another potential mechanism for dealing with low oxygen levels is, of course, fermentation. It is known that the obligate aerobe A. nidulans can withstand significant periods of anaerobic stress by converting pyruvate to ethanol [50, 51]. For A. parasiticus it has been shown that the fungus produces ethanol in shake cultures in response to anoxic conditions . In addition, several other obligate aerobic filamentous fungi were also found to have the capability to survive prolonged periods of hypoxia presumably through switching their respiration to an alcohol fermentation pathway. Cultures of A. fumigatus grown on rich carbon sources such as glucose have also been reported to produce ethanol, indicating the presence of an alcohol fermentation pathway in this infectious mould [53, 54].
Using a murine model of invasive pulmonary aspergillosis we recently have found by-products of alcohol fermentation in the lungs of A. fumigatus infected mice (Cramer et al., unpublished data). This suggests that A. fumigatus has switched from aerobic respiration (or at least partially utilizes) to alcohol fermentation due to the hypoxic conditions found at the site of infection. Alcohol fermentation should allow the fungus to replenish sources of NAD+ and thus generate ATP through continued functioning of glycolysis.
To test our hypothesis that A. fumigatus utilizes alcohol fermentation in hypoxic conditions, we grew two wild-type strains under normoxic and hypoxic (1% O2) conditions on minimal medium with glucose, ethanol, or glycerol as the sole carbon source (Fig. 3). Our results suggest that both wild-type strains of A. fumigatus, AF293 and CEA10, grow well under hypoxic conditions with glucose as the sole carbon source (Fig. 3). Interestingly, CEA10 does not produce conidia under hypoxic conditions and seems to grow faster than strain AF293. Importantly, however, both strains grew less when glycerol, a non-fermentable carbon source, was the sole carbon source under hypoxic conditions. While growth was not completely inhibited, the significant reduction in growth suggests that A. fumigatus utilizes fermentation as one mechanism to adapt to hypoxic conditions. We currently are testing mutants in the putative alcohol fermentation pathway to further explore our hypotheses (Grahl et al., unpublished data).
Another oxygen-independent energy-producing metabolic pathway that supports anaerobic growth of yeasts and filamentous fungi is ammonia fermentation. This pathway has recently been found to be widely distributed among fungi . Ammonia fermentation consists of the dissimilatory reduction of nitrate to ammonium coupled with the catabolic oxidation of electron donors (ethanol) to acetate and substrate-level phosphorylation [55, 56]. One key reaction in the ethanol-oxidizing pathway is catalyzed by the CoA-acylating aldehyde dehydrogenase that is specifically produced under anaerobic conditions to generate acetyl-CoA, which is a key compound that mediates numerous biosynthetic and energy-yielding metabolic pathways as well as regulates several key metabolic reactions . During ammonia fermentation the acetyl-CoA synthetase catalyzes the conversion of acetyl-CoA to acetate. In contrast to other fungi, A. nidulans uses this reaction to produce ATP . It is unknown if the same is true for A. fumigatus, and whether this pathway is important for in vivo growth and virulence.
We are just beginning to understand the intricacies of A. fumigatus metabolism in vivo during infection, and it is clear that much remains to be learned about the requirements for in vivo fungal growth. However, it is very likely that mechanisms of in vivo growth will be uncovered that are unique to A. fumigatus. An understanding of the host microenvironment is important for developing new therapies for IPA. The host environment, for example, can play an important role in determining the efficacy of antifungal drugs. Recently it was shown that the MIC for Amphotericin B was reduced under hypoxic conditions, indicating the importance of understanding microenvironments found in vivo during IPA . Moreover, targeting unique microbial respiration pathways has already been observed to be an effective antimicrobial strategy. For example, the discovery of new anti-tubercular drugs that target respiratory components of M. tuberculosis and ATP synthesis suggest that uncovering similar mechanisms in fungal pathogens may also yield new antifungal drugs [58-60]. In addition, other adjunctive therapies such as the use of hyperbaric oxygen may also become more warranted as we uncover the mechanisms of in vivo fungal growth and metabolism.
RAC is currently supported by funding from the National Institutes of Health, COBRE grant RR020185, and the Montana State University Agricultural Experiment Station. The authors would like to thank Dr. Yohannes G. Asfaw, Duke University Medical Center, for assistance with histopathology, and Dr. John Perfect and Dr. Judy Rhodes for their constructive comments.