In vivo veritas: Aspergillus fumigatus proliferation and pathogenesis – conditionally speaking

Keywords:

• Aspergillus fumigatus
• antifungal drugs
• conditional gene expression
• doxycycline
• fungal virulence
• fungal metabolism
• invasive aspergillosis
• virulence factor

This article refers to:

Mutant characterization and in vivo conditional repression identify aromatic amino acid biosynthesis to be essential for Aspergillus fumigatus virulence

Perspective on Sasse and Colleagues¹

Aspergillus fumigatus and related species are dynamic and successful members of our planet's microbiome where they degrade and recycle organic material.² Aspergillus spp. high environmental abundance strongly speaks to their ability to thrive in diverse conditions and it has been suggested that their richness correlates with metabolic flexibility.³ This so called metabolic flexibility is a proverbial double-edge sword when it comes to Aspergillus spp and humankind. On the positive, Aspergillus spp. metabolism produces many molecules beneficial to human well being, the cholesterol lowering Statin drugs a prime example. Unfortunately A. fumigatus and related Aspergillus spp are also associated with life limiting and threatening human diseases, with invasive aspergillosis (IA) causing significant levels of human mortality.⁴ While therapeutic strides have been made against Aspergillus in the last decade, advances in medical treatment of many cancers and autoimmune diseases continue to increase the patient population at risk for life threatening Aspergillus infections.⁵ While epidemiological data on A. fumigatus infections remains enigmatic, recent estimates and clinical experience strongly suggest the incidence of disease is far greater than previously appreciated.^6-8 Perhaps most importantly, regardless of disease incidence, the recalcitrant nature of Aspergillus infections to medical intervention too often leads to poor treatment outcomes.^5,9

Many reasons explain the unacceptably high treatment failures with Aspergillus infection. For one, as a fellow eukaryote, survival and proliferation of A. fumigatus relies on fundamental pathways of nutrient acquisition and utilization similar to those used by humans.^10,11 Thus, host toxicity with current and developmental antifungal drugs is a major impediment to improved treatment efficacy. Moreover, current treatment approaches that often rely on use of triazoles are now globally threatened by the emergence of triazole drug resistance.^9,12-16 To circumvent these problems, essential gene identification in A. fumigatus has been an important research goal for many years.^17-21 However, in this regard, a second impediment to antifungal drug efficacy and development is that pathways and genes essential under laboratory conditions may not be active or essential in vivo during infection. Consequently, the antifungal drug arsenal used to treat A. fumigatus infections remains limited compared to other pathogenic microbes.^5,22-24

In the context of improving therapeutics, targeting virulence of the fungus is an important ongoing research endeavor. Much has been written, researched, and debated on the nature of virulence in A. fumigatus and related environmental fungi.^3,11,25,26 The question is more than semantics as targeting a classical virulence factor, if one exists, could significantly improve treatment outcomes. In this regard, a significant number of Aspergillus genes are associated with pathogenicity and virulence as identified primarily via Molecular Koch's Postulates approaches.^27-29 However, majorities of these genes are also found in closely related Aspergilli or related fungi that are infrequent agents of human disease questioning their role as classical virulence factors. In addition, comparative genomics approaches have yet to yield the proverbial smoking gun of A. fumigatus virulence. An important caveat with regard to current interpretations of comparative genomics studies is that many A. fumigatus specific genes have been found; yet their role in virulence has not been examined in depth (perhaps due to their “hypothetical” and “unknown” functions). An important research direction in our field should be to characterize these fungal and species-specific genes of unknown function.

What is clear from the above studies is that A. fumigatus virulence is frequently attenuated when its ability to proliferate in the mammalian respiratory tract is inhibited. Indeed, inhibition of fungal proliferation and growth is the basic mechanism of action of the triazole and echinocandin classes of antifungal drugs most often used to treat A. fumigatus infections. Accordingly, Aspergillus genes essential for proliferation are sought after as antifungal drug targets and several studies over the years have identified a significant number of these genes. However, examples of virulent A. fumigatus mutant strains (as examined in murine models) with severe in vitro growth defects strongly argues for the need to examine the impact of all genes/mutant strains in vivo under conditions as close to human patients as models allow.^30,31

Therefore, identification of fungal metabolic pathways that are viable therapeutic targets must be augmented with in vivo analyses to assess the significance of the pathway in the host microenvironment. Macro-and micronutrients available in vivo may reconstitute presumed auxotrophies or growth defects identified in vitro. In addition, in vivo micro-environmental conditions may induce gene expression profiles in the fungus that activate cryptic metabolic pathways capable of bypassing perceived essential pathways identified in vivo. In a recent issue of Virulence, Sasse and colleagues provide new insights into metabolic pathways required for in vivo proliferation of A. fumigatus and further develop the use of a conditional gene expression system in this important fungus to address these important questions of in vivo gene essentiality and virulence.¹

The question asked by Sasse and colleagues is whether A. fumigatus requires endogenous biosynthesis of aromatic amino acids (AAA) during mammalian infection.¹ AAA biosynthesis proceeds from the Shikimate pathway that is absent in mammals and ultimately results in production of tyrosine, phenylalanine, and tryptophan as end products. An attractive aspect of targeting this pathway for antifungal drug development is that the intermediate chorismic acid and end product tryptophan are important precursors involved in respiratory growth and maintenance of proper cellular oxidation-reduction potentials. As expected Sasse et al. observe that loss of a chorismate mutase-encoding gene (aroC) in A. fumigatus results in a tyrosine/phenylalanine auxotroph. Unexpectedly, the aroC null mutant displays germination and growth defects in liquid culture conditions in the presence of appropriate supplemented AAA. Though the mechanism is unclear, this result suggests detrimental consequences of chorismic acid accumulation on the fungal cell that may alter levels of ubiquinone and impact mitochondria function and output. A hint in support of this hypothesis is observed in an experiment testing the impact of the mitochondria inhibitor sodium azide on solid growth medium (where the aroC null mutant displayed an increased growth defect compared to the wild-type). In contrast to the aroC null mutant, loss of the anthranilate synthase component I encoding gene, trpA, results in a strain that grows only in the presence of supplemental tryptophan or anthranilate, regardless of culture conditions. Taken together, these in vitro results generated strains to test the hypothesis that A. fumigatus requires AAA biosynthesis for virulence.

Importantly, both the aroC and trpA null mutants were severely if not completely attenuated in virulence in a pulmonary leukopenic murine model of IA. Thus, in the respiratory tract and lung tissue, inhibition of fungal AAA biosynthesis at the chorismate branch point prevents lethal disease. In contrast, in a systemic model where fungal conidia are directly injected into the tail-vein the aroC and trpA null mutants retained a significant level of virulence compared to the parent strain as the majority of mice eventually succumbed to the infection. These data suggest that AAA auxotrophies may potentially be complemented in vivo in other organ compartments. Alternatively, differences in the host microenvironment may differentially activate the cross pathway control system important in responding to amino acid starvation.^32,33 Additional studies including histopathology of organs and fungal burden analyses beyond the kidneys and lungs in this model is warranted to fully validate these target genes as promising antifungal drug targets. However, these studies illustrate the critical importance of examining fungal nutritional mutants in vivo in relevant animal models of disease.

Due to the potential in vivo complementation observed in the aroC null mutant, the authors attempted to generate an aroC/trpA double null mutant. Perhaps in line with the in vitro germination and growth defect in the presence of supplements in the aroC mutant, a double aroC/trpA was not obtainable by the authors. One potential hypothesis to explain this promising result is that a complete blockage in the pathway at the chorismic acid branch point results in toxic levels of chorismic acid that inhibit mitochondria function or other essential pathways. The authors were also unable to recover a genetic null mutant in the aroM gene encoding an enzyme further upstream of chorismic acid strongly suggesting that inhibition of this pathway upstream from the chorismic acid branch point is a promising antifungal drug target. In support of this hypothesis and observation, the author's employed a tetracycline based conditional promoter replacement (CPR) technique targeting the aroB gene encoding a chorismate synthase enzyme. The resulting strain grew in the presence of doxycycline, but consistent with the genetic null mutant results upstream of the chorismate branch point, did not grow when supplemented with AAA. In the author's pulmonary murine leukopenic IA model, in the presence of doxycycline the CPR strain was fully virulent. However, in the absence of doxycycline a rather surprising ˜50% reduction in murine mortality was observed.

Given their in vitro results that suggested a complete loss of virulence was the likely expected result, the authors further investigated the mortality in the doxycycline untreated mice. Fungi isolated from the doxycycline untreated mice challenged with the CPR strain that succumbed to the infection revealed a genetic recombination event in the CPR module that led to constitutive expression of the targeted gene, aroB. Thus, the unexpected mortality in these mice with the CPR strain is potentially due to the genetic recombination event that appears to be potentiated in vivo (given the low frequency of recombination observed by the authors in vitro). Alternatively, induction of the cross-pathway control (CPC) system may be a confounding compensatory mechanism for the fungus to deal with AAA starvation in vivo during infection. Induction of this pathway in vivo in the CPR strain could potentially induce enough fungal proliferation to cause mortality in a select number of animals. In support of this hypothesis, the authors observe that the Shikimate pathway inhibitor, glyphosate, had a significant inhibitory effect on A. fumigatus only in the presence of a cpcA null mutant deficient in the cross pathway control system.

While further research is needed to fully validate the fungal Shikimate pathway as a viable antifungal drug target, the approaches used by Sasse and colleagues are a welcome advance in the fight against Aspergillus.¹ The further refinement of the DOX system for use with Aspergillus spp. in vivo presents a promising tool in this fungus to explore the in vivo impact of presumed essential genes on disease.^17,34 The importance of probing gene function in vivo really cannot be over-stated, as the in vivo environment still remains a vast frontier with many hidden nuances and complexities. Recent in vivo gene expression studies in Candida albicans and A. fumigatus highlight how different fungal genetic programs are in vivo vs. favorite in vitro culture conditions.^35,36 As filamentous fungal RNA is often a limiting factor in in vivo fungal gene expression profiling, the DOX system presented here also presents an opportunity to regulate expression of key virulence associated genes in vivo allowing fungal growth to levels needed to acquire fungal RNA. Additionally, it is often under appreciated that the murine models utilized to probe virulence place pressure on the genes under study from the very initiation of infection. In the clinic, patients present with established fungal lesions, often as biofilms, and the degree to which genes currently associated with A. fumigatus pathogenicity and virulence are relevant in these established lesions remains unknown. Many genes that are critical for the establishment of infection may not be necessary at later clinically relevant time points. Conditional expression systems such as the one described by Sasse and colleagues provides the Aspergillus community with an expanding tool-kit to probe these critically important questions and identify new robust antifungal drug targets.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Funding

Research in the laboratory of RAC is currently support by grant 2R01AI081838 from the National Institute of Allergy and Infectious Diseases of the National Institutes of Health and a fellowship from the Burroughs Wellcome Fund (BWF) Investigator in the Pathogenesis of Infectious Diseases Program. Additional support comes from a Cystic Fibrosis Research Development Program Award (Dr. Bruce Stanton, PI) and a National Institute of General Medical Sciences of the National Institutes of Health grant, Award Number P30GM106394 (Stanton, B. PtdIns). The content is solely the responsibility of the author and does not necessarily represent the official views of the National Institutes of Health or the BWF.