• Loading metrics

Novel cutting-edge metabolite-based diagnostic tools for aspergillosis

Novel cutting-edge metabolite-based diagnostic tools for aspergillosis

  • Masha G. Savelieff, 
  • Lucia Pappalardo

Metabolite-based detection of aspergillosis

Aspergillosis is a group of diseases caused by the inhalation of ubiquitous Aspergillus spores, generally Aspergillus fumigatus [1], that evade the host immune system [2]. Its most aggressive form, invasive aspergillosis (IA), carries a particularly grim prognosis in immunocompromised patients [2]. High mortality and the associated socioeconomic burden [3] necessitate early, accurate, and sensitive detection of Aspergillus infection. Many clinical diagnostic tests are available [4]; however, the identification of improved methods for the detection of aspergillosis is still an active field of research.

Pathogens possess a rich metabolism and a vast array of secondary metabolites, many unique to their species, that constitute a pathogen “fingerprint” [5]. Pathogens can leave their imprint on the host in other ways; for instance, host–pathogen interactions can alter the host’s own metabolome, leaving a “signature” of the disease caused by the pathogen [6]. Detecting metabolic evidence of the pathogen’s presence underlies several emerging methods of disease diagnosis. Due to the absence of an ideal diagnostic test, we have seen new applications of metabolite detection for Aspergillus in recent years. Methods include detection of gliotoxin [7,8] and siderophores [9], siderophore uptake [10,11], volatile organic compounds (VOCs) [1214], and changes to the host’s metabolome in serum [15]. Although still at the research stage, all of these techniques offer distinct advantages for aspergillosis detection (Table 1).

Table 1. Emerging metabolite-based methods of aspergillosis detection: Advantages and disadvantages of metabolite-based methods of aspergillosis detection at the research stage of development.

Aspergillosis diagnosis based on metabolite detection or uptake

Gliotoxin is a secondary metabolite secreted during hyphal growth by a number of fungi, including A. fumigatus [7], and it has been postulated that its detection might coincide with the early stages of infection. Gliotoxin is a potent immune suppressor and may contribute to the failure of the host immune system to prevent opportunistic fungal infections [8]. Gliotoxin has been investigated as a possible biomarker for aspergillosis [16,17], and a comprehensive study employing high-performance liquid chromatography tandem mass spectrometry (HPLC-MS/MS) demonstrated high accuracy and sensitivity for gliotoxin quantification in analytical standards and human serum [7]. Using this accurate method, the authors compared the technique to the galactomannan (GM) assay, a clinical ELISA aspergillosis detection method. Serum samples negative for GM were generally lacking gliotoxin (85%); however, half the GM positive samples were also devoid of gliotoxin [7].

Several scenarios could explain the lack of correlation between positive GM and gliotoxin level: (1) false GM positives, (2) decline in gliotoxin at late stages of infection when hyphal growth might slow as fungal burden increases, or (3) chemical instability of the gliotoxin disulfide in vivo [8]. The study did not classify cases of IA according to European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) definitions, so it is not possible to ascertain the degree to which false positive GM assays may have influenced the results of this study. Accurate IA status and duration of illness are also required to determine whether gliotoxin levels correlate with early infection versus late infection, and this was similarly not reported. Therefore, while the accuracy and sensitivity of the test is promising, an in-depth analysis with IA status and time frame will be required to evaluate the precise relation of gliotoxin to IA diagnosis.

To address the possibility of in vivo chemical instability of gliotoxin, levels of an inactive methylated metabolite of gliotoxin, bis(methylthio)gliotoxin (bmGT), were assayed in patient sera by high-performance thin-layer chromatography (HPTLC). These studies indicated that bmGT may be a better metabolic biomarker than gliotoxin, possibly due to its greater stability [8]. In this study, the IA status of patients was assessed according to EORTC/MSG criteria by clinicians who were unaware of the results of the gliotoxin and bmGT assays. Then, prediction of IA status by the GM assay was compared to prediction by bmGT quantification. Employing bmGT quantification had greater sensitivity and positive predictive value for IA than GM, and, combined, the 2 diagnostic tests identified all positive IA cases and almost totally avoided false negatives [8]. This study incorporated EORTC/MSG definitions of IA, strengthening the clinical applicability of these findings compared to the findings of the earlier gliotoxin study [7].

Siderophores are secondary metabolites utilized by microorganisms to scavenge the micronutrient iron, necessary for growth, from the environment or host [9]. Therefore, their presence implies an actively dividing pathogen. Triacetylfusarinine C (TAFC) is produced by a few fungal genera, Aspergillus amongst them, with no known human function, so it is neither anticipated in healthy hosts [9] nor expected to be taken up by host cells [10]. TAFC in serum from patients at risk of IA can be quantified by ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) [9]. Fifty-eight suspected, probable, or proven IA cases classified according to EORTC criteria had serum assayed for GM (≥0.5 index threshold) and TAFC. A positive correlation was observed between GM score and TAFC (Pearson r value = 0.77), but, interestingly, TAFC levels above the threshold of detection were observed more frequently in suspected IA than in probable/proven IA, possibly because the latter received antifungals [9]. Furthermore, a subset of samples from patients with suspected IA were TAFC positive but GM negative, suggesting TAFC secretion may occur early during infection [9]; however, the GM results may also have been false negatives, and further work is required to clarify this.

The analytical methods for metabolite (gliotoxin, bmGT, and TAFC) quantification are accurate and sensitive, but they require further research to uncover the true correlation between serum level with IA status. These studies will require evaluating the time course of metabolite production during the natural history of the disease and in patients with a range of severity of infection. Use of EORTC/MSG definitions of disease will be critical to allow comparison of the results of these studies with those evaluating other diagnostic tests.

TAFC can bind the positron emitter gallium-68 (68Ga) creating 68Ga-TAFC complexes for positron emission tomography (PET) [10]. Biodistribution studies in healthy mice demonstrated rapid renal 68Ga-TAFC elimination, suggesting it was not actively taken up in healthy animals [10]. This was confirmed in Lewis rats in which 68Ga-TAFC uptake in the lungs was selective in rats with IA and not observed in healthy rats and Aspergillus-challenged rats that did not develop IA [10]. 68Ga-TAFC uptake was dependent on infection severity, suggesting deliberate uptake by Aspergillus.

Aspergillus-challenged rats that did not develop IA did not absorb 68Ga-TAFC, implying that the technique differentiates between actively growing, invading Aspergillus compared to exogenous, inert spores inhabiting the airways [10]. This observation also suggests that the method could detect IA early in the course of infection when Aspergillus is in a phase of rapid growth and prior to the development of a high burden of fungal disease [10]. In addition, whole body scans can reveal infection spread to organs other than lungs. 68Ga-TAFC uptake by Aspergillus was greater than by other fungi and almost nonexistent by other microorganisms, indicating it could be specific for Aspergillus [11]. The high uptake and retention time of 68Ga-TAFC by Aspergillus enhances sensitivity [10] but lengthens the duration patients would be exposed to radiation, a disadvantage of the technique. The application of gallium-68 complexes in nuclear medicine is well documented [18]; however, the safety of gallium-68 with TAFC in particular needs to be evaluated, but studies thus far indicate that 68Ga-TAFC uptake may be a useful diagnostic tool for IA.

Aspergillosis diagnosis based on VOCs

Aspergillus produces several volatile metabolites, which can be present in the breath of patients with lung infection [19]. Detection of these VOCs is an attractive strategy for the detection of IA, since breath collection is noninvasive and desirable for critically ill patients susceptible to IA. Earlier attempts to detect VOCs focused on the detection of single metabolites such as 2-pentylfuran, which was elevated in patients with Aspergillus airway colonization [20]. However, 2-pentylfuran is not unique to Aspergillus, and confounding factors affected diagnosis based solely on this VOC [21]. Therefore, more recent methods have included the detection of several VOCs or entire mixtures.

eNoses are affordable, portable options for VOC detection, possibly making point-of-care medical diagnosis a reality [22]. Numerous commercial eNoses exist that vary in the sensor material in order to make them suitable for different applications [23,24]. eNoses contain sensor arrays whose physical properties change upon adsorption of volatiles, e.g., electrical resistance, which is recorded. Volatile metabolites from a sample bind to the sensor arrays, and the combination of these volatiles will produce a pattern of response that reflects this mix of metabolites [23,24]. This response pattern is unique to that particular VOC mixture and forms the basis of detection.

The feasibility of eNose-based detection of aspergillosis was examined in patients with prolonged chemotherapy-induced neutropenia (PCIN) [12] and cystic fibrosis (CF) [13]. PCIN patients underwent full diagnostic workup for aspergillosis (classified by EORTC criteria) whilst CF patients were diagnosed by sputum culture. Each patient’s breath was analyzed by a Cyranose 320 eNose to produce a “breathprint,” a variation in the electrical resistance differential across all 32 eNose sensors. All “breathprint” data from uninfected controls and patients with proven and probable IA were subjected to principal component analysis, identifying signal components that accounted for 99.9% variance between control and experiment. These components were used to construct prediction models, which had cross-validation of about 89% or higher [12,13].

The eNose’s ability to detect aspergillosis in the presence of 2 underlying medical conditions (PCIN and CF) demonstrates its broad applicability [12,13]. Early eNose detection of VOCs from the host inflammatory response to infection may allow for the detection of infection prior to the development of significant fungal disease. It may even be possible to make eNose detection specific for Aspergillus since breathprints of CF patients with lung coinfections by other microorganisms were distinct from breathprints from patients with aspergillosis [13], as supported by in vitro tests [25]. The main drawback of eNose technology is that prior analysis is required to establish the prediction model, and a reliable calibration system is required to ensure the same operation between different eNose machines. These challenges will have to be addressed before eNose technology can become a widely used standardized diagnostic technique.

Gas-chromatography mass spectrometry (GC-MS) is another method that can be used to identify volatile metabolites based on their mass spectrometry (MS) profile. Combined with patient breath collection, it is noninvasive and can structurally identify metabolites in “disease breath.” GC-MS was used to identify secondary metabolite mixtures unique to A. fumigatus compared to other aspergilli [14]. The combination of monoterpenes (camphene, α-pinene, β-pinene, limonene) and sesquiterpenes (α-trans-bergamotene, β-trans-bergamotene) was found to be specific to A. fumigatus in vitro [14]. This observation was recapitulated in IA patient breath samples, which additionally contained 2 metabolites not detected in vitro, the terpenoid ketone trans-geranylacetone and a β-vatirenene–like sesquiterpene [14].

The IA status of 64 enrolled patients was evaluated independently by 2 doctors as “proven,” “probable,” or “possible” according to EORTC/MSG criteria. Breath VOCs were submitted to GC-MS and analyzed by heat map for the 8 metabolites found to define A. fumigatus. IA status of 60/64 cases were correctly identified by this GC-MS analysis, highlighting the potential of this technique for IA diagnosis [14]. GC-MS of patient breath is noninvasive and can be tailored to detect other invasive pathogens to extend its utility to the diagnosis of other diseases. That would require preliminary in vitro work to discover secondary metabolites or combinations unique to a pathogen/strain followed by validation in patient breath samples. This requires more labor and doesn't guarantee that findings in vitro translate to patient samples. In addition, it may not be possible to discover a combination of metabolites unique to the pathogen, which is the principal drawback of this method.

Aspergillosis diagnosis based on metabolomics

Metabolomics is the untargeted, system-wide detection of metabolites in biological samples and has great potential for clinical use. It may be employed to either (1) detect metabolite mixtures unique to infecting pathogens or (2) observe changes to the host’s metabolome caused by infection [26]. Typically, metabolite detection is by MS or nuclear magnetic resonance (NMR), followed by statistical analysis to reveal system-wide differences in all systematically varying metabolites between healthy versus infected patients [26].

NMR metabolomics is well suited to study biofluids in their native state and was applied to aspergillosis detection in falcons (gyrfalcons and gyr-x peregrine hybrids) [15]. Blood samples were withdrawn from clinically healthy falcons and falcons with confirmed aspergillosis and analyzed by 1D 1H-NMR spectroscopy. Multivariate statistical analysis of all identified host metabolites from the resultant spectra showed a clear metabolic separation of healthy versus diseased falcons, each with a distinct metabolic profile [15]. Additionally, because NMR metabolomics can identify metabolites, information about which contributed to the distinction between healthy versus diseased cohorts was revealed. In particular, 3-hydroxybutyrate was significantly elevated in the blood of aspergillosis falcons compared to healthy raptors [15].

NMR metabolomics produces a top-down systematic cataloging of host metabolites that vary consistently between infected and healthy patients [26]. Unlike other diagnostic methods that detect Aspergillus or its components, NMR metabolomics produces a host “disease metabolic profile” of aspergillosis. This indicator of disease differentiates from the detection of exogenous Aspergillus spores by some methods that could give false positives [27]. The promising, preliminary study in falcons suggests that 1H-NMR metabolomics could translate to humans as a detection tool for aspergillosis.

Despite the advantages of the systems-wide approach by NMR metabolomics, it suffers from low sensitivity and resolution. While the more abundant, primary metabolites may be readily identified, less abundant and more structurally complex secondary metabolites are more challenging to identify, although advances in data analysis are constantly being made.

Aspergillosis detection and beyond: Where else can metabolite detection take us?

Several metabolite-based methods have shown promise for improved Aspergillus detection, which satisfy the criteria of early detection, low invasiveness, low cost, and point-of-care [715]. The timely detection of aspergillosis will significantly facilitate the decision to treat infected immunocompromised patients—to improve their prognosis for recovery whilst also preventing unnecessary prophylactic administration of potentially toxic antifungals.

Beyond Aspergillus detection, metabolite identification and quantification in addition to the more recent system-wide metabolomics methods may be extended to other microbial and viral infections. Metabolomics can address questions pertaining to infection mechanisms and host–pathogen interactions [28]. These studies may reveal pathogen vulnerabilities that may be exploited to develop therapeutic strategies. Therefore, in addition to diagnosis, metabolomics may be applied in drug design, an essential area of research in this era of mounting microbial resistance [29]. The promise of metabolomics is only now unfolding, and its utility is being recognized in various research areas with novel potential future applications.


LP wishes to recognize PN Borer for his invaluable insight and thanks him for his suggestions. LP also gratefully acknowledge J Gannon’s contribution towards proofreading and comments made on an earlier draft of this manuscript. MGS would like to thank the Castro-Lowenstein laboratory at the University of Michigan for academic resources.


  1. 1. Sugui JA, Kwon-Chung KJ, Juvvadi PR, Latgé JP, Steinbach WJ. Aspergillus fumigatus and related species. Cold Spring Harb Perspect Med. 2015; 5:a019786. pmid:25377144
  2. 2. Segal BH. Aspergillosis. N Engl J Med. 2009; 360:1870–1884. pmid:19403905
  3. 3. Drgona L, Khachatryan A, Stephens J, Charbonneau C, Kantecki M, Haider S et al. Clinical and economic burden of invasive fungal diseases in Europe: focus on pre-emptive and empirical treatment of Aspergillus and Candida species. Eur J Clin Microbiol Infect Dis. 2014; 33:7–21. pmid:24026863
  4. 4. Steinbach WJ. Are we there yet? Recent progress in the molecular diagnosis and novel antifungal targeting of Aspergillus fumigatus and invasive aspergillosis. PLoS Pathog. 2013; 9: e1003642. pmid:24204250
  5. 5. Frisvad JC, Rank C, Nielsen KF, Larsen TO. Metabolomics of Aspergillus fumigatus. Med Mycol. 2009; 47:S53–S71. pmid:18763205
  6. 6. Ko ER, Yang WE, McClain MT, Woods CW, Ginsburg GS, Tsalik EL. What was old is new again: using the host response to diagnose infectious disease. Expert Rev Mol Diagn. 2015; 15:1143–1158. pmid:26145249
  7. 7. Cerqueira LB, de Francisco TMG, Gasparetto JC, Campos FR, Pontarolo R. Development and validation of an HPLC-MS/MS method for the early diagnosis of aspergillosis. PLoS ONE. 2014; 9:e92851. pmid:24690884
  8. 8. Vidal-García M, Domingo MP, De Rueda B, Roc L, Delgado MP, Revillo MJ et al. Clinical validity of bis(methylthio)gliotoxin for the diagnosis of invasive aspergillosis. Appl Microbiol Biotechnol. 2016; 100:2327–2334. pmid:26678078
  9. 9. Carroll CS, Amankwa LN, Pinto LJ, Fuller JD, Moore MM. Detection of a serum siderophore by LC-MS/MS as a potential biomarker of invasive aspergillosis. PLoS ONE.2016; 11: e0151260. pmid:26974544
  10. 10. Petrik M, Franssen GM, Haas H, Laverman P, Hörtnagl C, Schrettl M et al. Preclinical evaluation of two 68Ga-siderophores as potential radiopharmaceuticals for Aspergillus fumigatus infection imaging. Eur J Nucl Med Mol I. 2012; 39:1175–1183. pmid:22526953
  11. 11. Petrik M, Haas H, Laverman P, Schrettl M, Franssen GM, Blatzer M et al. 68Ga-triacetylfusarinine C and 68Ga-ferrioxamine E for Aspergillus infection imaging: uptake specificity in various microorganisms. Mol Imaging Biol. 2014; 16:102–108. pmid:23818006
  12. 12. de Heer K, van der Schee MP, Zwinderman K, van den Berk IAH, Visser CE, van Oers R et al. Electronic nose technology for detection of invasive pulmonary aspergillosis in prolonged chemotherapy-induced neutropenia: a proof-of-principle study. J Clin Microbiol. 2013; 51: 1490–1495. pmid:23467602
  13. 13. de Heer K, Kok MGM, Fens N, Weersink EJM, Zwinderman AH, van der Schee MPC et al. Detection of airway colonization by Aspergillus fumigatus by use of electronic nose technology in patients with cystic fibrosis. J Clin Microbiol. 2016; 54:569–575. pmid:26677251
  14. 14. Koo S, Thomas HR, Daniels SD, Lynch RC, Fortier SM, Shea MM et al. A breath fungal secondary metabolite signature to diagnose invasive aspergillosis. Clin Infect Dis. 2014; 59:1733–1740. pmid:25342502
  15. 15. Pappalardo L, Hoijemberg PA, Pelczer I, Bailey TA. NMR-Metabolomics study on falcons affected by aspergillosis. Curr Metabolomics. 2014; 2:155–161
  16. 16. Grovel O, Kerzaon I, Petit K, Du Pont TR, Pouchus Y-F. A new and rapid bioassay for the detection of gliotoxin and related epipolythiodioxopiperazines produced by fungi. J Microbiol Methods. 2006; 66: 286–293. pmid:16451813
  17. 17. Lewis RE, Wiederhold NP, Chi J, Han XY, Komanduri KV, Kontoyiannis DP et al. Detection of gliotoxin in experimental and human aspergillosis. Infect Immun. 2005; 73: 635–637. pmid:15618207
  18. 18. Banerjee SR, Pomper MG. Clinical applications of Gallium-68. Appl Radiat Isot. 2013; 76: 2–13. pmid:23522791
  19. 19. Heddergott C, Calvo AM, Latgé JP. The volatome of Aspergillus fumigatus. Eukaryot Cell. 2014; 13: 1014–1025. pmid:24906414
  20. 20. Syhre M, Scotter JM, Chambers ST. Investigation into the production of 2-pentylfuran by Aspergillus fumigatus and other respiratory pathogens in vitro and human breath samples. Med Mycol. 2008; 46:209–215. pmid:18404548
  21. 21. Bhandari S, Chambers S, Pearson J, Syhre M, Epton M, Scott-Thomas A. Determining the limits and confounders for the 2-pentylfuran breath test by gas chromatography/mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci. 2011; 879:2815–2820. pmid:21880557
  22. 22. Friedrich MJ. Scientists seek to sniff out diseases: electronic “noses” may someday be diagnostic tools. JAMA. 2009; 301:585–586. pmid:19211458
  23. 23. Wilson AD, Baietto M. Advances in electronic-nose technologies developed for biomedical applications. Sensors. 2011; 11:1105–1176. pmid:22346620
  24. 24. Wilson AD, Baietto M. Applications and advances in electronic-nose technologies. Sensors.2009; 9: 5099–5148. pmid:22346690
  25. 25. de Heer K, Vonk SI, Kok M, Kolader M, Zwinderman AH, van Oers MHJ et al. eNose technology can detect and classify human pathogenic molds in vitro: a proof-of-concept study of Aspergillus fumigatus and Rhizopus oryzae. J Breath Res. 2016; 10:036008. pmid:27447026
  26. 26. Nicholson JK, Lindon JC. Systems biology: metabonomics. Nature. 2008; 455:1054–1056. pmid:18948945
  27. 27. Khot PD, Fredricks DN. PCR-based diagnosis of human fungal infections. Expert Rev Anti Infect Ther. 2009; 7:1201–1221. pmid:19968513
  28. 28. Romani L, Zelante T, Palmieri M, Napolioni V, Picciolini M, Velardi A et al. The cross-talk between opportunistic fungi and the mammalian host via microbiota's metabolism. Semin Immunopathol. 2015; 37:163–171. pmid:25404119
  29. 29. Laxminarayan R, Matsoso P, Pant S, Brower C, Røttingen JA, Klugman K et al. Access to effective antimicrobials: a worldwide challenge. Lancet. 2016; 387:168–175. pmid:26603918