Alcohol pwns inflammation; Or, saga of alcohol dehydrogenase from Aspergillus
Aspergillus fumigatus and various other Aspergilli are ubiquitous molds. These are hardy aerobic saprotrophs, growing as easily on breads and potatoes as on plants and trees. However, many Aspergilli are capable of growing in nutrient-deficient or nutrient-absent environments, and surviving in extreme conditions, such as high temperature (up to 55oC) and pH; for example, A. niger, the Black Mold, can grow happily on damp walls. I have observed A. fumigatus grow on the surface of a highly alkaline buffer (pH9; one of the pH meter standards).

In addition, A. fumigatus has been found to be highly tolerant of a wide range of oxygen levels, from atmospheric (21%) to moderately aerobic (~14% as in the lung alveoli), low (2-4% as in tissues), or hypoxic (<1.5% as found in compost piles), even to as low as 0.1% (Source: References available with the article under review). For an obligate aerobe, A. fumigatus has evolved remarkably robust mechanisms that allow it to tolerate and thrive in extremely hypoxic conditions. Needless to say that such mechanisms are likely to come mighty handy when causing disease.

Many of the Aspergillus species are known to cause disease in human and animals; of these, A. fumigatus is the most common causal agent of invasive pulmonary aspergillosis, a frequent and life-threatening complication in several immunosuppressed patient populations. Microscopic airborne spores (‘conidia’) of the mold, produced copiously, are inhaled by the host (human and animals). Immunocompetent hosts mount an innate immune response that eliminates the conidia; however, both immunocompromised hosts (such as cancer and transplant patients), who cannot mount an efficient enough response, and patients of other chronic lung/airway diseases (such as allergic asthma), who mount an excessive and unmitigated inflammatory immune response, are both vulnerable to the disease produced by A. fumigatus conidia.

In order to cause disease, the conidia primarily colonize airways or the lungs and, if successful in breaching the innate immune defence, germinates into hyphae, long finger-like projections, that invade the tissues and blood vessels. The mechanisms by which A. fumigatus is able to survive and grow in the host environment are not all well-understood. Based on previous studies which demonstrated that inability of certain mutants of A. fumigatus and the yeast pathogen, Cryptococcus neoformans, to grow under hypoxic conditions correlated with their reduced virulence in mouse models, Grahl et al., leading a multi-institutional group of investigators, set out to discover if A. fumigatus encountered such hypoxic conditions in the host lung and how it dealt with it while infecting the host.

They found that
(a) hypoxic microenvironments do occur in three distinct immunosuppressed murine models; they discovered this by using a chemical hypoxia detection agent, pimonidazole hydrochloride, that enabled cool, visual estimation of hypoxia in tissue via immuno-fluorescence (Figure 2). From the observed extent of hypoxia, fungal growth, and host immune responses in different immunosuppressive models, the authors inferred that “the host inflammatory response plays an important, but not exclusive, role in the generation of the hypoxic microenvironment.”
(b) alcohol is involved (perhaps not surprisingly, since hypoxia must be a stressful situation!).

No, really. Using a 400 MHz 1H-NMR, Grahl et al. could detect substantial ethanol in 4 out of 10 immunosuppressed mice infected with A. fumigatus at day 3 post-infection, but none in uninfected mice.

Figure S1 Grahl et al., 2011, PLoS Pathog 7(7): e1002145

In order to ensure that the ethanol in the lung was of fungal origin (and not a result of, say, some wild Bacchanalian orgy the mice partook of in the middle of the night), the authors tested and established that A. fumigatus was indeed capable of fermenting glucose to ethanol in vitro in media containing minimal (1%) glucose under hypoxic conditions (1% oxygen, Figure 1) after 48, 72, and 96 hours of growth.

Analyzing fungal genes involved in the alcohol fermentation pathway, the authors zeroed in on the A. fumigatus gene alcC encoding an alcohol dehydrogenase whose expression is enhanced significantly in response to hypoxia. Interestingly, this gene appeared not to be contributing to A. fumigatus‘s ability to grow under hypoxic conditions, nor to the virulence of the mold – since mutants lacking this gene was as virulent as the wild type A. fumigatus in all models of immunosuppressed mice. However, in the model with cyclophosphamide induced neutropenia, as well as the one with corticosteroid induced immunosuppression, the mutant mold strain producing no alcohol had greatly reduced growth with evidence of significant inflammation when compared to the wild type (Figure 8); in the mice infected with the mutant mold strain, increased recruitment of immune effector cells, particularly neutrophils (Figure 9), and associated altered cytokine responses (Figure 10) were observed in the lung.

In other words, the alcohol of fungal origin may modulate the immune response by suppressing the inflammation, which may offer a survival advantage to the mold in the tissue. As always, alcohol makes everything better, especially when the mold brews it by itself.

The authors discuss one important caveat of the study: the observation of ethanol production in only 4 of the 10 infected mice. They offer several possible reasons that may have contributed to this, such as the lack of a more sensitive method of detection, unsuitability of bronchoalveolar lavage fluid as the site of interest and so forth. It’d be of interest to see if better detection methods – which they say they are developing – improve upon these results.

Another important caveat that the authors didn’t discuss lies in the model, particularly the method of immunosuppression. Corticosteroid treatment impairs the antifungal action of immune effector cells; in mice treated with a single dose of the corticosteroid Triamcinolone, it is perhaps not surprising that at day 3 post infection there was a rebound increase in inflammatory cells, led by neutrophils, which are after all the principal effectors against Aspergillus. Unfortunately, the authors didn’t check cellular infiltrate status in mice immunosuppressed with cyclophosphamide which they gave at day -2 and day 3 of infection. Cyclophosphamide causes profound neutropenia – as the authors have noted – and at the given dose, the neutropenia usually lasts for 96 hours. So, by day 3, one would expect a rebound neutrophilia in these mice prior, of course, to the second dose. It would have been interesting to see the cellular composition of the infiltrates in the cyclophosphamide-treated mice. One would expect the inflammation in this case to be largely macrophage/monocyte in nature, perhaps.

Overall, a rather interesting study with some intriguing findings; a good read.

Pathogens&rft_id=info%3Adoi%2F10.1371%2Fjournal.ppat.1002145&;bpr3.tags=Biology%2CMedicine%2CResearch+%2F+Scholarship">Grahl, N., Puttikamonkul, S., Macdonald, J., Gamcsik, M., Ngo, L., Hohl, T., & Cramer, R. (2011). In vivo Hypoxia and a Fungal Alcohol Dehydrogenase Influence the Pathogenesis of Invasive Pulmonary Aspergillosis PLoS Pathogens, 7 (7) DOI: 10.1371/journal.ppat.1002145


  1. Robert Cramer Jr

    August 2, 2011 at 6:49 pm

    Dear Kausik,

    Thank you for reading and commenting on our recent study. Yes, the difficulties in detecting the ethanol were frustrating for us as it prevented us from definitively attributing the in vivo phenotype of the alcC mutant to loss of ethanol production. The lack of cell wall alteration in the mutant argues strongly for a secreted factor being responsible, but without the robust ethanol data, we could not definitively state the mechanism to be due to ethanol. We are working on detecting EtOH in live animals, so it can be monitored over the time course of infection to further improve our understanding of the fungal-host interaction.

    With regard to the immunosuppression, the effects of triamcinolone and cyclophosphamide on the immune system are complex and still not fully understood. One important point: while the doses of cyclophosphamide used in our model and other murine models of IPA do induce leukopenia, most often this has been measured in uninfected animals. For example, see the IPA murine model papers by Steinbach et al. 2004 Medical Mycology and Sheppard et al. 2004 AAC. Of course, in the context of a microbial infection, we and others, find that the mice are never truly 100% leukopenic, particularly at the site of infection. In addition, I encourage you to take a look at work by David Cole’s group on the effects of Cyclophosphamide on host microenvironments (Salem et al. 2010, Journal of Immunology). Depending on dose, Cyclophosphamide can actually increase the number of precursor/circulating dendritic cells, which of course has important ramifications for IPA. This is further emphasized by Borna Mehrad’s recent work demonstrating an increase in dendritic cell recruitment to the lungs of neutropenic mice, and Stu Levitz’s recent work on the important of plasmacytoid dendritic cells in controlling A. fumigatus infection. We also did take a look at the cellular infiltrates in our cyclophosphamide treated animals. While the number of neutrophils was depleted compared to the triamcinolone only treated animals, a significant number of neutrophils were still observed at the site of infection, likely contributing to the hypoxia. So in the end, I think much work remains to be done regarding the impact of different immunosuppression regimens on the host response to Aspergillus fumigatus and how they contribute to the host microenvironment (along with how they affect the fungus!). Thanks again for your comments!

  2. Dr. Cramer,

    Thank you for your reasoned response. I am, by way of my work, familiar with the literature on cyclophosphamide-treated murine models. The way I understood Borna Mehrad’s observations was that the enhancement of circulating dendritic cells may have been a direct, and compensatory response to the profound neutropenia. Contrary to your observation in CD1 mice, I have found complete absence of neutrophils in C57 mice using a cyclophosphamide-only regimen – but as I noted in my comment, it may have been a function of the dosing. I use day -4, 0 and +4 at 150mg/kg body weight, which is different from yours. Marta Feldmesser’s work in Balb/c and C57 mice has shown this dosing to be adequate and safe for maintaining persistent neutropenia over a short study period. 

    However, I completely agree that our knowledge of immunosuppression regimens is incomplete, to say the least, as also is the knowledge of compensatory mechanisms that come into play when any particular componet of the innate defence system is MIA.

    Thanks again for your response, and for not minding my occasional flippancy about the outcomes! 😀

Leave a Reply

%d bloggers like this: