Aspergillus niger upregulated glycerolipid metabolism and ethanol utilization pathway under ethanol stress

Abstract

The knowledge of how Aspergillus niger responds to ethanol can lead to the design of strains with enhanced ethanol tolerance to be utilized in numerous industrial bioprocesses. However, the current understanding about the response mechanisms of A. niger toward ethanol stress remains quite limited. Here, we first applied a cell growth assay to test the ethanol tolerance of A. niger strain ES4, which was isolated from the wall near a chimney of an ethanol tank of a petroleum company, and found that it was capable of growing in 5% (v/v) ethanol to 30% of the ethanol-free control level. Subsequently, the metabolic responses of this strain toward ethanol were investigated using untargeted metabolomics, which revealed the elevated levels of triacylglycerol (TAG) in the extracellular components, and of diacylglycerol, TAG, and hydroxy-TAG in the intracellular components. Lastly, stable isotope labeling mass spectrometry with ethanol-d₆ showed altered isotopic patterns of molecular ions of lipids in the ethanol-d₆ samples, compared with the nonlabeled ethanol controls, suggesting the ability of A. niger ES4 to utilize ethanol as a carbon source. Together, the studies revealed the upregulation of glycerolipid metabolism and ethanol utilization pathway as novel response mechanisms of A. niger ES4 toward ethanol stress, thereby underlining the utility of untargeted metabolomics and the overall approaches as tools for elucidating new biological insights.

INTRODUCTION

Organic solvent-tolerant microbes play key roles in many industrial bioprocesses, such as biofuel production, biocatalysis, and bioremediation (Nicolaou, Gaida, & Papoutsakis, ). To obtain strains with tolerance to these solvents, an approach involves genetically engineering selected strains based on the knowledge of organic solvent-induced stresses and responses (Taylor, Tuffin, Burton, Eley, & Cowan, ; Torres, Pandey, & Castro, ), which include repression or activation of sporulation (Bohin, Rigomier, & Schaeffer, ), induction of stress proteins (Petersohn et al., ), biodegradation or secretion of toxic organic solvents (Aono, Tsukagoshi, & Yamamoto, ; Bustard, Whiting, Cowan, & Wright, ), alteration in cell morphology (Neumann et al., ), and adaptation of the cell surface and cell membrane (Aono & Kobayashi, ; Weber & de Bont, ). Because these responses might be triggered to counteract chemical toxins, the elevation in their levels might lead to the development of tolerance traits in these microorganisms (Kajiwara et al., ; Kang et al., ; Mahipant, Paemanee, Roytrakul, Kato, & Vangnai, ; Vinayavekhin & Vangnai, ).

Aspergillus niger is a filamentous ascomycete fungus, which can be found in almost every environment. It is known as the black mold on rotting fruits and vegetables. Yet, despite these common views of A. niger as an undesirable contaminant, it rarely causes disease in humans (Person, Chudgar, Norton, Tong, & Stout, ). In fact, it has the GRAS (generally regarded as safe) status for many of its processes (Frisvad et al., ) and is one of the most economically useful fungi in the biotechnological industry (Pel et al., ). It has been applied in the fermentation process for the production of organic acids, such as gluconic (Ramachandran, Fontanille, Pandey, & Larroche, ) and citric acids, with production of the latter exceeding one million metric tons annually (Baker, ), and of various extracellular enzymes, including α-amylase or β-glucosidase (Pariza & Cook, ). Apart from these industrial usages, A. niger has also been utilized in bioremediation processes (Coulibaly, Naveau, & Agathos, ; Srivastava & Thakur, ), as heterologous hosts for proteins and secondary metabolites production (Lubertozzi & Keasling, ) and as a cofermentation partner with Saccharomyces cerevisiae for the simultaneous saccharification and fermentation in bioethanol production (Izmirlioglu & Demirci, ). The further studies and usages of A. niger have also been facilitated by the availability of the genome sequence of three different A. niger strains (NRRL3, ATCC1015, and CBS513.88) (Baker, ; Pel et al., ).

Recently, one of the black spots found on the roof and outside upper wall of an ethanol tank of a petroleum company was investigated and identified as a living organism, A. niger strain ES4. The black spots could be found most densely near the valved chimney of the tank where ethanol was allowed to evaporate, which indicated the preference of this strain of A. niger for ethanol. This finding was rather surprising, as most non-ethanol-producing species are not capable of tolerating a high concentration of ethanol.

Relating to ethanol tolerance, A. niger isolated from spoiled pastry products was shown previously to be able to grow on potato dextrose agar containing ethanol up to about 3% (w/w) with almost no growth defect and up to 4% (w/w) with about 50% reduction in its growth (Dantigny, Guilmart, Radoi, Bensoussan, & Zwietering, ). Another unrelated study also exhibited the capabilities of A. niger to grow weakly on a plate containing 1% ethanol as a sole carbon source (O'Connell & Kelly, ). However, while A. niger was demonstrated to be capable of tolerating some concentrations of ethanol in many cases, and while a transcriptomic analysis of its closely related fungus Aspergillus nidulans revealed a 10-fold and twofold upregulation of alcohol dehydrogenase alcA and aldA genes in minimal medium containing ethanol compared with glucose, respectively (Mogensen, Nielsen, Hofmann, & Nielsen, ), there have so far been no reports on the metabolic responses of A. niger toward ethanol. The knowledge of which would allow for the engineering of the strain to have either higher tolerance toward ethanol for utilization in biotechnological industry or for designing novel methods for eradication of the strain in unwanted situations, such as in food spoilage or on a wall of an ethanol tank. We therefore decided to study the metabolic responses of A. niger ES4 toward ethanol further.

In this study, ethanol tolerance of A. niger ES4 was first examined. Then, to understand its metabolic responses toward ethanol, untargeted metabolomics analysis was conducted to assay extracellular and intracellular hydrophobic compound changes in A. niger ES4 when put under ethanol stress. Lastly, since it was possible that this strain of A. niger might intake ethanol for nutrients or substrates for production of some metabolites, the incorporation of ethanol into its metabolites was interrogated by stable isotope labeling mass spectrometry (MS) experiments using ethanol-d₆.

MATERIALS AND METHODS Fungal strain and growth conditions

Aspergillus niger strain ES4 was isolated from a black spot on the outside upper wall of an ethanol tank of a petroleum company by the serial dilution method (Clark, Bordner, Galdrich, Kabler, & Huff, ). It was identified based on morphological characteristics and then confirmed using molecular technique. The nucleotide sequence data were submitted into the DDBJ/EMBL/GenBank nucleotide sequence databases with accession number MK621333.

The fungus was grown on potato dextrose agar for 7 days, before three agar plugs were inoculated in 20 ml of potato dextrose broth and shaken at 180 rpm, room temperature for 3 days. The culture was then diluted 20-fold into 20 ml of minimal medium (MM; per liter: 6 g NaNO₃, 0.52 g KCl, 0.52 g MgSO₄.7H₂O, 1.52 g KH₂PO₄, 10 g glucose, 2 ml Hutner's trace elements, pH 6.8) (Barratt, Johnson, & Ogata, ) with ethanol (Merck, absolute, ≥99.9%), water (as control), or ethanol-d₆ (Merck, deuteration degree ≥ 99%; for stable isotope labeling MS) at the indicated concentrations and shaken further until the predetermined time.

Determination of the dry weight

Mycelia from three 20-ml cultures were combined and collected on a dry, preweighed Whatman paper no. 1 by vacuum filtration. They were then washed with distilled water (4 × 5 ml, then 2 × 20 ml) and dried on the filter paper at 70°C until at a constant weight (dry weight [DW]).

Metabolites extraction and analysis

Mycelia and supernatant from the 3-day-old 20-ml A. niger culture were separated by gravity filtration through a cotton ball. A mixture of 10 ml of chloroform and 5 ml of methanol was added to the supernatant, while mycelia were washed once with 10 ml of distilled water and soaked overnight in a mixture of 3 ml of chloroform and 1.5 ml of methanol, before 1.5 ml of MM without glucose was added to them. Subsequently, all mixtures were shaken vigorously, and centrifuged at 1500 g, room temperature for 3 min to separate the organic layer (bottom) from the aqueous layer (top). The organic layer was transferred to another glass vial, evaporated to dryness under a steam of nitrogen, and placed at –20°C for storage. The extracts were reconstituted in 200 µl of chloroform prior to analysis by liquid chromatography (LC)–MS.

For LC–MS and LC–MS/MS analyses, 40 µl of each sample was quantitated on an Ultimate DGP-3600SD LC coupled to a Bruker MicrOTOF Q-II MS instrument, both in the positive and negative ion modes, as described previously (Vinayavekhin et al., ).

LC–MS untargeted data analysis

The total ion chromatograms from each sample group (i.e., control vs. ethanol treatment) were obtained in triplicate. The total of six chromatograms for mycelia samples and six chromatograms for supernatant samples were then subjected to comparative data analyses separately as previously described (Vinayavekhin, Mahipant, Vangnai, & Sangvanich, ), except that (a) the data were normalized by the average DW of the cultures instead of the optical density at 600 nm and (b) the minimum integrated mass ion intensity (MSII) was set at 5,000 instead of 30,000.

Stable isotope labeling MS with ethanol-d₆

Intracellular metabolites from A. niger cultures treated with ethanol-d₆ were extracted and analyzed by LC–MS exactly as described above in the section “2.3 Metabolites extraction and analysis.” The resulting chromatograms were then inspected manually to obtain the mass spectra of the indicated ions.

RESULTS Ethanol tolerance of A. niger ES4

Since A. niger ES4 was isolated from the outside upper wall of an ethanol tank, we first assessed its ethanol tolerance. It, however, was not possible to monitor the growth of A. niger on the solid agar medium containing ethanol, which best mimicked its growth on the wall of the tank, because its spore interfered with the radial growth (data not shown). We therefore determined its ethanol tolerance using a cell growth assay in a defined liquid medium instead (Mahipant et al., ). In this assay, A. niger ES4 mycelia (60 ml) were cultured in MM adapted slightly from that used by Barratt et al. () for culturing Aspergillus nidulans, and with ethanol added at concentrations up to 5% (v/v). Then, their growth was monitored daily over a 5-day period by measuring the DW of mycelia.

The A. niger strain ES4 was able to grow in ethanol at all tested concentrations (2%–5% [v/v]), although at slower growth rates than the no-ethanol control (Figure ). Increasing ethanol concentrations decreased the DW at each measured time point in a dose-dependent manner and became lowest at 5% (v/v) ethanol. The DW amounted to 78%, 65%, 49%, and 30% of that of the ethanol-free control at 2%, 3%, 4%, and 5% (v/v) ethanol, respectively, at day 4 when the cells were solidly in the stationary phase. Overall, the data revealed some degree of tolerance toward ethanol by the A. niger ES4 strain.

Growth curves of A. niger ES4 in MM with concentrations of ethanol from 0% to 5% (v/v). Data are shown as the average DW of mycelia from a 60-ml culture ± standard error of the mean for triplicate experiments per concentration

Metabolomics of A. niger under ethanol stress

To further characterize the microbe and understand the responses of A. niger toward ethanol, metabolomics analysis was performed on both the extracellular and intracellular components of A. niger ES4 cultures in the presence and absence of 4% (v/v) ethanol at day 3, which was the condition that induced moderate stress levels to the fungi (i.e., 52% growth of that without ethanol) and at the day the cells entered early stationary phase (Figure ). The mycelia or culture supernatant was then extracted for analysis of the hydrophobic metabolites using a 2:1 (v/v) ratio of chloroform:methanol, and the extracts were concentrated and analyzed by LC–MS using a previously developed untargeted metabolomics platform (Vinayavekhin et al., ).

To identify differential metabolites related to ethanol stress responses, the XCMS program (Smith, Want, O'Maille, Abagyan, & Siuzdak, ) was used to obtain an MSII value for each detectable metabolite ion in each LC–MS chromatogram, and the MSII values were normalized by the DW to account for the differences in fungal growth. Ions were then regarded as potential responses to ethanol stress only if they were up- or downregulated by fourfold or more with statistical significance (Student's t test with p < .05) in the ethanol-treated samples compared with the controls, and only if they also met these criteria in another set of independent experimental repeat. Using these criteria, the unbiased comparative analyses revealed 68 and 7 upregulated ions and 1 and 1 downregulated ions in the supernatant, and 322 and 29 upregulated ions and 14 and 24 downregulated ions in the mycelia under ethanol stress in the positive and negative ion modes, respectively (Figure ).

Volcano plots of metabolite changes in A. niger ES4 at day 3 caused by 4% (v/v) ethanol. Each (a) extracellular and (b) intracellular metabolite ion in the hydrophobic components with an average MSII above 5,000 counts is plotted as its statistical significance (p-value) against the fold change of ethanol (EtOH) over the control. The ions that locate above the horizontal dash line and outside the two vertical dash lines have a p-value of less than 0.05 and a fold change of greater than 4, respectively. Each plot contains data from both negative (neg) and positive (pos) ion modes. However, only some positive-mode MS ions with p [less than] .05 could be identified (pos (ID)) in this study

Next, the structural characterization of these ions with altered levels following ethanol treatment was undertaken manually using the combined clues from the accurate mass, previously reported retention time (RT) (Vinayavekhin et al., , ), and tandem mass spectra (Appendix 1: Tables A1 and A2 and Figure S1). Structures could be assigned to five extracellular and 63 intracellular upregulated positive-mode ions. All of these ions were in the family of triacylglycerol (TAG) for the extracellular components, and the families of diacylglycerol (DAG), TAG, and hydroxy-(h)TAG for the intracellular components (Table , and Appendix 1: Tables A1 and A2 with the sn-1, sn-2 and sn-3 side chains written in random order and exact positions of the hydroxyl groups on hTAG unspecified). The most commonly found acyl chains in these altered lipids were 16:0, 18:0, 18:1, and 18:2. The remaining uncharacterized changed ions could not be grouped into the same families as other changed ions, were detected at relatively lower MSII, or were potentially classified as ion fragments or adducts of other smaller or larger molecules. As references, we also performed targeted analyses of other lipids in the biosynthetic pathways of DAG and TAG, such as phospholipids (see Figure for details), and found their levels under ethanol stress more or less undifferentiated from those of the controls (Table , Appendix 1: Table A3, and Figure S1). Together, the untargeted metabolomics analysis suggested the involvement of glycerolipids in response to ethanol stress in A. niger ES4.

Relative levels of identified ethanol-upregulated lipids and of other related lipids

Lipid class and acyl chain	Ion	m/z	RT (min)	EtOH/con
Upregulated lipids in ethanol-treated extracellular samples
Triacylglycerol (TAG)
16:0/18:1/18:2	[M + NH4]+	874.7830	48.3	4.8*
18:1/18:2/18:2	[M + NH4]+	898.7839	48.1	6.6*
18:1/18:1/18:2	[M + NH4]+	900.7989	48.6	4.1†
18:1/18:1/18:1	[M + NH4]+	902.8136	48.7	7.7†
18:0/18:1/18:1	[M + NH4]+	904.8272	48.9	8.3†
Upregulated lipids in ethanol-treated intracellular samples
Diacylglycerol (DAG)
18:2/18:2	[M + NH4]+	634.5394	43.8	7.5*
18:1/18:2	[M + NH4]+	636.5550	44.4	6.7*
18:1/18:1	[M + NH4]+	638.5700	44.9	6.7†
18:0/18:1	[M + NH4]+	640.5815	45.3	8.7†
18:2/20:2	[M + NH4]+	662.5667	44.6	8.0*
18:2/20:1	[M + NH4]+	664.5833	45.1	7.0*
TAG
12:0/18:2/18:2	[M + NH4]+	816.7036	47.6	8.4†
14:0/18:2/18:2	[M + NH4]+	844.7360	47.8	4.4‡
18:0/18:1/18:1	[M + NH4]+	904.8319	49.0	4.5§
18:0/18:0/18:1	[M + NH4]+	906.8451	49.1	5.4§
18:0/18:1/20:0	[M + NH4]+	934.8749	49.5	4.4§
18:0/18:1/24:0	[M + NH4]+	990.9377	50.2	4.1§
18:1/18:1/25:0	[M + NH4]+	1,002.9375	50.2	5.0†
18:0/18:1/26:0	[M + NH4]+	1,018.9654	50.5	4.7§
Hydroxy-TAG (hTAG)
16:0/16:1(OH)/18:2	[M + NH4]+	862.7455	46.9	4.6‡
16:0/16:0/18:2(OH)	[M + NH4]+	864.7565	47.2	4.2§
16:0/18:2/18:2(OH)	[M + NH4]+	888.7606	47.1	8.2*
16:0/18:1/18:2(OH)	[M + NH4]+	890.7736	47.3	5.4†
18:1/18:2(OH)/18:3	[M + NH4]+	912.7626	46.9	11.2*
18:1/18:2/18:2(OH)	[M + NH4]+	914.7771	47.2	11.8*
18:1/18:1/18:2(OH)	[M + NH4]+	916.7929	47.5	8.8‡
18:0/18:1/18:2(OH)	[M + NH4]+	918.8068	47.7	9.9†
Other intracellular lipids in the related pathways
Fatty acid (FA)
16:0	[M – H]–	255.2317	18.6	1.6†
18:2	[M – H]–	279.2336	18.5	1.7†
18:1	[M – H]–	281.2478	18.8	2.0‡
18:0	[M – H]–	283.2624	19.2	1.5‡
Monoacylglycerol (MAG)
16:0	[M + Na]+	353.2656	34.0	2.4*
18:2	[M + Na]+	377.2718	33.1	2.1
Phosphatidic acid (PA)
16:0/18:2	[M–H]–	671.4649	27.4	1.6
18:2/18:2	[M–H]–	695.4647	26.8	2.3*
Phosphatidylethanolamine (PE)
16:0/18:2	[M – H]–	714.5023	39.0	1.0
18:2/18:2	[M – H]–	738.5004	38.2	1.7
Phosphatidylserine (PS)
16:0/18:2	[M – H]–	758.4944	29.9	1.3
18:2/18:2	[M – H]–	782.4946	29.3	1.5
Phosphatidylglycerol (PG)
16:0/18:2	[M – H]–	745.4976	34.4	1.4
18:2/18:2	[M – H]–	769.4962	33.8	1.5
Phosphatidylinositol (PI)
16:0/18:2	[M – H]–	833.5160	34.0	1.2
18:2/18:2	[M – H]–	857.5147	33.4	1.1
Phosphatidylcholine (PC)
16:0/18:2	[M + H]+	758.5726	41.7	0.8†
18:2/18:2	[M + H]+	782.5751	41.3	1.8§

Abbreviations: m/z, mass-to-charge ratio; RT, retention time.

²EtOH/con value represents the ratio of the average mass ion intensity of ethanol-treated sample group and that of the control.

³Student's t test: *, p < .05; †, p < .01; ‡, p < .005; §, p < .001; N = 3.

Biosynthesis of glycerolipids and phospholipids in A. niger (Kanehisa & Goto, ), starting from ethanol. The green boxes indicate lipids that had statistically significantly elevated levels, whereas the gray boxes show other lipids whose levels were quantitated in this study and the yellow box emphasizes where ethanol locates in the pathways. Glycerolipids include monoacylglycerol (MAG), diacylglycerol (DAG), and triacylglycerol (TAG), whereas phospholipids shown are phosphatidic acid (PA), phosphatidylethanolamine (PE), phosphatidylcholine (PC), phosphatidylserine (PS), phosphatidylglycerol (PG), and phosphatidylinositol (PI)

Ethanol utilization by A. niger ES4

To survive on the wall of an ethanol tank, it might be necessary for A. niger strain ES4 to be capable of metabolizing ethanol for nutrients or incorporating ethanol into other molecules to reduce its toxicity. Because our metabolomics analyses above revealed the upregulation of DAG, TAG, and hTAG in the ethanol-treated A. niger samples compared with the controls, we set out to trace the possible incorporation of ethanol or parts of ethanol into some of these lipids by using stable isotope labeling MS with ethanol-d₆.

The A. niger ES4 cultures were grown in MM in the presence of 4% (v/v) ethanol-d₆ (or nonlabeled ethanol as controls) for 3 days, harvested for metabolites in the mycelia, and analyzed by LC–MS exactly as described earlier for metabolomics. The chromatograms were then inspected manually for the mass spectra of four representative metabolite ions: (a) DAG (18:2/18:2), (b) TAG (18:0/18:1/18:1) (significantly elevated under ethanol treatment compared to the untreated controls), (c) fatty acid (FA) (18:2), and (d) phosphatidic acid (PA) (18:2/18:2) (in the related metabolic pathways but with unchanged levels). The data showed varying shift in the detected mass-to-charge ratios (m/z) of all lipids in the ethanol-d₆ samples from the nonlabeled controls (Figure ). The monoisotopic peaks of all lipids, except for FA (18:2), in the nonlabeled samples were no longer dominant peaks in the ethanol-d₆ samples. The m/z with the highest intensities were shifted by +3, +5, and +2 mass units for DAG (18:2/18:2), TAG (18:0/18:1/18:1), and PA (18:2/18:2) from those in the nonlabeled samples, respectively. However, the m/z peaks at ±1, ±2, and ±3 mass units from the dominant m/z peaks were not very different in intensities from those of the dominant m/z peaks for all investigated lipids in the ethanol-d₆ samples. Overall, the results support the metabolism of ethanol in A. niger ES4 into other metabolites.

Mass spectra of representative ions from stable isotope labeling MS experiments with ethanol-d6. The A. niger ES4 was cultured in duplicate in the presence of 4% (v/v) either ethanol (EtOH)-d6 or EtOH (as control) and analyzed for intracellular metabolites exactly as conducted in the untargeted metabolomics analysis. Mass spectra were extracted from the total ion chromatograms at the retention time (RT) of 43.8 min for (a) diacylglycerol (DAG) (18:2/18:2) and 49.0 min for (b) triacylglycerol (TAG) (18:0/18:1/18:1) in the positive ion mode, and at a RT of 18.5 min for (c) fatty acid (FA) (18:2) and 26.8 min for (d) phosphatidic acid (PA) (18:2/18:2) in the negative ion mode

DISCUSSION

The ability to tolerate chemicals present in cultures is one of the most essential traits of microorganisms for utilization in bioprocesses. In this study, we found A. niger ES4 capable of growing even in 5% (v/v) (or 4% [w/v]) ethanol with its growth amounting to approximately 30% of that of the ethanol-free control. Interestingly, this level of tolerance was comparable with that of the ethalogenic filamentous fungus Fusarium oxysporum (Paschos, Xiros, & Christakopoulos, ) and higher than that of the natural ethanol-producing ascomycetous yeasts Pichia stipites, whose growth was inhibited at 3.4% (w/v) ethanol when grown on glucose (Meyrial, Delgenes, Romieu, Moletta, & Gounot, ). Assuming that the amount of ethanol produced by these ethalogenic microbes themselves stayed relatively low compared with that of the initially added ethanol concentration, our results would indicate that A. niger strain ES4 had a relatively high resistance toward ethanol. However, as mentioned earlier, Dantigny et al. () also demonstrated the ability of the A. niger strain isolated from spoiled pastry products to grow on potato dextrose agar containing ethanol up to 4% (w/w) with about 50% reduction in its growth, which seemed to be higher than ethanol tolerance of A. niger ES4 in this study. Nevertheless, it was known that the choice of culture media and conditions could affect ethanol tolerance of microbes greatly, and thus, how tolerance A. niger ES4 is compared to other A. niger strains or other organisms remained to be proven in the future study.

The subsequent investigation into the response mechanisms of A. niger ES4 to ethanol using an untargeted metabolomics approach revealed the accumulation of neutral glycerolipids (extracellular TAG, and intracellular DAG, TAG, and hTAG) under ethanol stress. DAG and TAG are interconnected in the glycerolipid metabolism (Figure ). Functionally, DAG is known to play multiple roles from being a component of the cell membrane and an intermediate in lipid metabolism to a second messenger in lipid-mediated signaling cascades (Carrasco & Mérida, ), whereas TAG is traditionally thought of as an energy storage lipid.

In terms of responses to stress, even though there have been no prior reports demonstrating the effect of upregulated DAG or TAG in the organic solvent tolerance of microbes, these neutral lipids were previously implicated in protecting plants from other abiotic stressors, such as coldness (Tan et al., ) and darkness (Fan, Yu, & Xu, ). In the case of coldness, for example, Arabidopsis accumulated PA, DAG, and TAG during freezing stress where disruption in the genes encoding the enzymes acyl-coenzyme A: DAG acyltransferase and DAG kinase, which catalyze the conversion of DAG to TAG and PA, subsequently resulted in decreased and increased TAG levels and tolerance to coldness, respectively (Tan et al., ). It is, therefore, possible that, with the unchanged levels of other phospholipids that are components of cell membranes (e.g., PA, phosphatidylethanolamine (PE), and phosphatidylcholine (PC) [Ianutsevich, Danilova, Groza, & Tereshina, ]) found in this study, the elevated DAG and TAG levels as the novel responses to ethanol stress might play some roles in defending A. niger against the toxic effects of ethanol. Yet, because cold stress rigidifies cell membrane, the effect that is opposite to that of ethanol, further genetic data are needed to pinpoint the relevance of these glycerolipids on ethanol tolerance of A. niger. In addition, it remains to be determined whether the upregulation in glycerolipids following ethanol exposure is specific to A. niger ES4 or general for all A. niger isolates.

Another class of metabolites elevated under ethanol stress was hTAGs, which have mainly been described in Ricinus communis castor oil (Kim et al., ), Lesquerella seed oil (Byrdwell & Neff, ; Hayes, Kleiman, & Phillips, ), and ergot oil from the fungus Claviceps purpurea (Morris & Hall, ). As the detection and quantitation of hTAG in routine work have remained quite limited compared with other classes of lipids, there is no evidence to support or refute their relevance to various stresses. Yet, because their substrates, hydroxy fatty acids, are known for their specialized medical and industrial usages (e.g. lubricants, paints, and coatings) (Hayes et al., ; Meesapyodsuk & Qiu, ), the discovery of upregulated hTAGs in this study suggested that, upon fully characterizing their structures, A. niger might be able to serve as another source of this industrially important class of lipids as well.

One of the known metabolic responses of the yeast Saccharomyces cerevisiae toward ethanol involves an increase in the unsaturated-to-saturated fatty acids ratio, whereby the relative contents of FA (16:1) and FA (18:1) increase while those of FA (16:0) and FA (18:0) decrease (Sajbidor, Ciesarova, & Smogrovicova, ). Interestingly, our metabolomics studies showed only slight elevation in contents of all four most abundant fatty acids found in A. niger ES4 (i.e., 16:0, 18:2, 18:1 and 18:0) under ethanol stress without an obvious shift in unsaturation index. The alteration in this index was also not immediately apparent in other lipid species, as most of the changing lipids contained both unsaturated and saturated acyl chains. The findings therefore suggested that A. niger might utilize different mechanisms to counteract toxicity of ethanol than S. cerevisiae. However, we could not rule out the possibilities that the observed effects might simply be specific to the choices of microbial strain, growth medium, or conditions depicted in this study.

The results from the stable isotope labeling MS showed that the isotopic patterns of all the molecular ions of representative lipids in the ethanol-d₆ samples differed from those in the nonlabeled ethanol samples. Because the only isotope present at an unnaturally high abundance in this case was deuterium, the finding indicated the incorporation of varying numbers of deuterium atoms into each lipid, which could potentially occur both via catabolism of ethanol-d₆ by A. niger and via deuterium exchanges with the deuteron on the hydroxyl group of ethanol-d₆ during the lipid biosynthesis. However, in this study, ethanol-d₆ was added into the cultures at 4% (v/v), which resulted in the mole ratio of proton to deuteron in the cultures being 100 to 0.58 (see Appendix 2 for details on calculation). Assuming that all hydrogen atoms on each representative ion have equal chances to undergo deuterium exchanges, and disregarding any kinetics isotope effects of deuterium, the predicted isotopic ratios, M:M + 1:M + 2:M + 3:M + 4, when considering only deuterium exchanges and natural isotopic abundance of each element would be equal to 100:84:36:10:2 for DAG (18:2/18:2), 80:100:63:27:9 for TAG (18:0/18:1/18:1), 100:38:7:1:<1 for FA (18:2), and 100:82:35:10:2 for PA (18:2/18:2) (Appendix 2). These ratios represent the highest probable signals of each isotope arisen from deuterium exchanges; yet, they alone still could not account for the observed high abundances of these isotopes, especially with M + 2 and higher m/z species, in the ethanol-d₆ samples (Figure ). The finding therefore suggested to us that A. niger ES4 was in fact capable of metabolizing ethanol-d₆ into other compounds in an existing metabolic pathway.

Metabolically, the ethanol utilization pathway has been well studied in the closely related fungus, Aspergillus nidulans (Felenbok, Flipphi, & Nikolaev, ). In this pathway, ethanol is first oxidized to acetaldehyde by alcohol dehydrogenase I. Further oxidation of acetaldehyde by alcohol dehydrogenase then yields acetate, which subsequently is converted to acetyl CoA by acetyl CoA synthetase (Figure ). In the form of acetyl CoA, these carbon and hydrogen atoms from ethanol can then enter into many metabolic pathways, along with the acetyl CoA synthesized from other carbon sources, such as glucose. For lipid synthesis, acetyl CoA is carboxylated to malonyl CoA and coupled with this product to yield acyl CoA, which is the substrate for production of fatty acids, phospholipids, and glycerolipids (Figure ) (Kanehisa & Goto, ). For A. niger, the genes encoding several homologs of alcohol dehydrogenases have been annotated in the genome of A. niger strain CBS513.88 (Pel et al., ). However, their activities in culture have not been confirmed. Our present data, therefore, represent the first piece of evidence to support the existence of this ethanol utilization pathway in A. niger ES4.

CONCLUSIONS

In total, by applying untargeted metabolomics to study the extracellular and intracellular hydrophobic components of the A. niger strain ES4 isolated from the wall of an ethanol tank, we demonstrated the upregulation of glycerolipids (i.e., DAG, TAG and hTAG) as novel responses of microbes to ethanol stress. The subsequent stable isotope labeling MS with ethanol-d₆ also supported the utilization of ethanol by A. niger ES4. Future work will aim to determine the relevance of these upregulated changes in glycerolipid metabolism and the ethanol utilization pathway in the ethanol tolerance of A. niger, as well as to elucidate the structures, biosynthesis, and functions of hTAG more thoroughly. More generally, we believe that untargeted metabolomics platforms and the overall approaches presented in this work will be powerful tools for the discovery of more novel responses of microbes to organic solvent stress, as well as to other external stimuli, in the future.

ACKNOWLEDGEMENTS

We thank Dr. Robert Douglas John Butcher (Research Clinic Unit, Office of Research Affairs, Chulalongkorn University) for editorial help in preparing this manuscript. This research was supported by the Office of the Higher Education Commission and the Thailand Research Fund (MRG6180002). The opinions expressed in this paper are the sole responsibility of the authors and do not necessarily reflect those of the funding agencies or Chulalongkorn University.

CONFLICT OF INTEREST

None declared.

AUTHOR CONTRIBUTIONS

Nawaporn Vinayavekhin coordinated the experiments, analyzed the data, drafted, and revised the manuscript. Wimonsiri Kongchai conducted the experiments and analyzed the data. Jittra Piapukiew and Warinthorn Chavasiri discussed the experiments and revised the manuscript. All authors read and approved the final version of the manuscript.

ETHICAL APPROVAL

None required.

DATA AVAILABILITY STATEMENT

The nucleotide sequence data were deposited into the DDBJ/EMBL/GenBank nucleotide sequence databases with accession number MK621333. Other data generated or analyzed during this study are included in this published article and its Supporting Information file.

1 APPENDIX 1: Identified positive-mode ions with statistically significantly elevated levels in ethanol-treated extracellular A. niger samples compared to the untreated control showing the mass-to-charge ratio (m/z), retention time (RT) and (a) potential identification and MS/MS spectrum, (b) integrated mass ion intensity (MSII) and (c) adjusted mass ion intensity (aMSII). The MSII and aMSII data are shown for three A. niger samples without (Con-1–3) or with ethanol treatment (EtOH-1–3) and their respective averages (Con-avg and EtOH-avg, respectively)

[Table omitted. See PDF]

A2: Identified positive-mode ions with statistically significantly elevated levels in ethanol-treated intracellular A. niger samples compared to the untreated control showing the mass-to-charge ratio (m/z), retention time (RT) and (a) potential identification and MS/MS spectrum, (b) integrated mass ion intensity (MSII) and (c) adjusted mass ion intensity (aMSII). The MSII and aMSII data are shown for three A. niger samples without (Con-1–3) or with ethanol treatments (EtOH-1–3) and their respective averages (Con-avg and EtOH-avg, respectively)

[Table omitted. See PDF]

A3: Other lipids in the pathways found in intracellular A. niger samples. Lipids are shown in terms of their lipid class and acyl chain as (a) detected ion adduct, measured mass-to-charge ratio (m/z), retention time (RT), MS/MS spectrum, (b) integrated mass ion intensity (MSII) and (c) adjusted mass ion intensity (aMSII). The MSII and aMSII data are shown for three A. niger samples without (Con-1–3) or with ethanol treatments (EtOH-1–3) and their respective averages (Con-avg and EtOH-avg, respectively)

[Table omitted. See PDF]

2: APPENDIX

Calculation and predicted isotopic patterns on the mass spectra of representative ions in the ethanol-d₆ samples when taking into account only deuterium exchanges of all hydrogens on the representative ions and natural abundances of elements.

Base calculation

As ethanol-d₆ (or CD₃CD₂OD) was added to the culture at the concentration of 4% (v/v), the medium had the following ratios of H₂O to CD₃CD₂OD:

By volume	96: 4
By mass (At 25°C, density of H2O = 0.99707 g/ml and of ethanol = 0.78522 g/ml)	95.7: 3.1
By mole (MW of H2O = 18.01528 g/mol and of CD3CD2OD = 52.10541 g/mol)	5.3: 0.060

Since H₂O has 2 protons, while CD₃CD₂OD has 1 deuteron, the mole ratio of proton (and some deuteron from water) to deuteron (from ethanol-d₆) is equal to 10.6: 0.060. However, considering that natural abundances of hydrogen and deuterium are 99.9844% and 0.0156%, respectively, the mole ratio of proton to deuteron then becomes 10.6: 0.062, which is equal to 100: 0.58.

(a) Diacylglycerol (DAG) (18:2/18:2)

Calculation

[M + NH₄]⁺ has the molecular formula of C₃₉H₇₂NO₅⁺.

The highest probability of each isotope when considering only deuterium exchanges and natural isotopic abundance of each element is then equal to the following:

(Note: Terms with relatively small contribution are ignored below.)[Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF]

Thus, M: M + 1: M + 2: M + 3: M + 4 = 100: 84: 36: 10: 2.

(b) Triacylglycerol (TAG) (18:0/18:1/18:1)

Calculation

[M + NH₄]⁺ has the molecular formula of C₅₇H₁₁₀NO₆⁺.

The highest probability of each isotope when considering only deuterium exchanges and natural isotopic abundance of each element is then equal to the following:

(Note: Terms with relatively small contribution are ignored below.)[Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF]

Thus, M: M + 1: M + 2: M + 3: M + 4 = 80: 100: 63: 27: 9.

(c) Fatty acid (FA) (18:2)

Calculation

[M – H]⁻ has the molecular formula of C₁₈H₃₁O₂⁻.

The highest probability of each isotope when considering only deuterium exchanges and natural isotopic abundance of each element is then equal to the following:

(Note: Terms with relatively small contribution are ignored below.)[Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF]

Thus, M: M + 1: M + 2: M + 3 = 100: 38: 7: 1.

(d) Phosphatidic acid (PA) (18:2/18:2)

Calculation

[M – H]⁻ has the molecular formula of C₃₉H₆₈O₈P⁻.

The highest probability of each isotope when considering only deuterium exchanges and natural isotopic abundance of each element is then equal to the following:

(Note: Terms with relatively small contribution are ignored below.)[Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF]

Thus, M: M + 1: M + 2: M + 3: M + 4 = 100: 82: 35: 10: 2.