Trehalose (α-D-glucopyranosyl-α-D-glucopyranoside) is a non-reducing disaccharide that is present in a wide variety of organisms. It has been isolated from plants, fungi, nematodes and insects [1–3]. In fungi, trehalose has been shown to accumulate in dispersal and survival structures such as spores (where it can constitute as much as 10% of the dry weight), sclerotia, and in yeast cells going into stationary phase [3, 4] . Since the sugar is rapidly degraded when these structures germinate or resume vegetative growth, early research concluded that trehalose serves as a storage molecule [5, 6]. However, later studies showed that the function of trehalose is more complex and diverse than just serving as an energy reserve; the molecule has been shown to function as a regulator of carbon metabolism [1], a signaling molecule and a protection molecule against various kinds of abiotic stress [3, 7]. Several fungal species have been shown to induce trehalose production as a stress response. Examples include: Saccharomyces cerevisiae[8, 9], Zygosaccharomyces bailii[10], A. nidulans[11], A. fumigatus[12], Rhizopus oryzae[13], and Botrytis cinerea[14]. Trehalose is known to protect both proteins and lipid membranes of living cells against stressors such as heat, desiccation and cold. Although the mode of bio-protection of trehalose is not fully elucidated, three main hypotheses are generally accepted, and the true mechanism is likely a combination of these. The hypotheses include: water replacement (direct interaction of trehalose with the protected structure through hydrogen bonds); mechanical entrapment (glass formation of trehalose that creates a protective coating around the structure); preferential exclusion (bulk water is ordered around trehalose and is thereby separated from the bio-molecule, which then becomes more compact and stabilized) [15, 16]. The physico-chemical properties of trehalose that lie behind these hypotheses include several crystalline forms, a high glass transition temperature, and the stereochemistry of the sugar [7, 15].

In fungi, trehalose is synthesized via the intermediate trehalose-6-phosphate (T6P) and involves two enzymatic steps. First, T6P is formed from one glucose-6-phosphate and one UDP-glucose catalyzed by T6P-synthase (here called TPS). In the next step, the phosphate molecule is removed by trehalose-phosphate-phosphatase (here called TPP) yielding trehalose [1, 11]. The organism in which trehalose synthesis has been most thoroughly studied is S. cerevisiae. Here, four homologous gene products responsible for trehalose synthesis physically interact forming a “trehalose synthase complex”, which consists of one TPS (called Tps1), one TPP (called Tps2), and two other subunits, Tsl1 and Tps3, with proposed regulatory and stabilizing functions [6, 17–19]. In filamentous fungi, the gene products involved in trehalose synthesis are not as thoroughly investigated as in S. cerevisiae, but have been studied with respect to germination [20], plant pathology [21] and human pathology [12, 22].

Within Aspergilli, several individual gene products have been identified and characterized. In A. niger, two Tps1 orthologs, tpsA and tpsB, have been identified and characterized. At ambient temperature, the trehalose level of ΔtpsA mycelia was lowered compared to wild-type. In contrast to the constitutively expressed tpsA, the expression of tpsB was induced by thermal stress [23]. In the opportunistic human pathogen A. fumigatus, four Tps1 paralogs, tpsA – D have been identified [12]. When deleting these genes, the authors found that either tpsA or tpsB was sufficient to maintain normal trehalose levels, but if both genes were deleted, the resulting mutant strain was depleted of trehalose and showed slower germination rates as well as higher susceptibility to heat and oxidative stress compared to wild-type. Another notable finding was that this double mutant was hypervirulent in infected mice [12]. In A. nidulans, a Tps1 ortholog, tpsA, has been identified and deleted. In this mutant, trehalose was not accumulated, and in addition, the authors could conclude that in A. nidulans trehalose is important for resistance to continual exposure to sub-lethal stress but not to short exposure of lethal stress [11]. In contrast to S. cerevisiae, tps mutants in Aspergilli are able to utilize glucose as carbon source [11, 23, 24]. All identified Tps1 orthologs in Aspergilli are generally much shorter than the S. cerevisiae Tps1, around 500 amino acids compared to 1447.

Besides Tps1 orthologs, two Tps2 orthologs have been identified within the Aspergilli, one in A. nidulans[25] and one in A. fumigatus[22]: In both species they are designated orlA. The ΔorlA mutant of A. fumigatus had a pronounced phenotype with abolished asexual reproduction as well as decreased virulence. However, the phenotype could be restored to wild-type appearance by growing the mutant on media containing an osmotic stabilizer (sorbitol or glycerol). As also observed in A. nidulans, the A. fumigatus ΔorlA mutant strain contained wild-type levels of trehalose but the T6P levels were elevated [22, 25].

In this study we focused on trehalose synthesis in filamentous fungi, and more specifically, in Aspergillus niger. This is a common food spoilage mould as well as an industrially important organism, utilized for production of citric acid, for instance [26]. Six genes, tpsA (ANI_1_1406074), tpsB (ANI_1_1078064), tpsC (ANI_1_1216124), tppA (ANI_1_1432094), tppB (ANI_1_48114) and tppC (ANI_1_2070064) were identified to be involved in trehalose biosynthesis. Expression of these genes was studied during conidial outgrowth. In addition, we deleted these genes and characterized the mutants in terms of trehalose and T6P content, protein interactions, and stress survival coupled to situations often occurring in foodstuff.

In this project we have studied six genes with a putative role in trehalose synthesis in A. niger: tpsA, tpsB, tpsC, tppA, tppB and tppC. All six genes encode homologous proteins and no similar gene products within the A. niger genome could be detected. Three proteins, TpsA, TpsB and TpsC, have previously been identified as orthologs to the yeast protein Tps1. As the orthologs are conserved in related species, it is plausible that there is a functional differentiation between the paralogs, e.g. one paralog could be essential for trehalose synthesis in conidia, whereas another paralog is strictly induced by stress. This assumption is in line with the previous observation in A. niger where the expression of tpsB is stress-induced whereas tpsA is constitutively expressed [23], although our data also suggest that tpsB has a role during differentiation (see Figure 3). When deleting the trehalose-phosphate-synthase paralogs, only ΔtpsA displayed a reduced trehalose content. The lower level in this mutant is in line with a previous report using a different target strain and deletion procedure [23]. In the related fungus, A. fumigatus, a tpsA/tpsB double deletion resulted in a strain with depleted trehalose content, and in the same study, it was shown that the expressions of tpsC and –D were very low at all time points [12]. These authors evaluated their expression data using RNA extracted from hyphae, and in the present study, the A. niger tpsC was expressed at very low levels at 72 h. Thus the results from the two fungi are not contradictory, and most likely an A. niger tpsA/tpsB deletion mutant would also have a depleted trehalose content. The results from A. niger and A. fumigatus are also in accordance with findings in A. nidulans where deletion of tpsA resulted in depleted trehalose content [11], as that species does not have the tpsB paralogue. A conclusion from studying the trehalose content from these three species is that TpsA is the most important trehalose-phosphate-synthase under normal conditions, but lack of the tpsA gene can be fully compensated by TpsB in A. fumigatus and partly by at least one of TpsB or TpsC in A. niger, but not by TpsD in A. nidulans.

The deletion mutant with the most distinctive characteristics in our experiments was ΔtppA, i.e. with an abnormal morphology and reduced levels of both trehalose-6-phosphate and trehalose. The altered morphology of the strain is probably due to toxicity of T6P as indicated for the corresponding deletion mutant in A. fumigatus[22]. However, in A. niger as well as A. fumigatus and A. nidulans[12, 25], mutants of tppA are not totally lacking in trehalose. Therefore, it is possible that under specific conditions, e.g. when TppA is absent, TppB, and also TppC where present, may contribute to some T6P activity. Another possibility is that the sugar can be synthesized by proteins other than Tps/Tpp, e.g. the Trehalose Phosphorylase pathway, for which putative genes have been identified and partially characterized in N. crassa[40] and A. fumigatus[22] and also exist in A. niger (ANI_1_2720024). However, it is possible to generate mutants, within the homologous Tps/Tpp group, in A. fumigatus and A. nidulans that totally lack trehalose [11, 12]. Therefore, we believe that this is the only active trehalose synthesis pathway in Aspergilli. However, internal trehalose contents may not solely be dependent on the presence and expression of these six genes, as in S. cerevisiae there is a strong linkage between trehalose synthesis and the degrading trehalases [41] as well as evidences of posttranscriptional activation of the genes involved in trehalose metabolism [42, 43].

Besides a putative phosphatase activity, TppB and TppC may have similar biological roles as the yeast proteins Tps3 and Tsl1, which also contain phosphatase domains – in yeasts, deletion of both genes is necessary before some reduction in internal trehalose content can be observed [17]. It is intriguing that tpsB and tppC are linked on the chromosome. We cannot explain why the conidial trehalose content in this double mutant was significantly higher after 28 days, but based on the expression patterns (see Figure 3), it is possible that the expression of the two genes are regulated by the same factors. In addition to the above-mentioned observations, some conclusions can be drawn from the gene expression data: All identified genes were expressed, indicating that the paralogs are not inactive duplicates. For tpsC and tppB, the expressions were consistently low after 6 h, indicating that the two genes may be regulated by the same mechanism. This assumption is supported by a previous observation using A. oryzae arrays where the tpsC and tppB orthologs were down-regulated in a deletion strain of atfA, a gene encoding a transcription factor [44]. To our knowledge, two previous studies describing the expression of trehalose synthesis genes in A. niger during germination, using microarray technology, or in combination with RNA sequencing, have been published [29, 45]. With the exception that van Leeuwen and co-workers [29] saw a drastic drop after 2 h and then a gradual up-regulation of tpsA and tpsB, those results are in line with our findings.

The extensive measurements of internal trehalose indicate that the trehalose contents, for all strains, were low in 5 day old conidia, significantly elevated in 14 day old conidia, and then maintained at the value of 14 days (Figure 7). A plausible hypothesis is that conidia of A. niger reach full maturity, at least in terms of trehalose accumulation, sometime between 5 days and 2 weeks. Consequently it is not advisable to perform stress experiments on young conidia because their trehalose content is not necessarily typical for the final level, especially not in kusA deficient strains that seem to have slower conidial maturation in terms of trehalose content.

We found that 2 week old conidia of ΔtppB were more susceptible to heat shock than wild-type conidia, indicating that trehalose protects the spores from thermal stress. These results are in line with earlier studies in Aspergillus species [11, 12, 23]. However, in contrast to results from A. fumigatus and A. nidulans, we could not detect any increased sensitivity of ΔtppB to oxidative stress [11, 12], salt or acid stress, or any decreased viability after long term storage. It should be noted that unlike ΔtppB in our experiments, which harbored approximately one third of wild-type trehalose content, the A. fumigatus and A. nidulans mutants were totally depleted of trehalose.

In S. cerevisiae it has been shown that, using a two-hybrid assay, the four homologous proteins physically interact. When repeating the experiments using the six identified A. niger proteins, we could observe interactions for four of six proteins. These results suggest that TppA and TpsA-C form a complex, while the phylogenetically more distant proteins, TppB and TppC, are present outside the complex. However, due to the experimental limits, it is possible that neither TppB nor TppC was correctly folded and therefore not interacting. It is notable that in S. cerevisiae, a truncated version of Tsl1 was necessary for the success of the interaction experiments [40], in contrast to our experiment in which we only used full-length proteins.

To conclude, in this study novel information about the six gene products involved in trehalose synthesis in A. niger has been generated. When characterizing deletion mutants, lack of the most conserved trehalose phosphate synthase tpsA, the trehalose phosphate phosphatase tppA, or the previously non-characterized tppB, resulted in lower trehalose contents. An additional insight is that the components in a putative trehalose synthesis complex differ among the Aspergilli, but some gene products are common throughout the fungal kingdom.