A major goal of this study was to discover the enzymes A. nidulans secretes to digest sorghum stover while growing under conditions mimicking its natural habitat. These findings should help us to design a mixture of enzymes that can convert sorghum stover into readily fermentable sugars in vitro.
From the visual development of a thick fungal mat, the appearance of numerous fungal hyphae in electron microscopic images of the cultures (Figure 1), and the accumulation of chitin during the first two days of culture (Figure 2) we conclude that A. nidulans grows vigorously on a semi-solid sorghum slurry. However, although A. nidulans maintains a certain level of healthy biomass over a prolonged period, the actual fungal growth seems to reach a plateau after two days. From day 3 onwards we observed high levels of autolytic enzymes in the secretome indicating autolysis of the fungus. Class V chitinase, ChiB, AN4871, belonging to family GH18, N-acetylglucosaminidase, NagA, AN1502, belonging to family GH20, beta-1,3-endoglucanases AN4825 and AN0472, from family GH55 and GH81 respectively, the alkaline protease PrtA, AN5558, metallo-protease PepJ, AN7962, and dipeptidase AN2572 have all been shown to be associated with autolysis [39–42]. All were found to be highly induced from day 3 onwards (Table 5). On the other hand, Catalase B, an enzyme associated with growing and developing hyphae was found in high abundance throughout the time course . Similarly enzymes associated with fungal cell wall synthesis and remodeling such as GPI-anchored cell wall organization protein Ecm33 (AN4390), and putative GPI anchored beta-1,3(4)-endoglucanase (AN2385)  were present in the secretome. Based on these findings, we propose that A. nidulans is maintaining growth throughout the time course but at the same time is recycling the older hyphae to support growth of hyphal tips and lateral branching. It is well established that lateral branching can be induced by fungal interaction with plants presumably leading to enhanced surface area and nutritional assimilation .
Powell et al.  studied the chemical composition of conventional grain sorghum, intermediate type sorghum and forage sorghum stover. They reported 31.3% cellulose content in grain type sorghum stover, 25.3% for intermediate sorghum stover, and 29.1% for forage type sorghum on a dry mass basis. The hemicellulose content was found to be 28.2% in grain sorghum stover, 21.7% in intermediate type sorghum stover, and 23.9% in forage type sorghum. The lignin content was between 7 to 7.3% for the three types of sorghum stover. In this study, the cellulose content was found to be 26.6% and the hemicellulose content was 22.3% of the total dry mass (Figure 5).
From the amount of xylanase and cellulase activities we found in the extracellular fluid collected from the culture on day 1 there should have been sufficient enzyme activity to break down all the xylan and cellulose in the sorghum in less than a day. However the concentration of soluble sugars (break-down products from the enzymatic hydrolysis), in the extracellular fluid was low (less than 1 mM), and less than 5% of the total xylan had been digested and seemingly none of the cellulose. This emphasizes the well-known recalcitrance of plant biomass to enzymatic degradation and suggests that the fungus is growing under carbon limited conditions. However, over time, the fungus did appear to be able to digest all classes of cell wall polysaccharides present and it did continue to maintain a certain level of healthy biomass.
We were interested in determining if the fungus attacked the most easily digestible polymers first and then moved on to using the more recalcitrant ones. It has been suggested that fungi first degrade the pectin in plant cell walls to make the hemicelluloses and cellulose more accessible . Sugars characteristic of pectin, i.e. rhamnose and galacturonic acid, were not detected in total hydrolysates of sorghum stover, although both of these sugars were found in the solubilized sugars rinsed from the cultures. This indicates the presence of a small amount of pectin in sorghum. Polygalacturonase activity was induced rapidly in the culture but xylanase and cellulase activity increased at the same time. Looking at the proteomics results one can deduce that almost all of the many enzymes involved in hemicellulose and cellulose degradation are induced in parallel. However in pectin degradation there is a steady decrease over time in exopolygalacturonases, one of the pectin methyl esterases, and one of the pectin lyases. In contrast there was a steep increase in other pectate lyases and rhamnogalacturonan hydrolases.
The recalcitrance of the sorghum or any other plant biomass to digestion is thought to arise from physical inaccessibility of the polymers to the enzymes designed to digest them. Two major factors are the incrustation of the polysaccharides with lignin and the crystalinity of the cellulose. We identified three classes of proteins which might play a key role in overcoming the recalcitrance of the sorghum to digestion.
There is strong evidence that some of the ferulic acids esterified to xylans crosslink to make diferulates which link xylans together thus making them less accessible to enzyme digestion. The ferulates can also be incorporated into lignin thus attaching xylan directly to lignin . Hydrolysis of the ferulate ester bonds is expected to decrease the recalcitrance of biomass to enzymatic hydrolysis [48–50]. Two feruloyl esterases were secreted by A. nidulans in high abundance throughout the time course.
Another enzyme in our cultures which might play a role in overcoming recalcitrance of the sorghum to digestion is cellobiose dehydrogenase (CDH). CDHs oxidize cellobiose, or other reducing sugars, and transfer the electrons to Fe+++, quiniones, radicals, or oxygen to make hydrogen peroxide . The combination of the resulting Fe++ and hydrogen peroxide can lead to the generation of hydroxyl radical which will attack polysaccharides and lignin. At least three putative sequences for CDHs are present in the A. nidulans genome. We only detected one form. Interestingly CDH was not present on day 1 and then increased gradually showing a spectrum count of 17 on day 3 up to 77 on day 14. No other studies have reported the secretion of cellobiose dehydrogenases by A. nidulans on solid state cultures. The presence of cellobiose dehydrogenase after day 3 may indicate its important role in the degradation of crosslinks between lignin, hemicelluloses and cellulose.
Whenever investigated, GH61 family endoglucanases lack measurable endoglucanase activity, but recently they have been reported to accelerate the hydrolysis of cellulose in cellulase preparations [52, 53]. GH61 was initially described as an endoglucanase, but more recently has been shown to be members of the family are copper mono-oxygenases that catalyse cleavage of cellulose oxidatively, releasing cellodextrins [53, 54]. Their active site contains a type II copper site, which, after being reduced to Cu+ by a reductant such as ascorbate or gallate, is thought to activate oxygen which can then oxidize glycosidic linkages on the surface of crystalline cellulose. This generates new chain ends rendering the substrate far more prone to attack by the classical endoglucanases and cellobiohydrolases [52, 53]. Cellobiose dehydrogenase appears to be able to mediate the reduction of the copper site . Recent studies have shown that cellobiose dehydrogenase may enhance cellulose degradation by coupling the oxidation of cellobiose to the reductive activation of copper-dependent polysaccharide monooxygenases [53, 54]. Four GH61 enzymes were found in our study.
Studies similar to ours were conducted by Schneider et al.  who grew A. nidulans and the mesophilic bacteria Pectobacterium carotovorum on leaf litter for 20 days, both individually and in co-culture, and identified the proteins secreted into the medium by 1-D PAGE-LC-MS/MS. The aim of their study was to determine the relative contribution of the fungus and the bacterium to the decomposition. They reported a total of 90 proteins secreted by A. nidulans on leaf litter during the course of 20 days in comparison with the 294 proteins identified in our sorghum cultures. Seventy- two proteins were found to be common in both sorghum and leaf litter cultures, 59 proteins were exclusively found in the sorghum cultures and 13 were only identified in leaf litter cultures. The identities of the proteins in the two cultures are reported in Additional file 2: Table S2. More than half of the cellulases and xylanases were common to both cultures with two xylanases and four endoglucanases unique to leaf litter cultures. Only one pectinase out of 22 was unique to leaf litter. Interestingly enzymes involved in degradation of crosslinks between lignin, cellulose and hemicelluloses such as cellobiose dehydrogenase and feruloyl esterases were only identified in sorghum cultures. Only half the proteases and less than one third of cell wall remodeling enzymes were common to both the cultures. In another recent study, Couturier et al.  identified total 66 proteins in the secretome of A. nidulans grown on maize bran. Out of 19 GHs identified in their study in A. nidulans secretome, seven hemicellulases belonging to family GH10, GH11, GH39, GH43, GH62 and GH93 and five beta-1,3-glucanases from the GH17, GH55 and GH81 families were identified. They could not detect any beta-1,4-endoglucanase or cellobiohydrolase or any enzyme activity on CMC . The difference in growth and enzyme activities of A. nidulans on sorghum and on leaf litter may be attributed to different substrates compositions and growth conditions .