Effects of cre1 modification in the white-rot fungus Pleurotus ostreatus PC9: altering substrate preference during biological pretreatment

Cellulose is the major component of plant cell walls that can potentially be converted to biofuel and other useful chemicals. Besides cellulose, the plant cell wall contains other polymers, such as hemicellulose and lignin, which encompass the cellulose and together creating a rigid, recalcitrant structure, which is difficult to hydrolyze enzymatically [1, 2]. In addition to its function as a mechanical barrier that restricts the accessibility of cellulolytic enzymes, lignin can also bind proteins in a nonspecific manner, leading to a decrease in the effective concentration of the cellulolytic enzymes available for the hydrolytic process [3–6]. Cellulose itself is an intractable molecule, containing crystalline regions, reinforced by inter- and intramolecular hydrogen bonds. To efficiently hydrolyze lignocellulosic biomass and to efficiently utilize the energy contained in cellulose, a pretreatment step is invariably required. The overall outcome of this pretreatment is the increase in digestibility of the available sugars. An optimal pretreatment strategy would enrich the cellulosic fraction, increase accessibility of the cellulose to the hydrolytic enzymes, remove or modify the lignin, and would be environmental friendly (i.e., minimized use of energy and chemicals along with reduced environmental wastes) [7, 8].

Numerous pretreatments, including physical (such as milling, steam explosion, and microwave), chemical (alkaline, acidic, etc.) or various combinations have been discussed in the literature [9]. Biological pretreatment, in which lignin is degraded or modified by microorganisms, has also been suggested. Biological pretreatment is considered to be environmentally friendly, since this approach exhibits low-energy demand/low carbon footprint and environmentally harmful chemicals are not used [9]. In contrast to the physical/chemical pretreatments used today, biological pretreatment does not require special facilities, since extreme heating or high pressures are not involved. Thus, biological pretreatment can also decrease the overall bioethanol production costs. Finally, unlike most other pretreatments, biological degradation of lignin is not accompanied by accumulation of compounds inhibitory towards yeast (such as furans), which may interfere with the downstream fermentation step [7, 8, 10–12]. On the other hand, biological pretreatment has some disadvantages, such as the requirement of long incubation periods and consumption of cellulose by the microorganisms during pretreatment [7, 10, 12, 13]. An efficient biological pretreatment strategy should thus result in a high ratio of lignin-to-cellulose degradation, with minimal cellulose loss.

Breakdown and consumption of lignocellulosic biomass by the fungus is mediated by a set of CArbohydrate-Active enZymes (CAZymes. Carbohydrate-Active Enzyme database, http://www.cazy.org/) [14]. The carbohydrates are decomposed by enzymes such as glycoside hydrolases (GHs), copper-dependent lytic polysaccharide monooxygenases (LPMOs), carbohydrate esterases (CEs), and pectate lyases (PLs), while the lignin modification is mediated by another set of enzymes, such as peroxidases, laccases, aryl alkyl oxidases (AAO), oxidoreductases, and copper radical oxidases (CRO). It is expected that altering the ratio between the lignin-modifying and the carbohydrate-decomposing enzyme activities in the secretome could influence the cellulolytic activities and utilization of cellulose by the fungus, and, consequently, the content of the residual lignocellulose biomass. The addition of metal salts (MnSO₄ and CuSO₄), phenolic and aromatic compounds (ferulic acid, xylidine, veratric acid, vanillic acid, cinnamic acid, and guaiacol) was found to increase selectivity by either increasing ligninolytic activity or by inhibiting cellulase activity [15–19]. Extracted natural cellulose inhibitor (NCI) from leaves of Vitis vinifera was also found to improve cellulose recovery and delignification, during biological pretreatment of paddy straw with P. florida [20].

Selecting a suitable microorganism for pretreatment is also a matter of consideration. A microorganism that demonstrates high ligninolytic capabilities but only minimal levels of cellulose consumption will be most appropriate for biotechnological applications. An extensive range of wood-decaying white-rot Basidiomycota (including P. ostreatus) were sequenced, enabling genomic, transcriptomic, and proteomic analysis of their plant cell wall-decaying enzymes [21–24]. Accumulating data show that there is no apparent genetic explanatory factor or phenotype pattern which would explain the preference for either lignin or cellulose degradation. In fact, based on comparative genomics of white-rot fungi, it was suggested that the functional diversity of these fungi could be better described as a continuum [25, 26] and could be also temporal (depend on growth kinetics) or related to the substrate chemical composition [21, 27]. A wide range of white-rot fungi, including Irpex lacteus, Pycnoporus cinnabarinus, Phanerochaete chrysosporium, Trametes versicolor, Ceriporiopsis subvermispora, P. eryngii, and P. ostreatus (alone or in consortia) were extensively studied for their potential for biological delignification [12, 28–36]. While evaluating the effect of rice straw pretreatment using several white-rot fungi, P. ostreatus was the most promising, demonstrating relatively selective removal of Klason lignin [35]. P. ostreatus was also found to be one of the preferred white-rot fungi for pretreatment of wheat straw, compared with 19 other tested white-rot fungi [36]. In a previous study conducted in our lab, we have genetically modified P. ostreatus PC9 to overexpress the ligninolytic enzyme versatile-peroxidase (VP1), independent of its natural Mn²⁺-dependent repression. The mutant displayed significantly higher levels of lignin mineralization, in vitro dry matter digestibility, and neutral detergent fiber digestibility [37].

Carbon catabolite repression (CCR) is a biochemical mechanism that down-regulates the synthesis of enzymes required for degradation and utilization of a carbon source when a preferred and easily metabolizable one (such as glucose) is available [38]. MIG1/CreA/Cre1 is zinc-finger transcription-factor homologs, conserved in most fungal species [38–45]. They were found to act as repressors in the CCR pathway in Saccharomyces cerevisiae as well as in some lower multicellular fungi, such as Aspergillus niger, A. nidulans, Trichoderma reesei, and Neurospora crassa. In the two latter species, CRE-1 was found to down-regulate cellulolytic genes and Δcre1 mutants exhibited increased cellulolytic activity. Transcriptome and secretome analyses of the Δcre1 mutants suggested that the regulation mechanism of CRE1 is complex, affecting both up- and down-regulation of genes and goes beyond CCR [43].

With the aim of altering the ratio between lignin and cellulose decomposition, the cre1 gene of P. ostreatus PC9 was either overexpressed or disrupted in the current study. The resultant cre1 overexpression (OEcre1) and cre1 deletion (KOcre1) demonstrated different levels of secreted-cellulase activities. By performing secrotome analyses, we identified 138 CAZymes secreted by the P. ostreatus PC9 and transformants, and determined the changes that occur in their expression following manipulation of cre1. In addition, the respective cell wall composition after pretreatment and the biological pretreatment efficiencies of the wild-type PC9 and both genetically modified transformants were investigated. Manipulation of cre1 expression altered the secreted CAZyme composition and the resultant pretreatment. Nevertheless, the parental PC9 demonstrated the highest pretreatment efficiency in terms of soluble sugar yield after pretreatment and hydrolysis.

Cellulose from lignocellulosic biomass is a promising resource for alternative energy (e.g., bioethanol), as well as for other biochemicals. Thus, the process of converting lignocellulosic biomass into soluble fermentable sugars is of great interest for the biotechnological industry. Lignin acts as a physical barrier that restricts the accessibility of the carbohydrate chains of the plant cell wall to the catalytic enzymes. Lignin is also known for its non-productive adsorption (and thus inhibition) of the hydrolytic enzymes [1–6]. Hence, a substrate pretreatment step is usually necessary to remove or displace lignin, to increase the efficacy of lignocellulosic biomass hydrolysis. White-rot fungi possess a promising potential for biological pretreatment, due to their unique ability to completely and efficiently break down lignin [12, 13, 36, 48–51]. Indeed, several white-rot fungi have been analyzed for their potential as lignocellulose decomposing agents and have been demonstrated to be able to reduce lignin content and increase hydrolysis efficiency [12, 28–36].

One of the major drawbacks of biological pretreatment is the consumption of cellulose by fungi, at the expense of glucose yield [7, 10, 12, 13]. A fungus that can degrade lignin with only minimal cellulose consumption would mitigate this shortcoming and increase the overall efficiency of bioethanol production, and thus can be defined as a selective [52, 53]. To this end, the potential of P. ostreatus PC9 and two mutants, modified in their CAZyme expression pattern to pretreat wheat straw, was investigated. As the lignocellulosic biomass is degraded outside the cells, the composition of the secreted enzymes, namely, the ratio between secreted lignin-modifying and carbohydrate-degrading enzymes, can affect the level of residual cellulose. The expression of these enzymes is regulated by CCR. CRE1, a conserved regulator in the CCR pathway, has been previously found to regulate cellulolytic genes in several ascomycetes, such as A. niger, N. crassa, and T. reesei [38–44]. In addition, homologues to CRE1 were found in 27 of the 31 basidiomycete genomes assessed [54]. Yet, no functional studies regarding the effect of CRE1 on the regulation of cell wall degrading enzyme was conducted in the basidiomycetes until now [55].

In view of the above, we hypothesized that a CRE1 homologue may also have a conserved role as a transcriptional regulator of carbon metabolism in the basidiomycete Pleurotus. Thus, two transformants were modified to either overexpress or eliminate expression of the cre1 gene (termed OEcre1 and KOcre1, respectively). Many studies have been conducted in fungi within the context of overproduction of cellulases, where the expression of CRE1 was eliminated or reduced [41, 43, 44, 56]. The purpose of the current study, however, was to explore the role of Pleurotus not as a cellulase producer but as a biological pretreatment agent. Consequently, unlike most previous studies, we examined the function of cre1 overexpression in addition to its deletion.

OEcre1 and KOcre1 exhibited decreased or increased levels of secreted cellulolytic activity, respectively, while grown on either soluble cellulosic substrate (CMC), insoluble cellulosic substrate (MCC) or on a lignocellulosic substrate (wheat straw), thereby demonstrating the role of cre1 repression in P. ostreatus (Fig. 1). Interestingly, OEcre1 was able to grow on cellulosic substrates despite overexpression of cre1, consistent with the previous observations, showing that basal expression levels of cellulolytic genes are not repressed by cre1 [43]. Similar levels of accumulated fungal biomass were observed in all three fungi, when grown on CMC. However, increased fungal biomass accumulation was observed when the KOcre1 was grown on a glucose-based medium. Likewise, the growth pattern reported for the Δcre1 mutant of N. crassa was similar to that of the wild type when grown on CMC but different when grown on glucose [39]. Glucose is known to regulate carbon catabolite enzymes via the cre1 repression pathway [57–60]. Thus, we expected that P. ostreatus that lacks cre1 would grow differently on glucose-based media. Since the differences in the secreted cellulolytic activity are not applicable in such glucose-containing media, our observation supports the contention that CRE1 regulates functions beyond CCR, including the rate of glucose assimilation, and can influence cell growth [43].

The degradation of lignocellulosic biomass by fungi is carried out via the collective activity of secreted enzymes. The secretome composition, in addition to cellulolytic activity levels, is, therefore, of great interest, since changes in secretome composition can influence the efficiency of biological pretreatment. Indeed, MS analysis of the secretome is a commonly used method for investigating the cell wall degrading machinery of fungi [61–65]. The availability of the genome sequence of the monokaryotic P. ostreatus [21] enables us to perform MS analysis of its secretome [48, 62, 66]. The wild-type and cre1-transformants were grown on MCC and wheat straw (as sole carbon sources) and their secretome compositions were analyzed. The three fungi revealed differences in the percentage of CAZymes within the secretome (Fig. 3a), as well as differences in the composition of CAZyme groups within each fungus (Fig. 3b). The secretome of OEcre1 grown on wheat straw exhibited a higher proportion of LMEs and lower abundance of carbohydrate-degrading enzymes (GHs, LPMOs, and PLs). In contrast, the secretome of KOcre1 contained less LMEs and a slightly higher proportion of carbohydrate-degrading enzymes. The changes in the compositions of the secretomes are consistent with the observed changes in the composition of the pretreated wheat straw, demonstrating increased/decreased ratios of cellulose-to-lignin content in the OEcre1/KOcre1-treated wheat straw (respectively, Fig. 5b). Our results thus highlight the effect of altering cre1 expression not only on the levels of secreted cellulolytic activity (an important parameter regarding cellulase production) but also on the potential to act as a biological pretreatment agent.

The effect of modifying cre1 expression on the individual CAZymes was further investigated (Fig. 4). An extended range of CAZYmes was affected by cre1 alteration: The relative abundance of approximately 70 and 55% of the detected CAZymes in the secretome of OEcre1 and KOcre1 (respectively) grown on MCC was either decreased or increased by at least one order of magnitude. While grown on wheat straw, 68 and 33% of the detected CAZymes in the secretome of OEcre1 and KOcre1 (respectively) was either decreased or increased. Interestingly, OEcre1 and KOcre1 did not demonstrate opposite regulatory response of CAZyme expression. Some proteins were less abundant in both transformants, regardless of whether cre1 was overexpressed or inactivated. Conversely, other proteins were highly expressed in both. Some were affected only in one of the transformant. These results can indicate an extensive and complicated regulatory pathway involving cre1 and/or the possible presence of overlapping/complementary regulatory pathways. Indeed, such complimentary regulation pathways of cellulose production have been reported for some ascomycetes. For example, cellulase expression in N. crassa was reported to be repressed by both CRE1-dependent CCR and by the glucose-sensing/metabolism transcription factor COL26, and a synergistic effect between the two regulation pathways was suggested [59]. In addition to repression, cellulase gene expression was also reported to be upregulated, inter alia by transcription factors CLR1 and CLR2 [67, 68]. Positive regulation of cellulase expression was also reported for other ascomycetes such as A. niger, T. reesei, A. aculeatus, and A. nidulans [67, 69, 70]. Moreover, the crosstalk between the two regulatory systems is currently not well understood [59]. Such complimentary regulation pathways can partially compensate for mutagenesis of cre1 and might provide a possible explanation for our observations.

CRE1 affected the various CAZymes differently. Namely, the relative abundance of some genes was increased, whereas the relative abundance of others was decreased in each of the transformants. These observations can indicate a role for CRE1 not just as a repressor, but as a regulator that can also increase expression. These results are in accord with the previous reports which identified genes that showed decreased expression levels in a Δcre1 mutant (namely positively regulated by CRE1) in both T. reesei and N. crassa [39, 43] (including genes that were annotated as members of carbohydrate metabolism pathways). In this context, Portnoy et al. [43] suggested that cre1 should be considered a master regulator (with up- and down-regulatory effects) rather than a repressor. It should be noted that OEcre1 was manipulated via non-homologous integration, which also might be a possible source of the differences between the wild-type PC9 and the OEcre1 transformant (e.g., neighboring effects).

The secretome of PC9 was previously analyzed when the fungus was grown on glucose, poplar wood, and wheat straw-containing media [66]. Our results highlight several similar trends in the secretome of PC9 grown on wheat straw samples, such as (i) the significant role of copper radical oxidases (CROs), rather than the aryl-alcohol oxidases (AAO) among the hydrogen peroxide-providing enzymes and (ii) the significant role of laccases [specifically lac10 (81117)], rather than the oxidoreductases (VP1 and MnP3) as the most abundance lignin-modifying enzymes.

Microorganisms are known to respond to available carbon sources and to change their secreted CAZymes accordingly [38, 42, 66, 71–74], e.g., via the CCR pathway [38, 39, 42, 43, 75]. Indeed, the secretome of PC9 has been found to change in response to different carbon sources [62, 66]. For example, the relative abundance of laccases, oxydoreductases, and peroxidases was shown to increase when grown on wheat straw compared to growth in a glucose-based medium [66]. By comparing the secretome of PC9 grown on MCC to wheat straw (regardless of altering cre1 expression), our results also demonstrate a similar influence of the carbon source. The relative portion of LMEs in the secretome of PC9 grown on wheat straw was higher than that of cells grown on MCC, demonstrating “differentiation” towards the wheat straw (where LMEs are required). Interestingly, the preference of laccases as the most abundant lignin-modifying enzymes was observed only in the wheat straw samples but not in the MCC samples where VP1 displayed higher relative abundance than laccases. Since CRE1 mediates regulation of CAZyme expression in response to the available carbon source, it logically followed that cre1 alteration would also be manifested differently when grown on different carbon sources. Indeed, the effect of modifying cre1 expression on several CAZymes was found to be carbon source-dependent. For example, the relative abundance of 115073 (GH3) was decreased only in the secretome of OEcre1 grown on wheat straw (where it was undetected) but not when grown on MCC (where no decrease was found). A similar trend was found for 52090 (GH5/Subf7). In general, it should be noted that only secreted soluble proteins were analyzed in the current work. Thus, the fact that some proteins were undetected (Fig. 4 and Additional file 3: Figure S3) does not necessarily mean that they were not expressed. The inability to detect a protein might also be related to substrate binding.

To investigate whether the changes in the secretome composition affect selectivity of lignin-to-cellulose degradation and the biological pretreatment efficiency of the mutants, the fungi were used to treat wheat straw. PC9 demonstrated the highest glucose yield after enzymatic hydrolysis (280%, compared to the untreated wheat straw). In comparison to PC9, KOcre1 exhibited increased CO₂ emission during pretreatment, a lower ratio of cellulose-to-lignin content and only a modest increase in the glucose yield compared to that of untreated wheat straw (121%). OEcre1 showed decreased CO₂ emission, which slightly increased the ratio of cellulose-to-lignin content and a glucose yield of 260% (slightly lower than that observed for PC9). It may thus be concluded that knockout of cre1 can reduce selectivity and pretreatment efficiency. Nevertheless, overexpression of cre1 can increase selectivity, but cannot improve the pretreatment efficiency.

Several fungi (alone or in a consortium), including P. ostreatus, were previously investigated for their biological pretreatment capabilities, demonstrating a wide range of lignin removal and increased hydrolysis yield upon treatment of woody and non-woody lignocellulosic biomass [12, 33, 35]. In the current work, the parental P. ostreatus PC9 was the most efficient, demonstrating 48.1% lignin removal and an increased hydrolysis yield of 2.8-fold after 35 days of pretreatment. Optimization process (time of growth, moisture percentage, additives in the medium, and other parameters) may increase pretreatment efficiency [34, 35]. Indeed, in a preliminary calibration step, two pretreatment periods (namely, 19 and 35 days) were compared. The latter (35 days) resulted in higher soluble sugar concentration (1.5- to 2-fold) after hydrolysis (Additional file 5: Figure S4), thus demonstrating the influence of pretreatment length on efficiency. For more efficient pretreatment using P. ostreatus, an extensive calibration process should be conducted.

Increased lignin degradation, accompanied by reduced cellulose decomposition, can mitigate a major disadvantage in the biological pretreatment process of lignocellulosic biomass—i.e., the consumption of cellulose by the fungus. In the current paper, the expression of the gene encoding the transcriptional regulator CRE1 of the white-rot fungus P. ostreatus PC9 was manipulated, changing the secreted cellulolytic activity levels, the secretome composition, the selectivity of lignin degradation, and the efficiency of biological pretreatment. Knockout of cre1 led to increased secreted cellulolytic activity, decreased selectivity of lignin degradation, and lower glucose yields. Overexpressing cre1 led to decreased secreted cellulolytic activity, slightly increased selectivity of lignin degradation, but failed to generate higher glucose yields compared to the wild-type. Thus, changing cre1 expression was found to be a suitable approach for altering substrate preference or for genetic engineering of overproduction of P. ostreatus cellulase, but not suitable for engineering a superior biological pretreatment organism. The regulation effect of the cre1 homolog on secreted CAZymes in the basidiomycete P. ostreatus provided only a partial answer to its role in fungal biology.