The transcriptional activator GaaR of Aspergillus niger is required for release and utilization of d‐galacturonic acid from pectin

Abstract

We identified the d‐galacturonic acid (GA)‐responsive transcriptional activator GaaR of the saprotrophic fungus, Aspergillus niger, which was found to be essential for growth on GA and polygalacturonic acid (PGA). Growth of the ΔgaaR strain was reduced on complex pectins. Genome‐wide expression analysis showed that GaaR is required for the expression of genes necessary to release GA from PGA and more complex pectins, to transport GA into the cell, and to induce the GA catabolic pathway. Residual growth of ΔgaaR on complex pectins is likely due to the expression of pectinases acting on rhamnogalacturonan and subsequent metabolism of the monosaccharides other than GA.

Abbreviations

AP, apple pectin

CM, complete medium

CP, citrus pectin

GA, d‐galacturonic acid

MM, minimal medium

PGA, polygalacturonic acid

RG, rhamnogalacturonan

SBP, sugar beet pectin

TF, transcription factor

XGA, xylogalacturonan

Pectins are complex heterogeneous polysaccharides found in plant cell walls. Four substructures of pectin have been identified, and include polygalacturonic acid (PGA) also known as homogalacturonan, xylogalacturonan (XGA), rhamnogalacturonan I (RG‐I), and rhamnogalacturonan II (RG‐II) 1. The backbones of PGA, XGA, and RG‐II are made up of α‐1,4‐linked d‐galacturonic acid (GA) residues. PGA, a linear polymer of GA, is the most abundant polysaccharide present in pectin 1. In XGA, β‐d‐xylose residues are β‐1,3‐linked to GA residues of the PGA backbone. The backbone of RG‐I is made up of alternating GA and l‐rhamnose residues 1, 2. Side chains of RG‐II contain at least 12 different types of monosaccharides, whereas the side chains of RG‐I are mainly arabinan and arabinogalactan comprising of l‐arabinose and d‐galactose residues 1.

In nature, pectin is an important carbon source for many saprotrophic fungi such as Aspergillus niger. Previous studies demonstrated that A. niger can produce more pectin‐degrading enzymes than other more specialized fungi such as Podospora anserina or Neurospora crassa 3, 4, 5. GA is the main product of pectin degradation. In A. niger, GA is transported into the cell by a GA‐induced sugar transporter named GatA 6. GA is then catabolized into pyruvate and glycerol 7, through a pathway consisting of four enzymes: GaaA, d‐galacturonate reductase; GaaB, l‐galactonate dehydratase; GaaC, 2‐keto‐3‐deoxy‐l‐galactonate aldolase; and GaaD, l‐glyceraldehyde reductase 7. Deletion of gaaA, gaaB, or gaaC abolished growth on GA as the sole carbon source 8, 9, 10. gaaD, also known as the l‐arabinose reductase gene, larA, is involved in the l‐arabinose catabolic pathway, and the ΔlarA strain showed a reduced growth on l‐arabinose as the sole carbon source 11.

The production of extra‐ and intracellular enzymes in A. niger is regulated by a network of transcription factors (TFs) 12. Small sugar molecules (mono‐ and disaccharides) act as inducers and stimulate TFs which can bind to conserved motifs in the promoters of their target genes and activate or repress their expression. Expression of pectinase genes is highly controlled and depends on both induction and carbon catabolite repression 13, 14. Induction of the genes required for pectin degradation, GA transport, and GA catabolism requires the presence of GA, and it has been shown that GA or a derivative of GA induces the expression of pectinase genes 9, 10, 13.

Coordination of the induction of genes encoding extracellular enzymes and sugar uptake systems in fungi are often mediated by Zn₂Cys₆ TFs that bind to conserved promoter elements in the coregulated genes 12, 15, 16. TFs inducing the genes required for the utilization of l‐rhamnose (RhaR), arabinan/l‐arabinose (AraR), xylan/d‐xylose (XlnR), d‐galactose (GalX), and cellulose (XlnR, ClrA, and ClrB) have been identified in A. niger 17, 18, 19, 20, 21. Although l‐rhamnose, l‐arabinose, d‐xylose, and d‐galactose are also present in complex pectins, knock out mutants in these TFs display no signs of reduced growth on pectin 17, 18, 20, suggesting that the utilization of GA, the main component of this substrate, is not affected.

Martens‐Uzunova and Schaap 7 have previously identified a set of GA‐induced genes in A. niger, containing several pectinases (pgaX, pgxA, pgxB, pgxC, paeA, pelA, and abfC), sugar transporter‐encoding genes (gatA, An03g01620, and An07g00780) and the GA catabolic pathway genes (gaaA‐D). These genes were suggested as the GA‐regulon and contain a common GA‐responsive element (GARE) in their promoter regions. The consensus element was defined as CCNCCAA 7. Deletion and mutational analysis of GARE showed that the element is required for GA‐induced gene expression in both A. niger and Botrytis cinerea 14, 22. A yeast one‐hybrid study using a GA‐responsive promoter in B. cinerea recently identified a novel Zn₂Cys₆ TF (BcGaaR) required for GA utilization 22. In this study, the GA‐responsive transcriptional activator, GaaR, of A. niger was identified by homology to BcGaaR. Deletion analysis and transcriptomic profiling studies performed in this study showed that the A. niger GaaR ortholog is required for growth on GA and PGA and for the induction of the GA‐regulon when grown on sugar beet pectin (SBP).

Materials and methods

Strains, media, and growth conditions

A. niger strains MA234.1 (cspA1, kusA::DR‐amdS‐DR) and N593.20 (cspA1, pyrG ⁻ , kusA::amdS) were used to create the ΔgaaR strains. N593.20 was made by transformation of N593 23 with a deletion construct (kusA::amdS) 24 resulting in the deletion of kusA. Strain FP‐1132.1 (cspA1, pyrG ⁻ ::AOpyrG, kusA::amdS) was obtained by transformation of N593.20 with pyrG from Aspergillus oryzae. MA234.1 was obtained by transformation of MA169.4 (kusA ⁻ , pyrG ⁻) 25 with a 3.8 kb XbaI fragment containing the A. niger pyrG gene, resulting in the full restoration of the pyrG locus.

Complementation studies were performed with JN35.1 (cspA1, kusA::DR‐amdS‐DR, gaaR::hygB). To restore functionality of the kusA gene to allow ectopic integration of the complementing fragment, the amdS marker was looped out of JN35.1 by FAA counterselection as described 26 to give JN36.1. The gaaR‐complemented strain, JN37.4, was created using JN36.1, by transformation of the gaaR gene including promoter and terminator regions (see below). All strains used are listed in Table S1.

Media were prepared as described 26. For growth phenotype analyses, strains were grown on minimal medium (MM) with 1.5% (w/v) agar and various sole carbon sources: 25 or 50 mm glucose (VWR International, Amsterdam, the Netherlands), GA (Chemodex, St Gallen, Switzerland), l‐rhamnose (Fluka, Zwijndrecht, the Netherlands), l‐arabinose (Sigma‐Aldrich, Zwijndrecht, the Netherlands), or d‐xylose (Merck, Amsterdam, the Netherlands), and 1% (w/v) PGA (Sigma, Zwijndrecht, the Netherlands), SBP (Pectin Betapec RU301 Herbstreith & Fox KG, Neuenbürg, Germany), citrus pectin (CP) (Acros Organics, Geel, Belgium), or apple pectin (AP) (Pectin Classic AU2022 Herbstreith & Fox KG). pH was adjusted to 5.8 with NaOH or HCl buffer. The plates were inoculated with 2 μL 0.9 % (w/v) NaCl solution containing 1000 freshly harvested spores and cultivated at 30 °C for 4 days. For gene expression analyses, freshly harvested spores were inoculated with a final concentration of 10⁶ spores·mL⁻¹ in 100‐mL complete medium (CM) (pH 5.8) with 2% (w/v) d‐fructose (Sigma‐Aldrich) and were pregrown for 16 h. For northern blot analysis, mycelium was harvested by filtration through sterile myracloth, washed twice with MM with no carbon sources (pH 4.5), and 1.5 g (wet weight) mycelium was transferred and grown in 50 mL MM (pH 4.5) with 50 mm GA or 50 mm d‐fructose for 2, 4, and 6 h. For RNA‐seq analysis, 2.5 g of pregrown mycelia were transferred to 50 mL MM (pH 4.5) with 25 mm GA and incubated for 2 h or to 50 mL MM with 1% SBP and incubated for 2, 8, or 24 h. All incubations were performed in rotary shaker at 30 °C and 250 r.p.m.

Construction of gene deletion and complementation strains

Protoplast‐mediated transformation of A. niger, purification of the transformants, and genomic DNA extraction were performed as described 26. To construct the deletion cassettes, 5′ and 3′ flanks of the gaaR gene were PCR‐amplified using the primer pairs listed in Table S2 and N402 genomic DNA as template. To create JN35.1 strain, the split marker fragments with hygB selection were created using fusion PCR 27 and transformed to MA234.1. To create FP‐1126.1 strain, the flanking regions were fused with a fragment containing the A. oryzae pyrG gene using GoTaq^® Long polymerase (Promega, Leiden, the Netherlands) and transformed into N593.20 strain. Parental strains and gaaR deletion mutants were deposited at the Centraal Bureau Schimmelcultures (CBS) under accession numbers indicated in Table S1. To complement the gaaR gene, the gaaR gene together with its 5′ and 3′ flanks was PCR‐amplified using the primer pairs listed in Table S2, ligated into pJET1.2/blunt cloning vector (Fermentas, Landsmeer, the Netherlands), amplified in the E. coli strain DH5α and transformed in to strain JN36.1 together with plasmid pMA357. pMA357 contains the A. nidulans amdS gene, cloned behind the A. nidulans gdpA promoter (Mark Arentshorst, unpublished vector). Deletion and complementation of gaaR were confirmed via Southern blot analysis or diagnostic PCR.

Gene expression analysis

For northern blot analysis, strains MA234.1 (reference strain) and JN35.1 (ΔgaaR) were pregrown in CM with d‐fructose. At the time of transfer (t = 0) and 2, 4, and 6 h after the transfer to MM with GA or d‐fructose, mycelium was harvested from cultures by filtration through sterile myracloth and frozen immediately in liquid nitrogen. Mycelium samples were stored at −80 °C. Total RNA was extracted from frozen mycelium samples after grinding in liquid nitrogen, using NucleoSpin RNA Kit (Macherey‐Nagel, Düren, Germany) following the protocol provided by the supplier, including the rDNase treatment. Total RNA samples were stored at −80 °C. Quantitation and purity assessment of total RNA was done by spectrophotometric method (NanoDrop 2000; Thermo Scientific, Breda, the Netherlands). Standard molecular techniques were applied as described 28. About 3.5 μg RNA was loaded per sample and hybridized with [α‐32P]‐dCTP labeled probes after blotting (DecaLabel DNA Labelling Kit; Thermo Scientific). Probes were PCR‐amplified using the N402 genomic DNA and the primer pairs are listed in Table S2. For RNA‐seq analysis, the mycelium of FP‐1132.1 (reference strain) and FP‐1126.1 (ΔgaaR) was ground in Tissue Lyser II (Qiagen, Venlo, the Netherlands) and RNA was extracted using TRIzol reagent (Invitrogen, Breda, the Netherlands) and purified with NucleoSpin RNA Clean‐up kit (Macherey‐Nagel) with rDNase treatment. RNA quantity of the samples was checked with a NanoDrop‐1000 spectrophotometer and the quality by RNA gel electrophoresis. Single‐read samples were sequenced using Illumina HiSeq^™ 2000 platform (http://illumina.com). Purification of mRNA, synthesis of cDNA library, and sequencing reactions were conducted in the BGI Tech Solutions Co., Ltd. (Hong Kong, China). Transfer experiments and subsequent RNA‐sequencing were performed in duplicates.

Bioinformatics

Raw reads were produced from the original image data by base calling. On average, ~ 13 million read of 51 bp per sample were obtained. After data filtering, the adaptor sequences, highly ‘N’ containing reads (> 10% of unknown bases), and low‐quality reads (more than 50% bases with quality value of < 5%) were removed. After data filtering, in average, ~ 97.5% clean reads remained in each sample. Clean reads were then mapped to the genome of Aspergillus niger NRRL3 (http://genome.jgi.doe.gov/Aspni_NRRL3_1) using bowtie2 29 and bwa software 30. In average, 63.8% total mapped reads to the genome was achieved. The gene expression level was measured in ‘fragments per kilobase of exon model per million mapped reads’ (FPKM) 31 using RSEM tool 32. Genes with expression value lower than 14 were considered low‐expressed (approximately bottom 50%) and differential expression was identified by Student's t‐test with a P‐value cutoff 0.05. The RNA‐seq data have been submitted to Gene Expression Omnibus (GEO) 33 with accession number: http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE80227. Homology searches were performed using the blastp algorithm from NCBI against the nonredundant database and proteins with an E‐value ≤ 1E‐50 were defined as homologous 34. Hierarchical clusters using the average expression values of genes were made via genesis 1.7.7 35 with Pearson correlation and complete linkage. Low‐expressed pectinases in all conditions were not included.

Results and Discussion

Identification of the A. niger GaaR by homology to B. cinerea BcGaaR

A putative A. niger GA‐responsive transcriptional activator was identified by homology to the recently identified B. cinerea Zn₂Cys₆ TF (BcGaaR) 22. The A. niger ortholog (named GaaR) is a 740‐amino acid‐long protein encoded by gaaR (An04g00780/NRRL3_08195) and the bidirectional best blast hit of the 817‐amino acid‐long BcGaaR (Bcin09g00170). Analysis of the presence of GaaR among 20 Aspergillus species using the Aspergillus genome database (http://www.aspgd.org/) revealed that all Aspergilli, except Aspergillus glaucus contain a GaaR ortholog in their genome (data not shown). Interestingly, A. glaucus is not able to grow on GA as the sole carbon source (http://www.fung-growth.org), indicating the requirement of GaaR for GA utilization. A. niger GaaR and BcGaaR show 50.3% identity on the amino acid level throughout the entire protein sequence (Fig. S1). GaaR contains a typical Zn₂Cys₆ DNA‐binding domain with the pattern of CX₂CX₆CX₆CX₂CX₆C close to its NH₂‐terminal end (residues 26–56) and a fungal‐specific TF domain (residues 139–518). Amino acid alignment and phylogenetic analysis of GaaR revealed no significant similarity (an E‐value cutoff < 1E‐50) of GaaR to other TFs involved in plant cell wall utilization such as XlnR, AraR, RhaR, GalX, ClrA, and ClrB or to any other TF in A. niger (data not shown).

Deletion and complementation of gaaR and growth analysis of the ∆gaaR in A. niger

To assess the function of gaaR in A. niger, several deletion strains (ΔgaaR) were created and verified by Southern blot analysis (Fig. S2 and data not shown). The growth phenotype of the ΔgaaR strains was analyzed on different monomeric and polymeric carbon sources (Fig. 1A). Deletion of gaaR in the AB4.1 background (MA234.1, Fig. S2) and N593 background (N593.20, Fig. 1A) resulted in an identical phenotype. Disruption of gaaR resulted in a strongly reduced growth on GA and PGA and in a reduced growth and sporulation on SBP, CP, and AP. No significant differences in growth and sporulation were observed on other carbon sources tested (Fig. 1A, Fig. S2). The strongly reduced growth of ΔgaaR on GA and PGA was fully complemented by reintroducing the gaaR gene ectopically (Fig. S2).

Figure 1.

Phenotypic and gene expression analyses of A. niger ΔgaaR (A) Growth profile of the reference strain (FP‐1132.1) and ΔgaaR (FP‐1126.1) on MM with 25 mm monomeric and 1% polymeric carbon sources. Strains were grown for 4 days at 30 °C. (B) Northern blot analysis of selected GA‐induced genes in the reference strain (MA234.1) and ΔgaaR (JN35.1). Mycelia were transferred from d‐fructose (preculture) to GA or d‐fructose. Total RNA was isolated at the time of transfer (0 h) from mycelia grown in CM with 2% d‐fructose and at different time points (2, 4, and 6 h) after the transfer from mycelia grown in MM containing 50 mm GA (in bold) or d‐fructose.

GaaR is required for the induction of genes related to d‐galacturonic acid utilization

The presence of GA has been shown to induce genes involved in PGA degradation (e.g., pgxB, pgxC), GA transport (gatA), and catabolism (gaaA‐D) 7, 14. As a first indication for the involvement of GaaR in the induction of a subset of these genes on GA, a northern blot analysis was performed. The reference strain and ΔgaaR that made the AB4.1 background were pre‐grown in d‐fructose medium and transferred to either GA or d‐fructose medium. For the reference strain, transfer of mycelium to GA resulted in a rapid induction of pgxB, pgxC, gatA, gaaB, and gaaC, whereas this induction was not observed in ΔgaaR (Fig. 1B).

To analyze the expression of a larger number of genes involved in pectin degradation, GA transport and catabolism, a genome‐wide gene expression analysis was performed using RNA‐seq. The reference strain and ΔgaaR in the N593 background were again pre‐grown in d‐fructose medium and transferred to GA medium. RNA‐seq analysis indicated that the GA‐induced expression of all genes that were previously identified as part of the GA‐regulon 7 is dependent on GaaR (Table 1 and Fig. 2). The only exception is a putative GA transporter (An03g01620) that is expressed more than threefold less in ΔgaaR for which the P‐value did not pass our significance level (0.05). In general, these observations show that the genes in the suggested GA‐regulon 7 showed a significant reduction in ΔgaaR compared to the reference strain on GA (Table 1) and that GaaR is required for the induction of those genes.

Table 1.

RNA‐seq analysis on GA of the genes that depend on GaaR for induction. Expression values (FPKM) are averages of duplicates. Fold changes ≥ 2 and P‐values ≤ 0.05 are highlighted. GARE position is given with respect to the transcription start site

NRRL3 protein ID	CBS 513.88 gene ID	Gene name	Ref GA 2 h	ΔgaaR GA 2 h	Fold change Ref/ΔgaaR GA 2 h	P‐value	GARE (CCNCCAA) position
NRRL3_00958	An14g04280	gatA a	888.35	13.32	66.69	1.54E‐03	+ strand −360
NRRL3_08663	An03g01620	GA transporter (putative)a	106.09	30.34	3.50	1.25E‐01	+ strand −673
NRRL3_04281	An07g00780	GA transporter (putative)a	90.41	1.86	48.74	7.77E‐03	− strand −42 and −994
NRRL3_05650	An02g07710	gaaA a	2599.98	117.53	22.12	1.69E‐04	+ strand −414 and −100
NRRL3_06890	An16g05390	gaaB a	11309.00	344.03	32.87	1.88E‐03	+ strand −326
NRRL3_05649	An02g07720	gaaC a	5658.32	106.21	53.27	2.98E‐04	− strand −292 and −606
NRRL3_10050	An11g01120	gaaD a	8104.43	506.79	15.99	7.01E‐03	− strand −538, −583, −801 and −813
NRRL3_03144	An12g07500	pgaX a	698.90	24.27	28.80	1.19E‐02	+ strand −388
NRRL3_09810	An11g04040	pgxA a	10.65	0.34	31.32	9.10E‐03	− strand −594
NRRL3_08281	An03g06740	pgxB a	200.31	12.39	16.17	2.62E‐02	− strand −298 and −823
NRRL3_05260	An02g12450	pgxC a	99.93	4.10	24.40	6.24E‐04	+ strand −268 and − strand −642
NRRL3_06053	An02g02540	paeA (putative)a	522.81	22.99	22.75	4.57E‐03	+ strand −1238
NRRL3_04916	An07g08940	paeB (putative)c	13.41	10.57	1.27	7.42E‐01
NRRL3_08325	An03g06310	pmeA	6.54	0.42	15.75	1.18E‐02	+ strand −983 and − strand −308
NRRL3_07470	An04g09690	pmeB (putative)	30.16	4.67	6.46	1.41E‐02	+ strand −389
NRRL3_05252	An02g12505	pmeC (putative)b	558.37	24.68	22.62	4.20E‐03	+ strand −275, −246 and −35
NRRL3_02571	An01g11520	pgaI	56.38	6.56	8.59	6.96E‐04	+ strand −221
NRRL3_05859	An02g04900	pgaB c	15.10	3.11	4.86	6.74E‐02	− strand −753 and −934
NRRL3_08805	An05g02440	pgaC	5.26	0.59	8.99	3.65E‐02	+ strand −374, −196 and −865
NRRL3_02835	An01g14670	pgaE c	4.26	2.40	1.78	4.12E‐01
NRRL3_00965	An14g04370	pelA a	56.54	9.74	5.80	2.12E‐04
NRRL3_09811	An11g04030	pelC	0.51	0.00	NA	4.77E‐03
NRRL3_01237	An19g00270	pelD	18.95	0.34	55.74	6.03E‐04	− strand −409 and −465
NRRL3_04153	An15g07160	pelF c	35.48	37.02	0.96	8.73E‐01	− strand −644
NRRL3_10559	An18g04810	rgxC (putative)	20.00	0.90	22.22	1.26E‐02	+ strand −880 and −852 and − strand −250
NRRL3_00684	An14g01130	rglA	5.77	1.03	5.63	9.23E‐03	− strand −188
NRRL3_01606	An01g00330	abfA c	87.96	59.62	1.48	5.81E‐01
NRRL3_10865	An08g01710	abfC (putative)a	201.62	67.16	3.00	4.04E‐02
NRRL3_07094	An16g02730	abnD (putative)	4.57	1.53	2.99	2.66E‐02	+ strand −246
NRRL3_02479	An01g10350	lacB (putative)	137.63	41.24	3.34	1.36E‐02
NRRL3_11738	An06g00290	lacC (putative)	28.91	9.08	3.19	3.55E‐02	+ strand −267
NRRL3_10643	An18g05940	galA	105.64	29.24	3.61	3.53E‐02	+ strand −307

^aGenes identified as the GA‐regulon by Martens‐Uzunova and Schaap 7. ^b pmeC not present on the Affymetrix microarray. ^cGenes not significantly differentially expressed on GA, but differentially expressed on SBP (Table 2).

Figure 2.

Transcript levels of pectin utilization genes in A. niger reference and ΔgaaR on GA or SBP. (A) GA transporters and GA catabolic pathway enzymes; (B) exo‐polygalacturonases and pectin acetyl‐ and methylesterases; (C) endo‐polygalacturonases; (D) pectin lyases and endo‐xylogalacturonan hydrolase; (E) α‐l‐arabinofuranosidases, arabinan endo‐1,5‐α‐l‐arabinofuranosidase, endo‐arabinanases, ferulic acid esterases, and feruloyl esterase D; (F) β‐galactosidases, galactan 1.3‐β‐galactosidase, and β‐1.4‐endogalactanase; (G) the l‐rhamnose regulator rhaR and l‐rhamnose catabolic pathway enzymes; and (H) the l‐arabinose and d‐xylose catabolic pathway enzymes. Mycelia of the reference strain (FP‐1132.1) and ΔgaaR (FP‐1126.1) were pregrown in CM with 2% d‐fructose, washed and transferred to MM with 25 mm GA or 1% SBP and incubated for 2, 8 or 24 h.

To identify additional pectinase genes controlled directly or indirectly by GaaR, the expression of all 58 pectinolytic genes 3 was examined (Table S3). An overview of the gene abbreviations and their (putative) function is given in Martens‐Uzunova and Schaap 3. This analysis resulted in the identification of several additional pectinase genes for which the expression on GA is dependent of GaaR (Table 1 and Fig. 2, Fig. S3). This difference could be caused by higher sensitivity of the RNA‐seq analysis compared to the previously used Affymetrix microarrays. In general, these newly identified genes were lower expressed compared to the genes in the GA‐regulon described previously 7. The gene encoding the putative pectin methylesterase C (pmeC) was missing on the Affymetrix chips, and therefore missed previously, but the RNA‐seq study clearly indicated that induction of pmeC on GA is GaaR dependent. Inspection of the promoter regions of the newly identified members of the GA‐regulon indicated the presence of putative GaaR‐binding sites in the promoter regions of most genes (Table 1), enabling us to expand the GA‐regulon to a larger set of genes.

GaaR is required for the induction of genes related to polygalacturonic acid degradation and d‐galacturonic acid utilization on complex pectin

Both the strongly reduced growth phenotype on GA and PGA and the expression analysis in ΔgaaR suggest that that GaaR is required for GA utilization in A. niger. Growth and sporulation of ΔgaaR on complex pectins such as SBP was also reduced, but not as severe as on GA and PGA (Fig. 1A). This could be explained by two (not mutually exclusive) hypotheses. The first explanation could be that A. niger has alternative mechanisms (independent of GaaR) to induce genes involved in GA utilization. The second possibility is that additional sugars such as l‐arabinose, d‐galactose, d‐xylose, or l‐rhamnose that are present in SBP are metabolized and used for growth. To gain insight in the expression of pectinase genes in ΔgaaR on complex pectin, the reference strain and ΔgaaR were transferred from d‐fructose to SBP and grown for 2, 8, and 24 h before harvesting mycelia and extraction of RNA.

Expression profiles of pectinase genes in the reference strain and ΔgaaR were pairwise compared for identical time points (Table 2 and Fig. 2, Fig. S3). Most of the genes in the GA‐regulon, including those required for GA transport and catabolism, are dependent on GaaR for induction on SBP (Fig. 2A–D). This observation strongly suggests that ΔgaaR is not utilizing GA from SBP. The expression of gaaD/larA can be explained by the dual activity of the enzyme encoded by this gene as both an l‐glyceraldehyde reductase and an l‐arabinose reductase 11 and the utilization of l‐arabinose from SBP in ΔgaaR (see below). The expression profile of exo‐polygalacturonases, pectin acetyl‐ and methylesterases, endo‐polygalacturonases, and pectin lyases (Table 2 and Fig. 2B–D), all acting on the PGA backbone, support the conclusion that the GaaR target genes are not induced during growth on SBP in ΔgaaR.

Table 2.

RNA‐seq analysis on SBP of the genes that depend on GaaR for induction. Expression values (FPKM) are averages of duplicates. Fold changes ≥ 2 and P‐values ≤ 0.05 are highlighted

NRRL3 protein ID	CBS 513.88 gene ID	Gene name	Ref SBP 2 h	ΔgaaR SBP 2 h	Fold change Ref/ΔgaaR SBP 2 h	P‐value	Ref SBP 8 h	ΔgaaR SBP 8 h	Fold change Ref/ΔgaaR SBP 8 h	P‐value	Ref SBP 24 h	ΔgaaR SBP 24 h	Fold change Ref/ΔgaaR SBP 24 h	P‐value
NRRL3_00958	An14g04280	gatA a	849.85	12.60	67.45	6.18E‐05	1077.70	21.92	49.17	2.84E‐03	57.92	1.82	31.91	5.06E‐02
NRRL3_08663	An03g01620	GA transporter (putative)a	2647.36	642.68	4.12	1.01E‐04	387.81	273.28	1.42	2.41E‐01	2.69	3.20	0.84	7.16E‐01
NRRL3_04281	An07g00780	GA transporter (putative)a	33.99	7.43	4.58	1.89E‐01	14.34	6.24	2.30	5.60E‐02	34.49	2.63	13.11	8.63E‐03
NRRL3_05650	An02g07710	gaaA a	4649.59	70.19	66.24	2.16E‐03	2785.77	78.98	35.27	1.73E‐03	215.02	24.06	8.94	9.85E‐02
NRRL3_06890	An16g05390	gaaB a	11722.91	113.85	102.97	2.38E‐03	9634.93	229.60	41.96	2.72E‐04	208.56	65.62	3.18	8.85E‐02
NRRL3_05649	An02g07720	gaaC a	7306.08	92.88	78.66	2.95E‐05	6041.04	78.17	77.29	2.63E‐03	527.35	34.38	15.34	1.02E‐01
NRRL3_10050	An11g01120	gaaD a	11412.45	3807.08	3.00	5.06E‐03	7573.39	1434.68	5.28	8.61E‐04	621.40	409.01	1.52	4.46E‐01
NRRL3_03144	An12g07500	pgaX a	948.28	5.71	166.22	1.16E‐02	1154.06	21.72	53.13	4.37E‐03	133.41	4.97	26.87	1.59E‐03
NRRL3_09810	An11g04040	pgxA a	19.61	0.11	186.71	1.72E‐02	72.87	0.60	121.44	2.09E‐03	3.23	0.00	NA	NA
NRRL3_08281	An03g06740	pgxB a	483.28	3.15	153.66	1.81E‐02	784.73	13.84	56.72	6.23E‐02	364.35	3.40	107.32	4.37E‐02
NRRL3_05260	An02g12450	pgxC a	206.37	3.08	67.11	8.92E‐03	191.37	4.23	45.29	2.90E‐04	7.90	8.21	0.96	7.98E‐01
NRRL3_06053	An02g02540	paeA (putative)a	585.00	14.43	40.55	1.48E‐02	1836.37	30.62	59.97	1.75E‐02	300.70	11.83	25.43	1.46E‐02
NRRL3_04916	An07g08940	paeB (putative)c	166.46	7.51	22.16	1.47E‐02	427.08	62.09	6.88	9.60E‐02	151.00	10.65	14.18	9.63E‐04
NRRL3_08325	An03g06310	pmeA	15.92	1.30	12.24	3.81E‐02	27.43	0.99	27.70	3.18E‐03	0.90	0.12	7.50	2.43E‐01
NRRL3_07470	An04g09690	pmeB (putative)	41.34	4.84	8.55	3.34E‐02	130.79	52.94	2.47	1.77E‐02	33.31	2.06	16.17	2.20E‐01
NRRL3_05252	An02g12505	pmeC (putative)b	957.92	7.59	126.29	1.59E‐02	1917.18	26.77	71.63	1.09E‐02	249.81	0.40	624.53	1.67E‐02
NRRL3_02571	An01g11520	pgaI	18.75	4.38	4.29	1.44E‐02	106.37	6.59	16.14	8.39E‐03	9.88	4.83	2.04	4.97E‐01
NRRL3_05859	An02g04900	pgaB c	7.84	11.01	0.71	2.52E‐01	31.37	17.51	1.79	3.24E‐02	4.29	2.07	2.07	3.62E‐01
NRRL3_08805	An05g02440	pgaC	2.97	0.65	4.60	3.05E‐01	58.43	6.22	9.40	2.60E‐02	17.67	0.41	43.09	2.87E‐01
NRRL3_02835	An01g14670	pgaE c	3.16	1.35	2.34	8.30E‐02	177.86	19.69	9.03	3.46E‐03	28.57	15.13	1.89	1.99E‐01
NRRL3_00965	An14g04370	pelA a	40.64	11.09	3.67	1.14E‐02	861.91	9.02	95.55	2.19E‐02	25.43	5.41	4.70	1.82E‐01
NRRL3_09811	An11g04030	pelC	2.39	4.58	0.52	4.27E‐01	0.00	0.00	NA	NA	0.00	0.00	NA	NA
NRRL3_01237	An19g00270	pelD	11.72	0.61	19.37	1.10E‐02	17.97	0.58	31.24	1.26E‐03	1.28	0.42	3.05	8.54E‐03
NRRL3_04153	An15g07160	pelF c	18.03	21.02	0.86	4.67E‐01	44.85	14.09	3.18	2.56E‐02	64.95	25.60	2.54	5.26E‐04
NRRL3_10559	An18g04810	rgxC (putative)	86.41	0.94	92.41	6.26E‐02	207.96	3.76	55.38	4.39E‐02	116.98	1.11	105.39	1.02E‐01
NRRL3_00684	An14g01130	rglA	2.46	0.00	NA	NA	6.60	2.32	2.84	2.96E‐02	3.98	2.30	1.73	6.60E‐01
NRRL3_01606	An01g00330	abfA c	3864.76	705.67	5.48	6.85E‐03	133.72	176.93	0.76	3.59E‐01	1.56	4.72	0.33	1.76E‐01
NRRL3_10865	An08g01710	abfC (putative)a	2406.40	527.90	4.56	2.46E‐02	441.88	483.86	0.91	2.48E‐01	97.31	36.96	2.63	2.77E‐02
NRRL3_07094	An16g02730	abnD (putative)	3.92	2.82	1.39	1.06E‐01	14.38	6.20	2.32	4.33E‐02	55.45	12.53	4.43	2.40E‐01
NRRL3_02479	An01g10350	lacB (putative)	487.09	48.89	9.96	1.33E‐03	115.20	24.31	4.74	8.71E‐03	13.85	8.31	1.67	2.24E‐01
NRRL3_11738	An06g00290	lacC (putative)	297.53	12.04	24.71	6.23E‐03	314.50	8.56	36.74	3.61E‐02	113.92	2.81	40.61	3.74E‐02
NRRL3_10643	An18g05940	galA	154.02	30.49	5.05	1.82E‐02	398.55	26.21	15.21	2.85E‐04	19.05	20.72	0.92	5.82E‐01

^aGenes identified as the GA‐regulon by Martens‐Uzunova and Schaap 7. ^b pmeC not present on the Affymetrix microarray. ^cGenes not significantly differentially expressed on GA, but differentially expressed on SBP.

The results described above indicate that the residual growth of ΔgaaR on SBP is due to the utilization of other monosaccharides released from SBP. Analysis of the monosaccharide composition of the SBP used in this study was performed as described previously 36 and showed that it contains 55 mol% GA, as well as 17 mol% l‐arabinose, 16 mol% d‐galactose, and 10 mol% l‐rhamnose. Analysis of the expression of the genes involved in the degradation of RG‐I such as exo‐rhamnogalacturonases (rgx), rhamnogalacturonases (rhg), rhamnogalacturonan acetyl esterases (rgae), rhamnogalacturonyl hydrolases (urhg), arabinofuranosidases (abf), endo‐arabinanases (abn), ferulic acid esterases (fae), and β‐galactosidases (lac), and the genes responsible for catabolism of l‐rhamnose, l‐arabinose, and d‐xylose showed that these genes were still expressed in ΔgaaR (Fig. 2E–H, Fig. S3), indicating that the degradation and metabolism of RG‐I support the growth of ΔgaaR on SBP.

A clustering analysis of the expression of genes encoding the (putative) GA transporters, GA catabolic pathway genes, and pectinases provided further insight in the groups of coregulated genes (Fig. 3). Clusters E and G consist of genes that are members of the GA‐regulon (Table 1) and represent genes involved in the release and utilization of GA. Cluster F also consists mostly of genes that are part of the GA‐regulon (Tables 1 and 2). Genes in Cluster F, like genes in Clusters E and G, are expressed in the reference strain on GA and SBP at 2 and 8 h, but unlike genes in Clusters E and G also expressed in the ΔgaaR strain on SBP at 2 and 8 h. Cluster F mainly includes pectinases acting on RG‐I side‐chains. Their expression profile indicates that they are regulated by GaaR as well as other TFs involved in pectin degradation. Genes in Clusters A, B, C, and D are generally expressed in a GaaR‐independent fashion and represent pectinases acting on RG‐I and XGA. Pectinase genes of Cluster D are predominantly expressed in the ΔgaaR strain on SBP at 2 and 8 h. Genes in Clusters A, B, and C are expressed predominantly in the reference strain and ΔgaaR on SBP at 24 h or in ΔgaaR on GA, suggesting that these genes are likely induced on starvation or derepressed conditions.

Figure 3.

Hierarchical clustering of pectin utilization genes according to their expression in the reference strain (FP‐1132.1) and ΔgaaR (FP‐1126.1) on GA and SBP. The color code displayed represents the transcript levels of the genes. Clusters E and G include genes that are members of the GA‐regulon.

In conclusion, in this paper we showed that the conserved Zn₂Cys₆ TF GaaR of A. niger is required for the utilization of GA and PGA. We also showed that GaaR is essential for GA utilization from complex pectic substrates and that residual growth of ΔgaaR on complex pectins is likely due to induction of pectinases releasing l‐rhamnose from the RG‐I backbone and l‐arabinose and d‐galactose from the RG‐I ‘hairy regions’. These monosaccharides are metabolized independently of gaaR. With the identification of the GaaR in A. niger, we identified the missing link to further understand the interplay between several TFs involved in plant cell wall degradation. Insight in the regulation of pectin degradation and GA utilization in A. niger can help in exploiting A. niger for more efficient pectinase production.

Author contributions

EA, JN and JEK performed experiments, MP, MVAP and EA performed bioinformatics analysis, all authors analyzed results, EA, JEK, JALK, JV, RPV and AFJR wrote the manuscript with input of all authors, RPV and AFJR designed experiments and supervised the study.

Supporting information

Fig. S1. Alignment of AnGaaR and BcGaaR using EMBOSS Needle with standard settings (http://www.ebi.ac.uk/Tools/psa/emboss_needle/).

Fig. S2. Verification of the gaaR deletion strain in the MA234.1 background.

Fig. S3. Transcript levels of pectinases acting on RG‐I backbone in A. niger reference and ΔgaaR on GA or SBP.

Table S1. Strains used in this study.

Table S2. Primers used in this study. Overlapping sequences for fusion PCR are written in bold.

Table S3. RNA‐seq analysis of pectinases on GA and SBP.