The Weak-Acid Preservative Sorbic Acid Is Decarboxylated and Detoxified by a Phenylacrylic Acid Decarboxylase, PadA1, in the Spoilage Mold Aspergillus niger

Abstract

Resistance to sorbic and cinnamic acids is mediated by a phenylacrylic acid decarboxylase (PadA1) in Aspergillus niger. A. niger ΔpadA1 mutants are unable to decarboxylate sorbic and cinnamic acids, and the MIC of sorbic acid required to inhibit spore germination was reduced by ∼50% in ΔpadA1 mutants.

Naturally occurring weak organic acids are added to foods and beverages as flavor compounds (e.g., cinnamic and butyric acids), acidulants (e.g., acetic and citric acids), and preservative agents (e.g., sorbic and benzoic acids). Cytoplasmic acidification during weak-acid stress has been demonstrated in both yeasts and molds (9, 15) and causes inhibition of enzyme activity and amino acid transport (7, 9). Accumulation of the anion fraction of dissociated preservatives is reported to cause oxidative stress and increased turgor pressure within the cell (14).

Some mold and yeast species decarboxylate and thereby detoxify sorbic acid, and their contamination of foods and beverages supplemented with sorbic acid typically leads to “hydrocarbon-like” or “kerosene-like” odors due to 1,3-pentadiene production (3, 4, 8, 10). We have shown recently that S. cerevisiae Pad1p, an enzyme that has been shown previously to decarboxylate cinnamic acid to styrene (5), facilitates the decarboxylation of sorbic acid to the volatile compound 1,3-pentadiene (17). Although several mold species have been shown to degrade sorbic acid to 1,3-pentadiene, the gene(s) and the enzyme(s) that facilitate this degradation were unknown. We show here that Aspergillus niger, a significant spoilage mold (15), decarboxylates sorbic and cinnamic acids to the volatile compounds 1,3-pentadiene and styrene, respectively, and this reaction is facilitated by PadA1, a Pad1p homologue, and constitutes a strong resistance mechanism to weak-acid toxicity.

Previous studies showed that the concentration of sorbic acid in the culture medium decreases during the germination and early outgrowth of A. niger conidia (15). Our hypothesis was that A. niger conidia decarboxylated sorbic acid to 1,3-pentadiene, and this accounted for the disappearance of sorbic acid from the culture medium. To confirm this hypothesis, A. niger N402 (2) conidia (10⁷) were cultured with 1 mM sorbic acid in sealed McCartney bottles containing 10 ml of ACM (pH 4) (15), shaken at 120 rpm at 28°C. One-milliliter headspace samples were removed by piercing the rubber cap septa with a syringe needle. Samples were immediately injected into stainless steel thermal desorption tubes (89 by 5 mm, inner diameter) packed with 200 mg of Tenax TA 60 to 80 mesh (Phase Separations, United Kingdom), followed by a 10-ml air flush before being examined by gas chromatography-mass spectrometry. The concentration of 1,3-pentadiene in the headspace of cultures containing sorbic acid increased over time (Fig. 1). Similarly, the concentration of styrene in the headspace of A. niger cultures increased when cultured with 1 mM cinnamic acid (Fig. 1). After 10 h of incubation, 60 and 88% of the sorbic and cinnamic acids had been converted to 1,3-pentadiene and styrene, respectively. In contrast, S. cerevisiae decarboxylated ∼1.5% of the acids over the same period (our unpublished data).

FIG. 1.

Time course of decarboxylation of sorbic acid to 1,3-pentadiene (•) and cinnamic acid to styrene (○) by A. niger N402. The results presented are from independent cultures (n = 2 for each).

To identify likely candidate genes in A. niger that facilitate the degradation of sorbic and cinnamic acids, the A. niger genome database, held by DSM at the time of the present study but now publicly available (http://genome.jgi-psf.org/Aspni1/Aspni1.home.html), was mined for Pad1p homologues. Three A. niger genes encoding protein with homology to the amino acid sequence of Pad1p were identified (designated PadA1, PadA2, and PadA3) (nucleotide sequences have been deposited in GenBank under accession numbers EF215454, EF215455, and EF215456, respectively), and the proteins showed 63.0, 56.9, and 58.8% identities with the S. cerevisiae Pad1p amino acid sequence, respectively. Other PadA/Pad1p homologues have been identified in molds and yeast species whose genomes have been published.

To evaluate the importance of the A. niger padA1 gene in conferring resistance to sorbic and cinnamic acids, a gene disruption vector (pAXP6-2) was constructed, in which the open reading frame (ORF) of the padA1 gene was replaced by the A. oryzae pyrG gene. The primers PaddisF and PaddisRII (5′-GCTTCTACGTATTATTCCGCAGCG-3′ and 5′-AACAAGGCGAA TAAGGTCCCCTC-3′, respectively) were used to amplify by PCR an ∼4.7-kb fragment containing the A. niger padA1 gene. PCR cycling was as follows: 1 cycle of 3 min at 98°C; followed by 35 cycles of 20 s at 98°C, 30 s at 55°C, and 2 min at 72°C; followed finally by 1 cycle of 10 min at 72°C. The resulting PCR product was digested with EcoRI and NsiI, and a 3.5-kb DNA fragment containing the padA1 ORF and its flanking sequence was ligated into an EcoRI/NsiI-digested pGEMT-Easy plasmid (Promega) to produce pAXP5. pAXP5 was digested with BamHI and NheI to remove a 1.325-kb fragment containing the padA1 ORF (684 bp) plus the flanking sequence on either side. A DNA fragment containing the A. oryzae pyrG gene was obtained as a BamHI/NdeI fragment from pAO4-13 (6) and ligated into the BamHI/NdeI-digested pAXP5 plasmid to produce pAXP6-2. pAXP6-2 was linearized with NaeI and used to transform A. niger MA70.15 (Δkus70::AmdS) (12). A. niger protoplast formation and transformation procedures followed standard methods (1). Strain MA70.15 was also transformed with pAB4.1 (21), containing the A. niger pyrG gene, to produce a reference pyrG⁺ strain (AXP8-1). A. niger pyrG⁺ transformants were purified and screened by using PCR for putative deletion mutants, and Southern blot analysis was used to further corroborate the deletion of the A. niger padA1 gene (data not shown). Genomic DNA was extracted from A. niger mycelia as described previously (13), and Southern blot analyses were performed as described previously (16). A 907-bp DNA fragment downstream of the A. niger padA1 ORF (bp +900 to bp +1806) was amplified by PCR from A. niger genomic DNA by using the primers padkoF and padkoR (5′-CAACTACAAGCACGTTGTGTTTGC-3′ and 5′-ACAGTGCCGAGGGCAGTTTC-3′, respectively) under the following conditions: 1 cycle of 3 min at 98°C; followed by 35 cycles of 20 s at 98°C, 30 s at 55°C, and 30 s at 72°C; followed finally by 1 cycle of 10 min at 72°C. With this DNA fragment, a [α-³²P]dCTP-labeled probe was synthesized by using the Megaprime DNA labeling kit (Amersham Pharmacia Biotech). Southern blot analysis revealed the padA1 genes of the four transformants under investigation were all disrupted, and transformant AXP6-2.21a was selected for further investigation. Spore germination and the growth rate in ACM broth (without sorbic or cinnamic acids) of the ΔpadA1 mutant compared to the reference strain (AXP8-1) were identical in both cases (data not shown).

The ability of the ΔpadA1 mutant to degrade sorbic and cinnamic acids and the sensitivity of this strain to both acids was investigated. Disruption of the padA1 gene completely prevented the decarboxylation of sorbic and cinnamic acids to 1,3-pentadiene and styrene, respectively (Fig. 2). The MIC of sorbic acid required to completely inhibit spore germination and subsequent growth of A. niger was reduced by ∼50% by the ΔpadA1 mutant compared to the reference strain (AXP8-1) (Table 1). MIC tests were carried out in McCartney bottles containing 10 ml of ACM (pH 4) supplemented with various concentrations of weak acid, inoculated with 10⁴ conidia, and incubated 28°C for 28 days. The A. niger ΔpadA1 mutant showed increased sensitivity to cinnamic acid, but it was not possible to assign a direct measure of the difference between the MICs for the ΔpadA1 mutant strain and the reference strain because the latter was capable of growing in medium containing cinnamic acid at concentrations up to the solubility limit of 6 to 7 mM at pH 4. The MIC of benzoic acid required to inhibit spore germination by A. niger was, as expected since benzoic acid is not a PadA1 substrate, identical in both the mutant and the reference strains (AXP8-1) (Table 1). The degradation of benzoic acid by A. niger and other fungi has been well documented (18-20, 22) and differs in mechanism from that for breakdown of sorbic and cinnamic acids.

FIG. 2.

Percent conversion of sorbic acid to 1,3-pentadiene (▪) and cinnamic acid to styrene (□) by conidia of A. niger AXP6-2.21a (ΔpadA1), AXP8-1 (pyrG⁺, reference strain), and N402 (wild type). A total of 10⁷ conidia were incubated in ACM (pH 4) containing 1 mM sorbic or cinnamic acids for 10 h at 28°C. The results presented are from independent cultures (n = 2 for each).

TABLE 1.

MICs of sorbic, cinnamic, and benzoic acids required to completely inhibit spore germination and subsequent growth of A. niger (ΔpadA1) for up to 28 days^a

Strain	MIC (mM)
	Sorbic acid	Cinnamic acid	Benzoic acid
AXP8-1	5	>6.5	4
AXP6-2.21a	2	4	4

^aMcCartney bottles (30 ml) containing 10 ml of ACM (pH 4.0) were supplemented with increasing concentrations of sorbic, cinnamic, or benzoic acids (AXP8-1 = A. niger pyrG⁺ reference strain; AXP6-2.21a = ΔpadA1). The results presented are from independent cultures (n = 3 for each). Note that the solubility limit for cinnamic acid prevents MIC determinations in excess of 6.5 mM at this pH.

Complementation of the ΔpadA1 was achieved by transforming this strain with ScaI-linearized pAXP5. The reintroduction of the native padA1 gene and the subsequent loss of the A. oryzae pyrG gene marker was selected for by supplementing the growth medium with 2 mg of 5-fluoroorotic acid/ml and 10 mM uracil. Transformants were purified and screened by using PCR to confirm the presence of the padA1 gene. To ensure that the complemented transformants degraded sorbic and cinnamic acid to the same level as the wild type, three transformants (AXP6-2.21aRa, AXP6-2.21aRh, and AXP6-2.21aRm) were cultured with either 1 mM sorbic or 1 mM cinnamic acid, and headspace samples were taken as described previously. The results showed that the growth of padA1-restored strains and the amount of degradation of both acids by all three of the padA1-restored transformants were comparable to those of the wild type (a comparison of one transformant with wild type and the ΔpadA1 mutant strain is shown in the supplemental material).

To further corroborate the function of A. niger PadA1, the padA1 ORF, under the control of the S. cerevisiae ADH1 promoter, was expressed in S. cerevisiae Δpad1. The A. niger padA1 ORF was amplified by PCR from genomic DNA using primers padoeF and padoeR (CGGCTAGCATGTTCAACTCACTTCTATCCGGTAC and CGGGATCCTTATTTTTCCCATCCATTCCAAC, respectively). The underlined portions indicate artificial NheI and BamHI restriction sites, respectively, to aid subsequent ligation reactions. PCR cycling consisted of 1 cycle of 3 min at 98°C; followed by 35 cycles of 20 s at 98°C, 30 s at 55°C, and 20 s at 72°C; followed finally by 1 cycle of 10 min at 72°C. The PCR fragment containing the padA1 ORF was digested with BamHI and NheI and was ligated into pVTU260 (11) to produce pVTU-padA1. pVTU-padA1 was used to transform S. cerevisiae Δpad1 (Euroscarf strain Y05833) using standard techniques. The resulting strain, unlike the control (transformed with empty pVTU260), was able to decarboxylate sorbic and cinnamic acids at a level similar to that of the PAD1-containing wild-type yeast. This demonstrates that A. niger padA1 is able to complement the Δpad1 phenotype of S. cerevisiae (see the supplemental material).

Of the mold species whose genome sequences have been published, A. fumigatus and A. clavatus, unlike other Aspergillus spp., are unusual in not containing an apparent padA orthologue.

In conclusion, the results show that the A. niger padA1 gene (encoding PadA1) represents the major mechanism of resistance by this spoilage mold to sorbic and cinnamic acids. Disruption of the padA1 gene renders A. niger sensitive to sorbic and cinnamic acids. The precise roles of A. niger padA2 and padA3 are under investigation. Inhibition of the degradation of sorbic and cinnamic acids, facilitated by PadA1, presents an attractive target for the development and refinement of future food and beverage preservative strategies.