Microbiological Safety and Sensory Quality of Cultivated Mushrooms (Pleurotus eryngii, Pleurotus ostreatus and Lentinula edodes) at Retail Level and Post-Retail Storage

In this study, the microbiological and sensory quality of cultivated mushrooms (Pleurotus ostreatus and eryngii and Lentinula edodes) available at the Austrian retail level were determined. Aerobic mesophilic bacteria (AMC), Enterobacteriaceae (EB), Pseudomonadaceae (PS), lactic acid bacteria (LAB), yeast, moulds and presumptive Bacillus cereus were enumerated at the day of purchase and after storage at 4 °C for 7 or 12 days. Additionally, the presence of Salmonella spp. and Listeria monocytogenes was investigated. Isolates of presumptive spoilage bacteria were confirmed by partial 16S rRNA sequencing. At the day of purchase, 71.2% of the samples were of high microbiological quality and grouped into the low contamination category (AMC < 5.0 log cfu/g), while the sensory quality of 67.1% was categorized as “very good or good”. After storage, the number of samples with high microbial quality was 46.6%, and only 37.0% of the samples scored as “very good or good”. The most abundant species across all mushroom samples were the Pseudomonas fluorescens species complex (58.4%) and the potential mushroom pathogen Ewingella americana (28.3%). All mushroom samples tested negative for Salmonella spp., L. monocytogenes and Bacillus cereus. The microbiological and sensory quality of the analysed mushrooms at the day of purchase and after storage was considered to be good overall. Longer transport distances were found to have a significant influence on the microbiological and sensory quality.


Introduction
In the last ten years, global mushroom and truffle production has increased from 7.5 million tonnes (t) in 2009 to 11.8 million t in 2019 [1]. Indeed, a wide range of cultivated mushrooms are available on the global market, the most common being white or champignon de Paris (Agaricus bisporus), Pleurotus species (ostreatus, sajor-caju, cystidus, citrinopileatus, flabellatus, eryngii), shiitake (Lentinula edodes) and enoki or winter mushrooms (Flammulina velutipes) [2]. In Austria, the self-sufficiency rate for cultivated mushrooms is only 9%. Imported cultivated mushrooms traded in Austrian retail markets mainly originate from farms in Germany, Poland and South Korea [3].
Fresh foods from stable to retail and customers undergo various processes (e.g., harvesting, washing, cutting, packaging, storing and transporting), which may influence the quality and the shelf life [4]. The food industry must ensure food safety and quality standards at all stages of the chain. Nevertheless, foodborne pathogens are found in various foods at European retail level. Mushrooms may also be contaminated with foodborne

Schematic Overview of the Experimental Program
The experimental program, comprising the main steps from the origin of cultivated mushroom samples, collection in retail markets, transportation and laboratory analyses, is illustrated in Figure 1. The pilot study primarily evaluated the microbial loads in association with the sensory quality of Austrian-grown and imported fresh cultivated mushrooms at the retail level and the effects of storage. The occurrence of foodborne pathogens in cultivated mushrooms at retail were studied. In single experimental approaches, the microbial status and the sensory quality of each batch were analysed at the day of purhase and after storage.

Collection and Preparation of Mushroom Samples
Oyster, king oyster and shiitake mushrooms (Pleurotus ostreatus, Pleurotus eryngii, Lentinula edodes) from six mushroom producers (A, B, D, E, F, G) and one packaging station (C) located in Austria, Germany, Poland and South Korea were investigated ( Figure 1, Table 1). Mushroom samples (each 150 to 400 g) were randomly sold at local retailers (n = 5) depending on their availability from January 2018 to May 2019. The mushrooms at retail level were either stored at room temperature (2/5 retail markets) or refrigeration temperature (7 • C). In total, 73 mushroom samples were collected in duplicates from the same batch and transported under refrigerated conditions (4 • C). At the day of purchase the initial microbial and sensory status (initial state, IS) of each mushroom sample was determined. At the end of shelf life (after storage at 4 • C, AS) the microbial and sensory status was determined for the duplicate sample set (Figure 1). The average storage time for oyster and shiitake mushrooms was seven days, and for king oysters it was 12 days, as the best-before date was defined by a retail chain after the day of packaging.
Microbial quality of mushrooms before (IS) and after storage was categorised according to the AMC load in five contamination levels: "low", < 5.0 log cfu/g; "medium", 5.1-6.5 log cfu/g; "high", 6.6-8.0 log cfu/g; and "very high", > 8.1 log cfu/g (according to our internal laboratory standard for fresh products).
Partial 16S rRNA gene sequencing was performed by Sanger sequencing technology (LGC Genomics GmbH, Berlin, Germany). The nucleotide sequences were quality evaluated, trimmed via Finch TV 1.4.0 (Geospiza Inc.; https://digitalworldbiology.com/ FinchTV) and subsequently compared with sequences from GenBank databases using the Nucleotide BLAST (Basic Local Alignment Search Tool) algorithm from the National Center for Biotechnology Information (https://blast.ncbi.nlm.nih.gov/Blast.cgi) [33]. Sequences were assigned to genus or species level according to best matches and highest similarities (similarity cut-off ≥ 97%). The sequences are provided under the accession number SUB8135232: MT997995-MT998248. Isolates suspicious for Listeria and Salmonella spp. were confirmed with a species-specific PCR reaction. Listeria species differentiation was performed by a multiplex PCR assay targeting the invasion-associated protein (iap) gene [34]. Salmonella spp. were confirmed by PCR amplification of the invasion protein (invA) gene [35].

Sensory Tests
Sensory quality evaluation of mushrooms including appearance, texture and consistency, aroma and taste was performed by three trained panellists [36]. Generally mushroom samples have to be clean, firm, undamaged and largely free from maggot damage, mineral and organic impurities according to Codex Alimentarius Standard, CXS 38-1981 [37]. The typical aroma of oyster mushrooms is described as intense umami and a lightly meaty note and that of king oyster mushrooms is slightly umami [38,39]. Shiitake mushrooms have an intense, sharp and slightly turniplike aroma [40]. Taste was determined by sautéing freshly sliced mushroom cap and stem (10 g) in canola oil (1:10, oil:mushrooms) for five minutes.
The maximum achievable score was five points, which meets the quality standard for edible fungi and fungus products according to Codex Alimentarius Standard [37]. The sensory quality of mushrooms (appearance, consistency and texture, aroma and taste) before and after storage was evaluated according to a grading system 1 to 5 (Supplementary  Table S1): 5.0 to 4.0 "very good or good"; 3.9 to 3.0 "satisfactory"; 2.9 to 2.0 "still acceptable" and below 1.9 "not acceptable" and inadequate for human consumption [41].

Statistical Analyses
Descriptive statistics (mean, standard deviation, minimum and maximum values) of variables (AMC, EB, PS, LAB and sensory quality scores) were performed with IBM SPSS V.24 (SPSS Inc., Chicago, IL, USA). Logarithmically transformed microbial counts resulted in normally distributed data and were assessed by Kolmogorov-Smirnov test. Mixed model analysis was used to identify the difference in microbial counts and sensory score between the two time points (initial status and after storage) where producer and country of origin were added as factors to the model. Multiple comparisons were performed using Sidak's alpha correction procedure. Cross-tabulation table were created to test frequency distributions and associations between categorical variables (microbial counts and sensory quality of initial/stored mushrooms) and were calculated via Pearson's chi square test. A p-value of p < 0.05 was considered statistically significant.

Microbial Counts, Sensory Quality Score and Detection of Foodborne Pathogens
The results of three mushroom species are presented in Table 2. The average counts of AMC, EB and PS (3.8, 2.6 and 3.1 log cfu/g) increased significantly after storage (5.1, 3.3 and 4.0 log cfu/g; p < 0.05). Average sensory quality scores decreased in oyster (4.3 to 3.5), king oyster (4.4 to 3.2) and shiitake mushrooms (3.7 to 3.0) significantly after storage (p < 0.05). In general, in 60.3% (n = 44/73), 47.9% (n = 35/73) and 34.2% (n = 25/73) of mushroom samples the increase of EB, PS and AMC was ≤ 0.5 log cfu/g after storage. An increase of more than 2.0 log cfu/g for PS, AMC and EB was observed for 31.5% (n = 23/73) 28.8% (n = 21/73) and 24.6% (n = 18/73) of all batches. In particular, a significant increase of AMC and EB counts was observed for king oyster mushrooms after storage (1.9 and 1.3 log cfu/g; p < 0.05) ( Table 2).  The lowest average AMC (initially: 3.2 log; stored: 4.5 log cfu/g), EB (initially: 1.5 log; stored: 2.5 log cfu/g) and PS counts (initially: 2.4 log; stored: 3.3 log cfu/g) were found in mushrooms produced in Austria. Mushrooms from South Korea were exclusively king oyster. These samples had highest AMC (initially: 5.0 log; stored: 7.6 log cfu/g) and EB counts (initially: 4.7 log; stored: 6.9 log cfu/g). The highest initial PS counts were seen in oyster mushrooms from Poland (4.4 log cfu/g) and after storage in South Korean king oyster mushrooms (5.4 log cfu/g), all from the same packer C.
Salmonella spp., L. monocytogenes, Listeria spp. and presumptive Bacillus cereus were not detected in any mushroom sample at the day of purchase or after storage at 4 • C (n = 146).

Pseudomonadaceae and Ewingella americana Attribution to Samples and Microbial and Sensory Quality
P. fluorescens SG and E. americana were predominantly isolated from king oyster mushrooms, which can be seen in Figure 4. The most abundant species or bacterial subgroups isolated from all sample types (n = 146) were P. fluorescens SG (54.7%; n = 80), E. americana (45.2%; n = 66), P. gessardii SG (26.0%; n = 38) and P. mandelii SG (8.9%; n = 13) ( Figure 4). P. azotoformans (17.8%; n = 26), P. proteolytica (13.6%; n = 20), P. tolaasii (12.3%; n = 18) and P. brenneri (11.6%; n = 17) were the most observed Pseudomonas species detected in the samples.  More sensory deficits (QS ≤ 2.9, "still acceptable") and higher microbial loads (AMC ≥ 6.6 cfu/g) were seen in samples containing E. americana and at least one Details can be seen in Table 3. In ten batches of stored oyster and king oyster mushrooms, an increase in the microbial load of ≥ 4.0 log cfu/g was found, most often in samples from packer C and producer A. In all cases, the increase was associated with an increase of PS (P. fluorescens SG) or/and EB (E. americana) counts. Abbreviations: a QS-score of sensory quality, "still acceptable"; b AMC-aerobic mesophilic count, "high" and "very high" microbial contamination; c Ewingella (E.) americana in co-occurrence with at least one of the listed Pseudomonas (P.) spp.; d one or more of the listed P. spp.

Discussion
This study was initiated to evaluate the microbial and sensory quality of fresh cultivated mushrooms sold in the Austrian retail market and to examine the shelf life stability during storage at 4 • C based on the proposed expiration date. The microbiological and sensory quality of three mushroom species (Pleurotus ostreatus, Pleurotus eryngii, Lentinula edodes) at the day of purchase and after storage was considered to be good overall. "High" and "very high" levels of contamination was observed in only 8.9% (n = 13/146) of mushroom samples, most often in king oyster (11.3%, n = 7/62) and oyster (9.1%, n = 4/44) mushrooms (Figure 2A). The microbial load of mushroom samples varied widely at the day of purchase and after storage: AMCs at the retail level ranged from log 1.7 to 7.8 at day of purchase and from 1.7 to 9.4 log cfu/g after storage (Table 2). Overall, these findings are in accordance with data from other studies. AMCs at retail level have been reported as follows: 7.7 to 8.4 log cfu/g for champignon, 5.0 to 5.3 log cfu/g for oyster, 4.9 to 6.9 log cfu/g for shiitake and 5.9 log cfu/g for king oyster [23,42,43]. AMCs at the harvest level ranged from 3.1 to 5.8 log cfu/g for champignon, 3.5 log cfu/g for oyster and 4.0 to 4.5 log cfu/g for shiitake [18,20,21,44,45]. Counts of LAB, yeasts and moulds were low (1.1, 2.2 and 2.9 log cfu/g, respectively, Table 2), which is in accordance with findings by Venturini et al. [23]. At the day of purchase, the majority (71.2%) of mushroom samples investigated in our study were categorized as having a "low" level of contamination and similar results were found in the sensory quality, as 67.1% of the samples scored as "very good or good" (Figure 2A,B). After storage, the number of samples with high microbial quality ("low" contamination level) was 46.6% and those with "high" and "very high" contamination increased by 25.9% (Figure 2A). Changes of the visual appearance (cap, stem discoloration or brown or yellow spots, gill sticky, macerated), consistency and texture (soft, rubbery, fibrous), aroma (loss, old, musty, putrid) and taste (loss, bitter) resulted in reduced sensory quality. Only 37.0% of the samples scored as "very good or good"; in total, 21.9% (n = 16/73) were deemed "not acceptable". Most often king oyster (29.0%, n = 9/31) and shiitake (25.0%, n = 5/20) mushrooms were unfit for consumption ( Figure 2B). In contrast to oyster and king oyster, in shiitake mushrooms with sensory deficits a "low" or "medium" microbial contamination was often seen (Figure 2A,B). This observation could be explained by the fact that packed shiitake mushrooms tend to dry out during storage. Other studies reported an AMC increase in shiitake mushrooms during storage at 4 • C for six to eight days by 2.0-3.0 log cfu/g as well as a deterioration of the sensory quality [20,21,45].
In our study high AMCs were associated with the presence of Pseudomonas spp. and E. americana. This is consistent with the literature [18,23,43,46]. The majority of Pseudomonas isolates were assigned to P. fluorescens (54.7%) and P. gessardii (26.0%) subgroups (Figure 4). In particular, P. azotoformans, P. proteolytica, P. tolaasii and P. brenneri were often detected. Frequently, bacterial mushroom spoilage is associated with P. tolaasii, leading to brown blotches and the resulting yellow to dark brown lesions on the mushroom cap, which are induced by the toxin tolaasin [47][48][49]. Additionally, other bacteria such as P. azotoformans and P. brenneri are reported to cause spoilage of diseased mushrooms [50]. In industrial cultivation, other bacteria such as P. azotoformans, P. fluorescens, P. putida and Bacillus spp. are used as antagonists that suppress the aforementioned mushroom pathogenic bacteria [51][52][53]. This could explain why these bacteria were seen in our samples as well. E. americana was found in 45.2% of the samples and was present in all mushroom species (Figure 4). This is in accordance with Reyes et al. (2004) where E. americana was highly prevalent in oyster (76.7%) and shiitake mushrooms (73.3%) purchased at retail markets [43]. E. americana has been reported to be an opportunistic pathogen in humans, isolated in patients with immunosuppressed status [54]. According to the literature, E. americana might either cause internal stipe necrosis in champignon and king oyster mushrooms, or soft rot and mild browning in oyster mushrooms. Further, the bacteria is a "commensal" in healthy champignon, oyster and shiitake mushrooms [43,55,56]. Recent results show that blotch symptoms of champignon are often associated with bacteria of the Enterobacteriaceae family, while Pseudomonadaceae were present in healthy and diseased mushrooms [57]. In our study, in stored mushroom samples with increased Enterobacteriaceae counts E. americana was dominant (83.3%; n = 15/18). We found that samples containing E. americana and at least one Pseudomonas species more often had sensory deficits and higher microbial loads (p < 0.05) ( Table 3). The bacterial community of samples classified as unfit for consumption (OS < 1.9) contained high levels of E. americana and diverse Pseudomonas species (76.2%; n = 16/21) including potential mushroom pathogens (P. azotoformans, P. brenneri and P. tolaasii). Three samples (two king oyster and one oyster) also showed symptoms as described by Gonzales et al. (2012) [56]. It seems that the co-occurrence of E. americana and Pseudomonas species in one sample influences the storage stability of mushrooms.
L. monocytogenes, Salmonella spp. and presumptive B. cereus were not detected in any of our samples. A Spanish study was also unable to isolate L. monocytogenes and Salmonella spp. in cultivated mushrooms [23]. However, in oyster, king oyster and shiitake mushrooms sold in Chinese markets, L. monocytogenes was found in 6.7%, 4.4% and 2.9% of the samples, respectively [12]. Recently, two studies reported an incidence of L. monocytogenes of 0.8% [14] and 50.0% [58] in processed mushrooms. There is increasing evidence that L. monocytogenes results from contamination during processing. The pathogen has many stress adaptions that enable survival in a wide range of environmental niches [59,60]. Further, L. monocytogenes has also been isolated from the mushroom farm environment, emphasizing the importance of monitoring the production chain from the substrate production to harvest, processing and packaging [8,10,61].

Conclusions
This study focused on the quality of fresh mushrooms at the retail level, analysed on the day of purchase and after storage. The majority of mushrooms investigated at the day of purchase had a high microbiological quality. After storage, the AMC increased, especially in oyster and the South Korean king oyster mushrooms. The sensory quality of mushrooms did not always change depending on the microbiological load. Changes in consistency and texture, including drying and increased toughness, occurred during storage, and might have been influenced by the type of packaging. Poor hygiene facility management, re-packaging and long transport distances might lead to product spoilage and reduced shelf life. The king oyster mushrooms from producer B had the highest product stability and the shortest transport distances to the supermarket compared to packer C, which re-packed the mushrooms and had the longest transport distances. Severe changes of the sensory quality were mainly observed in such stored samples where E. americana were found in combination with different Pseudomonas species. Therefore, the monitoring of these bacteria species could be useful in the assessment of the shelf life. It is known that the stability of mushrooms during storage is influenced by many interacting factors. To obtain microbiologically safe mushrooms and maintain their sensory quality, further research should therefore focus on various steps along the processing chain and on the effectiveness of packaging systems to prevent moisture loss and microbial growth.