Blue mold, caused by Penicillium expansum and other Penicillium spp. is the most common and economically significant postharvest disease of apples and pears worldwide and losses have been estimated to exceed 4.4 million dollars annually [1-3]. In addition to P. expansum, P. carneum, P. paneum, P. crustosum and P. solitum have been isolated from apples with blue mold symptoms from the major tree fruit producing regions in Washington State and Pennsylvania [4,5]. Blue mold is characterized by a soft, watery rot that is light brown with blue-green colored conidia, often forming coremia, that develop on the fruit surface following advanced decay. P. expansum and other Penicillium spp. generally enter through wounds caused by stem punctures, bruises, and fingernail scratches that occur during harvest, but can also infect via natural openings (i.e., cracked lenticels, stem ends and open calyx). Penicillium spp. is troublesome because they reduce the availability of fresh fruit for consumption, contribute to food waste and produce mycotoxins (i.e., patulin, citrinin and penicillic acid) that contaminate processed fruit products [6]. Patulin is carcinogenic and is of primary concern in the United States and in Europe where strict limits on its amount in fruit juices and processed pome fruit products are set to a maximum level of 50 μg/L [6,7].

There is no source of host resistance in commercial apple cultivars as they are all susceptible to blue mold [8]. However, pioneering studies involving the identification and characterization of wild apple germplasm with resistance to blue mold have revealed multiple mechanisms associated with host resistance against Penicillium spp. infection [9-12]. It is envisioned that these findings will be utilized to incorporate blue mold resistance from specific wild apple accessions into existing commercial apple cultivars (e.g. ‘Gala’, ‘Fuji’, ‘Honeycrisp’, etc.). Current blue mold management options involve the use of postharvest fungicides, biological control formulations (BioSave™) and sanitation of bins and storage buildings [13,14]. Although, fungicideresistant populations of Penicillium spp. have emerged in the packinghouse environment, thus greatly reducing the efficacy of fungicidal controls [15,16]. Hence, the lack of host resistance in cultivated apples, coupled with reduced efficacy of chemical controls, has demonstrated the need to develop targeted methods to manage blue mold decay which are facilitated by functional genetics studies in the fungus.

The first public release of an assembled and annotated P. expansum genome was published recently and has been available at the National Center for Biotechnology Information [17,18]. This has stimulated much interest in identifying genes and gene products that mediate pathogen virulence, spore germination and mycotoxin production [19-21]. Additionally, an Agrobacterium tumefaciens Mediated Transformation (ATMT) system for P. expansum using hygromycin and the hygromycin phosphotransferase (hph) gene as a single dominant selectable marker (DSM) has allowed researchers to analyze gene function, which has been developed [22]. Therefore, both functional genetic tools and corresponding genomic platforms have enabled researchers to utilize ATMT coupled with a single DSM to analyze Penicillium spp. gene function [19-21]. However, the use of one selectable marker is not optimal or desirable to conduct molecular genetics investigations. Several situations arise when additional dominant selectable markers are needed for example: complementation of a mutant strain with a different dominant selectable marker, when several pathway enzyme mutants (e.g. patulin) are transformed in the same strain, and where dominant markers allow selection in prototrophic strains [23,24]. Therefore, the specific objective of this research was to screen multiple commercially available compounds suitable for Penicillium spp. selection, determine their MICs and evaluate their efficacy on different isolates of the same species to ensure broad functionality in Penicillium spp. causing blue mold decay of pome fruit.

Penicillium spp. isolation, culturing and storage

Nine of the Penicillium spp. isolates (F16, G2, G9, P24, R14, R19, SAH1, SAR1, SA4) examined were obtained from decayed apple fruit exhibiting blue mold symptoms from a commercial packing and storage facility located in Pennsylvania and a farm stand in Beltsville, Maryland by Dr. Wayne M Jurick II. Briefly, fruit surfaces were surface sanitized by spraying with 70% ethanol until run off and the asymptomatic tissue was removed from the lesion margin using aseptic techniques and a sterile scalpel. Fragments of apple tissue were placed on Petri plates containing Potato Dextrose Agar (PDA) and incubated at 25°C until fungal growth was evident. Pure cultures were obtained by touching the sporulating culture with a sterile loop and streaking out onto fresh PDA Petri plates. Monoconidial Penicillium spp. isolates were maintained on PDA plates and on PDA slants stored at 4°C. Additional isolates, F-Fr-J8, 16104, 42710 and Stanley A, were provided by Dr. Wojciech J Janisiewicz, USDA-ARS AFRS in Kearneysville WV. Isolate 3354 was provided by Dr. Richard Kim at Pace International in Wapato, Washington and RS1 was a gift from Dr. Robert A Spotts in Hood River, Oregon.

Morphological and genetic identification of Penicillium spp.

To determine the species-level identity of all 15 Penicillium isolates, two standard identification methods were utilized. Single spore cultures were preliminarily identified using morphological methods in vitro  (Figure 1) [25]. Confirmation of species-level identity was achieved using conventional PCR by sequencing ~700 bp of the β- tubulin locus as previously described [26]. Purified PCR products were subjected to Sanger sequencing using both forward and reverse primers. Geneious software was used to assemble a 2X consensus of each amplicon and was subjected to MegaBlast analysis (Table 1).

Figure 1: Photograph of the A) top and B) reverse of the 15 isolates analyzed in this study. From top to bottom, left to right P. expansum R19, P. solitum RS1, P. expansum SAR1, P. solitum Stanley A, P. expansum 3354, P. solitum 16104, P. expansum 42710, P. carneum G2, P. paneum G9, P. expansum F16, P. expansum F-Fr-J-8, P. expansum P24, P. expansum SA4, P. crustosum R14, and P. expansum SAH1. Penicillium spp. cultures were grown in a temperature controlled incubator at 25°C for 4 days and 25 μl of 1 × 10⁴ conidia/ml suspension was inoculated as three spots on a PDA plate for each isolate.

Table 1: Penicillium species examined in this study. All isolates were propagated as monoconidial cultures, subsequently identified using standard morphological methods, and via sequencing the β-tubulin gene.

Minimum inhibitory concentration (MIC)

MIC of six commercially available compounds for five different Penicillium spp. were conducted using conidial suspensions from seven-day-old Penicillium spp. cultures grown on PDA that were harvested with 1 ml of filter sterilized 0.05% Tween 20-treated water. Conidial suspensions were vortexed for ten seconds, quantified using a hemacytometer, and adjusted to 1 × 10⁴ conidial/ml. Technical grade compounds of benlate, chlorimuron ethyl, bialophos, G148, neomycin and phleomycin D1 ranging from 0 to 1000 μg/ml were added to Richards defined medium containing sucrose as a sole carbon source pH 7.0 with 15 g/l phytagel agar (Table 2). Amended medium was added to 96-well plates and allowed to cool. Twenty five microliters of quantified conidial suspension from each isolate were pipetted into three independent wells of a 96 well plate, and placed in a temperature controlled incubator for 4 days at 25°C with natural light. Plates were examined using dissecting and compound microscopes for fungal growth. MIC was determined as the lowest concentration of compound that halted conidial germination and inhibited mycelial proliferation in vitro . Experiments were conducted twice using separate 96-well plates and different PDA plates for each isolate as the source of conidia. Additionally, the efficacy of benlate, chlorimuron ethyl, and bialophos on ten different P. expansum and P. solitum isolates was carried out using Richards defined medium in 96-well plates amended with 0, 1, 5 and 10 μg/ml benlate or chlorimuron ethyl and 0, 100, 250, 500, 750 and 1000 μg/ml bialophos (Table 3). The experiments were conducted as indicated above, MIC determined and was repeated.

Table 2: Minimum Inhibitory Concentration (MIC) of 6 compounds (μg/ml) for five monoconidial Penicillium spp. isolates that cause blue mold on apple fruit. MIC is defined as the concentration of chemical that halted conidial germination and prevented fungal growth four days post inoculation.

Table 3: Minimum Inhibitory Concentration (MIC) of three compounds (μg/ml) for 8 monoconidial Penicillium expansum and 2 P. solitum isolates that cause blue mold on apple fruit. MIC is defined as the concentration of compound that halted conidial germination and prohibited fungal growth four days post inoculation.

Morphological and genetic identification of Penicillium species

Fifteen Penicillium isolates, representing four different species from six U.S. apple producing states and one from the Netherlands, were obtained from a variety of sources (Table 1). The majority were isolated from decayed apple fruit with blue mold symptoms from five commercial apple cultivars, one from peach seed, and four from unknown sources. All isolates were identified using morphological and molecular methods as described [25,26]. A ~700 bp portion of the β- tubulin amplicon was subjected to MegaBLAST analysis and, was found to be 99-100% identical with 0.0 E values, to cognate Penicillium spp. from published sources, previously deposited in Genbank. Thus, confirming the morphological determination of each species (Table 1).

Inhibition of Penicillium species growth in vitro

Minimum Inhibitory Concentration (MIC) was observed four days post inoculation for benlate, chlorimuron ethyl and bialophos at 1, 5 and 250 μg/ml respectively for P. expansum, P. crustosum and P. solitum (Table 2). Penicillium carneum and P. paneum had high levels of resistance to bialophos and grew on 1000 μg/ml but had MIC of 1 and 5 μg/ml for benlate and chlorimuron ethyl. All Penicillium spp. isolates had abundant growth on three compounds G418, neomycin and phleomycin D1 that developed vigorously growing fungal colonies at 1000 μg/ml for all five Penicillium spp. isolates.

Variation in MIC amongst Penicillium expansum and P. solitum isolates

To evaluate variation between isolates, several P. expansum and P. solitum isolates, were selected and tested against three compounds that were most effective to inhibit Penicillium spp. growth. Ten single spore Penicillium spp. isolates (8 P. expansum and 2 P. solitum ) obtained from five different states (Maryland, New York, Oregon, Washington State, West Virginia and Pennsylvania,) including one international isolate (Netherlands), from five apple cultivars (‘Fuji’, ‘Golden delicious’, ‘Honeycrisp’, ‘Pink Lady’, and ‘Red Delicious’) isolated over multiple years from 1988-2011 were tested against three compounds (Table 2). Benlate, chlorimuron ethyl, and bialophos had MICs that were consistent between isolates. Benlate was the most inhibitory at 1 μg/ml for all isolates except for SA4 and RS1 (>10 μg/ml and 5 μg/ml respectively), inhibition by chlorimuron ethyl ranged from 5-10 μg/ml for all isolates, and bialophos was consistent at 250 μg/ml except for isolates F16 (750 μg/ml) and 16104 (500 μg/ml) (Table 3).

In this study, we conducted a screen of commercially available compounds to test against Penicillium spp. growth in vitro and to identify new materials that could be used as DSM’s for Penicillium spp. transformation. We focused on chemicals that have been effective in transformation of other fungal species (Ustilago maydis, Aspergillus flavus, Beauveria bassiana and Colletotrichum acutatum ) as DSMs with cloned corresponding fungal/bacterial resistance genes that function in different fungal pathogens [23,24,27-29]. We determined that G148, neomycin and phleomycin, were ineffective as fungal growth was observed for 5 different Penicillium spp. isolates on concentrations as high as 1000 μg/ml of each compound. This result was surprising for two of the three compounds (neomycin and phleomycin) which have similar modes of action to bleomycin that has been reported as an effective DSM for P. chrysogenum [30]. However, we found that benlate was excellent at inhibiting Penicillium spp. growth at low MIC values (~1 μg/ml) for the majority of Penicillium spp. isolates tested. This is not unexpected, since benlate is a member of the beta-tubulin inhibiting class of fungicides, in which thiabendazole (active ingredient in postharvest fungicide Mertect^®) is routinely used in drenches and dips to control blue mold of apple [26]. It is hypothesized that previous exposure to TBZ, which is used as a postharvest decay control, may have contributed to higher levels (>1 ppm) of tolerance in one isolate of P. expansum and one of P. solitum obtained from decayed apple fruit. However, a higher MIC may also reflect natural variation in the level of resistance in the fungus which also may be present.

Benlate, also marketed as Benomyl^® (DuPont), was used as a broad spectrum fungicide that inhibits the polymerization of β-tubulin monomers composing the cytoskeleton of the cell. Benlate, and corresponding β-tubulin resistance genes, have been successfully used as DSMs in other pathosystems for fungal transformation (i.e., U. maydis , A. flavus and P. expansum ). Polyethylene glycol-mediated (PEG) transformation of U. maydis was accomplished using benlate and the Tub gene, which functioned just as well as hygromycin and thus expanded the molecular toolkit for this fungus [23]. Previous work [31] demonstrated that benomyl could function as a DSM in P. expansum via PEG-mediated transformation. However, results from their study showed low transformation efficiency and unstable transformants following selection on benlate-amended medium. Our results show that benlate can be used for a variety of Penicillium spp. and adds to the current body of knowledge as we defined a MIC for benlate based on 15 isolates from five different Penicillium spp. Our findings serve as a platform for transformation and selection of Penicillium spp. carrying a benlate resistance gene from Aspergillus spp. [24]. The next logical step would be to build a binary vector containing a ben resistance gene flanked by the A. nidulans Trp promoter and terminators for use in combination with ATMT to ensure higher transformation efficiency and stable transformation of Penicillium spp.

Chlorimuron ethyl is the active ingredient in the herbicide Classic^® (DuPont) that inhibits acetolactate synthase mediating isoleucine and valine biosynthesis in plants. Interestingly, this compound is also inhibits fungal growth in vitro in Colletotrichum acutatum and Beauvaria bassiana and has been successfully used as a DSM [27,29]. PEG-mediated and ATMT systems have utilized the sur gene from Magnaporthe grisea which confers resistance to chlorimuron ethyl. PEG-mediated transformation has been used with this DSM, had acceptable transformation efficiency and stable transformation with little impact on cultural morphology and or virulence for C. acutatum [27]. The same system was used as a DSM but in an ATMT system to transform B. bassiana and achieved stable, high transformation efficiency [29]. Our studies are the first to show that this compound can also inhibit Penicillium spp. growth (5-10 μg/ml) and has established a guide for its use when transforming Penicillium spp. with a binary vector carrying the sur gene construct to screen Penicillium spp. transformants.

Bialophos is the active compound in the herbicide Ignite^® (Bayer Crop Science) which interferes with glutamine synthesis in crops. Bialophos, along with corresponding bar resistance gene from Streptomyces hyrgroscopicus , has been used as a DSM in the transformation of Sclerotinia sclerotiorum and Pleruotus ostreatus [28,32]. It is possible that bialophos may function successfully as a DSM in Penicillium spp. However, functionality may be limited as this compound exhibited the greatest variability in inhibiting Penicillium spp. growth in vitro . Based on our findings, we suggest that researchers use this compound with caution, and perform a series of MIC tests using their isolates of interest before undergoing fungal transformation with a bar construct in a given Penicillium spp. strain.

In summary, we were able to show effective inhibition of fungal growth for 15 different Penicillium spp. isolates from 5 different species using two different compounds which resistance genes have been cloned and characterized. These genes and corresponding constructs are available from the Fungal Genetics Stock Center located at Kansas State University and directly from researchers. The MIC values serve as a reference for others to evaluate these compounds to determine the precise concentration needed to screen putative Penicillium spp. transformants. In the future, we will couple our findings from this study with ATMT, which has been established [22]. The use of routine molecular methods (i.e., marker exchange/in fusion reaction) to swap the hph gene of the pPK2 binary vector [33] with the Ben-resistant gene from a closely related species like A. flavus [24] or the sur gene from M. grisea [27] would enable fungal transformation with both of these selectable markers. Construction of these vectors serves as the next logical step to transform various Penicillium spp. to determine frequency of integration, mitotic stability, and impact on cultural phenotype of Ben-resistant and Sur-resistant Penicillium spp. transformants. Additionally, one could also use the binary vector that was constructed for B. bassiana [29], to transform Penicillium spp. and select on chlorimuron-ethyl amended medium. We foresee that these new compounds, in combination with ATMT, will increase the molecular tool kit for Penicillium spp. that cause postharvest decay of pome fruit and facilitate future molecular genetics investigations to target multiple gene pathways involving fungal virulence, toxin production and fungicide resistance.

We thank Dr. Richard Kim and Dr. Robert Spotts for sharing Penicillium spp. isolates used in this study. This work was supported by base funds provided to USDA-ARS project plan #1245-42430-014-00D entitled “Methods for Rapid Identification and Functional Analysis of Fungi Causing Postharvest Decay of Pome Fruit” via National Program 303 Plant Diseases. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture.