AflJ belongs to a unique family of proteins found only in fungi (Figure 1). The protein is predicted to possess three membrane-spanning domains. Only AflJs from A. parasiticus (and other section Flavi species) and A. nidulans (A para and A nid in Figure 1) contain putative C-terminal microbodies-targeting sequences [21]. The aflJ orthologs that were chosen for complementation analysis are associated with gene clusters involved in sterigmatocystin and monodictyphenone gene clusters of A. nidulans [28], and the dothistromin gene cluster of Dothistroma septosporum [29] (Figure 1). aflJ orthologs are also found in the cercosporin gene cluster from Cercospora nicotianae [30] and in partial clusters from A. terreus, A. fumigatus, Coccidioides immitis, and Penicillium marneffei [31]. The A. flavus, A. parasiticus, A. nidulans and D. septosporum aflJ homologs are believed to code for proteins with a similar function in biosynthesis since the end metabolites are related to aflatoxin precursors. The function of the predicted AflJ homologs in the other species is unknown but, in each, the aflJ ortholog adjoins a Zn₂Cys₆ factor gene that resembles AflR.

Complementation of A. parasiticus ΔaflJ was performed with constructs carrying the following aflJ homologs: ANID7819 (A. nidulans sterigmatocystin cluster), ANID_10021 (A. nidulans monodictyphenone cluster), Ds-aflJ (D. septosporum dothistromin cluster (personal communication, R. Bradshaw), and AflJ (AAS90096) from A. flavus (positive control for complementation). Of these, only aflJ from A. flavus complemented the deletion of aflJ in A. parasiticus ΔaflJ as evidenced by the ability of such transformants to produce of OMST (Figure 2).

We confirmed, by end-point RT-PCR, the earlier study of Meyers, et al. [21] that some aflatoxin gene transcripts accumulated in the A. parasiticus-ΔaflJ mutant and in the transformants of ΔaflJ with putative aflJ homologs (Figure 3a). ΔaflJ and the other transformants expressed the aflatoxin biosynthesis genes hexA, pksA, and ver-1 but not omtA. PCR studies used primer sets that flanked introns in the genes and the sizes of PCR products from cDNA were smaller than those products from amplification of DNA or unprocessed RNA. The control for these reactions was the PCR done in the absence of reverse transcriptase (Figure 3a). We confirmed that the introduced aflJ homologs, were processed correctly when placed under the control of the gpdA promoter (shown only for transformants expressing gpdA-Ds-aflJ, Figure 3b).

As a transcriptional co-activator that binds AflR we expected that AflJ would co-localize with AflR at aflatoxin gene promoters that are recognized by AflR. ChIP analysis was performed on A. parasiticus and A. flavus mutants transformed with gpdA-c-myc::aflJ to determine if binding of AflJ could be detected at four different aflatoxin biosynthesis gene promoters. Fusion of AflJ with the cMyc epitope allowed recovery of the chromatin-bound protein by use of highly purified commercial antibodies to c-Myc. The fusion protein was used because we were unable to develop antibodies to AflJ or AflR that were suitable for chromatin immunoprecipitation. Since AflJ was previously shown to bind to AflR, we reasoned that c-Myc::AflJ would bind to AflR and allow immunoprecipitation of DNA in chromatin in promoter regions of genes from fungi grown in media conducive for aflatoxin production. As a positive control, ChIP analysis was also performed on A. parasiticus and A. flavus mutants transformed only with gpdA-c-myc-aflR. Although binding of c-Myc::AflR was detected at known AflR-binding sites in the aflJ, pksA, fasB, and ver-1 promoter regions (Figure 4), binding of c-Myc::AflJ to these promoters could not be detected at a level significantly different from that of the no-antibody control.

Using a yeast two-hybrid assay, Chang demonstrated that AflJ binds to AflR [18] and binding required essentially intact AflJ. In order to obtain additional information on possible interactions of AflJ with other A. parasiticus proteins, we used a 24 h A. parasiticus cDNA library as prey and A. parasiticus aflJ as bait in a yeast two-hybrid (Y2H) assay (Table 1).

Some of the clones selected in the Y2H experiment are predicted to encode proteins associated with components of the COP9 signalosome and a putative Nedd8-activation. Others detected in the assay, including Con-10 and SteA, are proteins known to be involved in developmental processes in fungi, while another (AflL) is a cytochrome P450, a type of enzyme typically associated with membranes in peroxisomal organelles. Although AflR was not found among the interacting clones, a different Zn₂Cys₆ factor AmdR was detected. When tested separately using yeast transformants expressing AflR as bait and AflJ as prey as was done by previously [18], we were able to confirm the ability of these proteins to interact as demonstrated by the formation of yeast colonies on selective medium (Table 2).

Split enhanced yellow fluorescent protein (eYFP) assays (see below) also confirmed that AflJ is able to bind to AflR. Taken together, these results suggest that AflJ, besides binding to AflR, is able to bind to proteins associated with peroxisome-like vesicles.

Plasmids containing aflJ fused to the jellyfish green fluorescent protein (aflJ-GFP) were transformed into A. parasiticus ΔaflJ. These were used to investigate the intracellular localization of AflJ in live A. parasiticus cells. The GFP constructs were prepared with either the glycerol phosphate dehydrogenase (gpdA) or the aflJ promoter driving gene expression. Images of the intracellular localization of AflJ-GFP in live hyphae were taken at various times after growth in PDB liquid medium (Figure 5). At 24 h, large fluorescent structures were observed in the cytoplasm. These large structures seem to represent protein aggregates (Figure 5a). Generally, these fluorescent aggregates do not co-localize with DAPI-stained nuclei. At later times (48 h) AflJ-GFP localized both to aggregates and to small dot-like structures reminiscent of endosomes in approximately 30% of examined cells (Figure 5b). When the aflJ promoter was used to drive aflJ expression, most of the green fluorescence still co-localized with endosomes rather than with the DAPI-stained (blue fluorescent) nuclei (Figure 5c).

A similar fusion construct was created with the gpdA promoter driving expression of aflR fused to GFP. A. parasiticus transformants containing this plasmid also showed fluorescence in cellular aggregates that were similar to those seen in the AflJ-GFP fusion studies (Figure 5d). Both discreet fluorescent aggregates and smaller dots consistent with endosomes were visible in the mycelia. The majority of the GFP fluorescent signal did not appear to co-localize with the DAPI-stained nuclei.

Because previous studies using the yeast two-hybrid assay concluded that AflJ interacts with AflR in nuclei [18], we investigated the interaction of AflJ with AflR using a split eYFP assay. Transformants containing both the amylase gene promoter (amyB) driving expression of the N-terminal half of the eYFP coding sequence fused to aflR (amyB-Nt-eYFP-aflR) and the amyB promoter also driving expression of the C-terminal half of the eYFP coding sequence fused to aflJ (amyB-Ct-eYFP-aflJ) were grown in either maltose (inducing) or glucose (non-inducing; data not shown) minimal media. Only induced co-transformants showed Ct-eYFP::AflJ and Nt-eYFP::AflR interaction as evidenced by fluorescence within small and large dots consistent with vesicles and endosomes within the mycelia (Figure 5e) similar to those observed for the AflR-GFP transformants. Only a few of the dots (the smaller ones mainly) co-localized with DAPI-stained nuclei when the cultures were grown on minimal media containing maltose.

To investigate if AflJ is involved in endosome/aflatoxisome biogenesis we examined wild-type and ΔaflJ A. parasiticus cultures grown for 44 h or 68 h cultures using bright field microscopy. At both time points, wild-type cultures, grown similarly, produce aflatoxins. We hypothesized that disturbance of endosome phenotype could indicate changes in biogenesis of endosomes and/or trafficking of endosomal proteins that are involved in aflatoxin biosynthesis. As part of the analysis, parameters such as the number, size, timing of appearance, and sub-cellular localization of structures consistent in size and shape with endosomes were characterized. In previous work we obtained biochemical and genetic evidence that these structures are endosomes [25].

At 44 h in wild-type A. parasiticus, endosomes were just starting to form; they were small and were present in <10% of the cells (data not shown), whereas in the ΔaflJ mutant, at this same time, nearly 50% of the cells carried distinct endosomes (Figure 6a,b). By 68 h, larger more elongated clusters (Figure 6c) of organelles were observed in more than 50% of the cells from both the wild-type and the mutant. In the wild-type these appeared mainly as discreet units. However, for a small number of examined mycelia, endosomes were detected in small clusters located primarily near a septum. The number of endosomes per cluster ranged from 3 to 6, and the number of endosomes per cell ranged from 30 to 46 (Figure 6c). In ΔaflJ, by 68 h, large clusters of organelles were observed in 75% of cells and were located near septa, as in the wild type, and in the middle portions of the cells. The number of endosomes per cluster ranged from 5 to 14.

A. parasiticus SU-1 (ATCC 56775) and A. parasiticus BN009E (BN9) [45] were used as wild type aflatoxin producing species and for preparation of transformants for the split eYFP (BiFC) studies, A. parasiticus SRRC2043::ΔaflJ (ApΔaflJ) [18] was used for complementation studies and preparation of some of the transformants in which aflJ or aflR was fused to GFP coding sequence. For the ChIP studies A. flavus CA14 Ku70⁻ [46] was used to prepare the transformants. Cultures were maintained on V8 agar (5% V8 juice/2% agar) [47] and grown on YES (2% yeast extract and 6% sucrose, pH 5.8) or potato dextrose agar (PDA) or broth (PDB) liquid medium for production of aflatoxin or its precursor, OMST. Maltose was substituted for glucose in A&M-medium for growth of transformants expressing the split eYFP fusion proteins. Conidiospores (spores) were inoculated into liquid growth medium at approximately 10⁴ spores per mL and incubated for appropriate time periods at 30 °C with shaking at 150 rpm in the dark (standard growth conditions).

For ChIP analysis, each recombinant protein was tagged at the N-terminal end with a c-Myc epitope tag (EQKLISEEDL) by first cloning the gene into the Clontech vector pGBKT7 followed by recovering the tag and gene by PCR with appropriate primers. After recloning the PCR product into pPTRI-gpd::trpC [48] and transforming into A. flavus CA14, ChIP was performed on mycelia grown on PDB for 18 to 90 h, using a modification of the procedure outlined in the product brochure for Millipore EZ ChIP™ (Millipore, MA, USA). Mycelia were treated with formaldehyde (final concentration of 1%) for 10 min and cross-linking was terminated by adding glycine to a final concentration of 0.125 M. After rinsing with phosphate buffered saline (PBS), the mycelia were ground to a powder in liquid nitrogen. Ground mycelia (100 mg) was suspended in 1 mL SDS lysis buffer (1% SDS, 10 mM EDTA and 50 mM Tris, pH 8.1), PMSF (1:100 of 0.125 M stock in iso-propanol) and protease inhibitor cocktail (1:100; Sigma-Aldrich, St Louis, MO, USA; catalog number P8215). The suspension was sonicated (Sonics Vibracell Newtown, CT, USA; catalog number VCX130) with the 3 mm microtip on a freezer block at an amplitude of 25% for 4 pulses of 30 s and a 30 s rest between pulses. The average product size after sonication was 300 to 400 bp. After centrifugation for 10 min at 14000 × g at 4 °C, the supernatant (200 μL) was diluted with 1.8 mL ChIP Dilution Buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl, pH 8.1, 167 mM NaCl with PMSF and protease inhibitors to 200 μL). From each sample 2 μL was saved for determination of the concentration of input chromatin DNA. Before immunoprecipitation, the sample was precleared by adding 80 μL salmon sperm DNA/Protein G Agarose mixture (Millipore, Billerica, MA, USA) and incubating on a rocker platform for 1 h at 4 °C. After centrifugation for 1 min at 5000 × g to remove beads 5 μL anti-c-Myc (abcam ab9132) was added to samples and incubated overnight at 4 °C. A sample was also incubated identically without including the antibody. Detection was with anti-c-Myc (abcam Eugene OR, ab9132) antibody (Ab) by incubation overnight at 4 °C. The negative control was the sample with no Ab added. Chromatin bound to c-Myc tagged protein was precipitated by adding 60 μL salmon sperm DNA/Protein G Agarose beads for samples that received anti-c-Myc Ab. Incubation was for 1 h at 4 °C. The beads were harvested by centrifugation at 1000 rpm for 1 min and washed for 4 min at 4 °C with 1 mL of Low Salt Immune Complex Wash Buffer, High Salt Immune Complex Wash Buffer, LiCl Immune Complex Wash Buffer (reagents supplied with a Kit from EMD Millipore, Darmstadt, Germany), and TE Buffer (2 washes). The beads were then re-suspended in 100 μL elution buffer (50 mM Tris-Cl, pH 7.5, 10 mM EDTA, 1% SDS) and incubated for 10 min at 65 °C, centifuged as done previously, and the supernatant collected. The extraction was repeated once and 20.8 μL of 5 M NaCl (200 mM NaCl final) and 1 μL proteinase K (20 mg/mL stock) were added. The solution was first incubated at 45 °C for 2 h and then at 65 °C overnight to reverse crosslinks. The DNA was purified using a QIAquick PCR purification kit (Qiagen, Valencia, CA, USA) and qPCR was performed using 2 μL of the 100 μL recovered DNA as the template. qPCR was conducted on the recovered DNA and input DNA using the Power Sybr Green Kit (Applied Biosystems, Carlsbad, CA, USA; Catalog # 4367659) and 40 cycles of amplification. qPCR determinations followed the guidelines suggested by Bustin, et al. [49]. Oligonucleotides were designed using Vector NTI to give amplicons ranging from 100 to 150 bp (Table 1). Primer efficiency for DNA amplification was determined to be 80% to 110%.

The coding sequences of putative aflJ homologs from A. nidulans and D. septosporum were PCR-amplified from genomic DNA of these species using oligonucleotide primers to the N- and C-terminal coding sequences of each gene. After cloning into pPTRI-gpdA-trpC, A. parasiticus ΔaflJ was transformed with the resulting plasmid DNA as previously described [48], and pyrithiamine-resistant transformants selected. The ability to produce OMST by these transformants was assessed by thin layer chromatography (TLC) on Baker, Phillipsburg, NJ Silica250 plates after elution with toluene:ethylacetate:acetic acid (6.5:3.5:1).

Yeast two-hybrid studies were performed using the Matchmaker Library Construction and Screening Kit (Clontech, Mountain View, CA, USA). aflJ from A. parasiticus BN9 was used as the bait by insertion of the gene into plasmid pGBKT7 while cDNA from BN9 grown for 48 h on PDA medium was introduced into pGADT7-Rec to generate the GAL4 activation domain plasmid. All procedures were conducted according to published Matchmaker protocols. Selection of Saccharomyces cerevisiae AH109 colonies was achieved by growth for 5 days at 30 °C on SD/-Ade/-His/-Leu/-Trp (quadruple dropout, QDO) agar. Colonies were analyzed by PCR using the Matchmaker 5' and 3'AD LD-insert screening primers specific to pGADT7. The PCR products were sequenced and the sequences analyzed by BLAST analysis of the non-redundant DNA database in GenBank as well as against the A. flavus NRRL3357 sequenced genome.

For subcellular localization of AflJ and AflR at different time points after germination, A. parasiticus ΔaflJ transformants with pUC18-gpdA promoter-aflJ::GFP-nmt1 terminator or pUC18-gpdA-aflR::GFP-nmt1 were prepared, where aflR::GFP and aflJ::GFP are in-frame fusions of the coding sequences for aflJ or aflR, and the coding sequence for the enhance green fluorescent protein modified from Aequorea victoria. For split eYFP analysis [37], the vectors used to introduce aflR and aflJ were pMCB17apx-amyB-Ct-eYFP-trpC and pPTRI-amyB-Nt-eYFP-trpC. The former plasmid contains the maltose-inducible amyB promoter [50], and the C-terminal (Ct) portion of eYFP followed by the A. nidulans trpC transcriptional terminator, while the latter plasmid contains the amyB promoter, the N-terminal portion of eYFP and the trpC terminator. aflR was inserted in-frame after the Nt-eYFP sequence while aflJ cDNA was inserted in-frame after the Ct-eYFP sequence. The two plasmid constructs were cotransformed into A. flavus CA14 pyrGΔ,Ku70Δ and transformants containing both vectors were selected by growth on Czapek’s medium containing 0.1 μg/mL pyrithiamine. The negative control for the split eYFP assays was a cotransformant containing Nt-eYFP-aflR and Ct-eYFP-veA. The positive control was a cotransformant containing Nt-eYFP-laeA and Ct-eYFP-veA [37]. Transformants were grown on minimal media with either glucose (non-inducing) or maltose (inducing) as the sole carbon source. Fluorescence was examined on a Nikon, Melville, NY, Eclipse E600 microscope with either a Endow GFP HQ filter cube (Ex 450–490, DM 495, BA 500–550 or a Yellow HYQ filter cube (EX 490–510, DM 515, BA 520–550). A 10× eyepiece and a 40× objective were used for most microscopy. Nuclei were stained with 1 μg/mL DAPI (4',6-diamidino-2-phenylindole) and examined using the UV-2E/C filter cube (EX 330-380, DM 400, BA 435-485). For bright-field microscopy, hyphal filaments were observed under a Nikon Eclipse E600 microscope. Computer-assisted image analyses of endosomes in wild-type and ΔaflJ A. parasiticus were performed using CMEIAS^© Ver. 1.27 (Center for Microbial Ecology Image Analysis System, MSU, MI, USA) [32].