A convenient workflow to spot photosensitizers revealed photo-activity in basidiomycetes

Photodynamic therapy (PDT) is an alternative approach for the treatment of neoplastic diseases employing photosensitizers activated by light. In order to discover new natural photosensitizers, a convenient workflow was established. To validate the workflow, fungi were selected, because we hypothesized that fruiting bodies and mycelia are an overlooked source. The results proved the hypothesis, as exorbitant high photo-cytotoxicity values were detected. For example, the acetone extract of Cortinarius croceus was characterized by an EC50, 9.3 J cm−2 of 1 µg mL−1 against cells of a lung cancer cell-line (A549). In sum, a low-cost workflow for the detection and biological evaluation of photosensitizers is presented and discussed. Furthermore, this paper provides the first experimental evidence for phototoxic metabolites in basidiomycetes. This hints towards a new assignable function of fungal pigments, i.e. photochemical defense.


General Experimental Procedures.
All used solvents for the preparation of the extracts were purchased from VWR International (Vienna, Austria). Solvents used for HPLC analysis were obtained from Merck (Darmstadt, Germany). Ultrapure water for the HPLC analysis was obtained employing a Sartorius Arium 611 UV water purification system (18MΩ/cm, Sartorius AG, Göttingen, Germany). Solvents used for NMR spectroscopy were purchased from Euriso-top SAS (Saint-Aubin Cedex, France). 9,10-Dimethylanthracene (DMA) was purchased from TCI (TCI Deutschland GmbH, Eschborn, Germany). Berberine was obtained from the stock of standards of the Institute of Pharmacology and Pharmacognosy, University Innsbruck. The Berberis ilicifolia root extract was prepared starting with the milled and dried plant material (5 g). The material was sequentially extracted with ethyl acetate (3 x 30 mL) and methanol (3 x 30 mL) in the ultrasonic bath (20 min). The collected extracts were concentrated under vacuum. Berberine was detected in the methanol extract and thus the methanol extract was used.

Extraction
In Table 1 an overview of the obtained yields is given.

DMA-Assay
The established DMA-assay is based on the photosensitized oxidation n of 1 O2 to 9,10-dimethylanthracene disrupting the dislocated π-system ( Figure S2) and thereby the typical absorbance of anthracenes.

Calculation of the Relative Singlet Oxygen Yield
The relative singlet oxygen yield was calculated as follows in equation (1).
This equation is derived from the "relative quantum yield measurement method". 1 Here the quantum yield of an unknown PS is calculated in comparison to a known one. Instead of quantum yields as outcome, we chose "relative singlet oxygen production" ( 1 O2 [%]) as output of our assay to obtain an intuitive measure for the putative phototoxic extracts. In detail, the reduction of the DMA-absorbance by illuminated berberine (EtOH) was normalized to "100% singlet oxygen". We utilized the absorbance of DMA, 2 instead of the commonly used fluorescence, 3 to avoid falsenegative results. The latter would be obtained, if a metabolite of the extract-matrix absorbs the emitted photons of DMA. As the fluorescence emission of DMA lies between at 400 and 450 nm, each yellow extract would lead potentially to a false-negative result. Furthermore, as visualized in equation (1), the difference of the DMA absorbance (i.e. ∆ ( )) was corrected by the difference in absorbance of the pure extract at λ = 377 nm (i.e. ODExtract, 377 nm). This accounts for potential changes in the OD due to photochemical processes happening in the extract. The 468 ( As correction, the intensity ratio (i.e. �. 4 The latter is an approximated expression of the former 1 and is only valid for OD < 0.2. This assumption is for the investigated extracts in the shortdistanced cavity of a 96 well plate generally not valid and therefore negated. Important to note, if another emission source is intended to be used, equation (S2) needs to be adjusted to the emission maximum of the other light source and therewith the measurement. Each measurement was done at least as triplicate of technical duplicates.

Analytical Conditions.
HPLC-DAD-MS experiments were performed on a 1260 Infinity II LC system (Agilent Technology, Waldbronn, Germany) equipped with degasser, binary pump, autosampler, and column heater. The system was fitted with a MS Detector from Agilent. Best separation for the Basidiomycetes was achieved using a Synergi 4µ MAX-RP 80Å (150 mm × 4.6 mm, Phenomenex, Torrance, CA, USA) and utilizing a mobile phase consisting of water (A) and acetonitrile (ACN) with 0.1% formic acid (FA) (B). The linear gradient started with a ratio of 90:10 (A to B), reached after 30 min a ratio of 50:50, after additional 4 min 10:90 and ended with this ratio after 50 min. The column was re-equilibrated for 20 min. Column temperature, flow rate, and sample volume were set to 40 °C, 0.5 mL/min, and 10 μL, respectively. The MS parameters were: split 1:1; API-ES, negative/positive-ion mode; spray voltage, 4.5 kV; dry temperature, 320 °C; drying gas flow rate, 12.00 L/min; nebulizer gas, 25 psi; mode, scan range: m/z 200 -1200. The ascomycete extracts were investigated with the same MS parameters. However, a different stationary phase (i.e. Luna Phenyl Hexyl column (150 mm x 3.0 mm, 3 µm)) and mobile phase were used. The solvent mixture (H2O : ACN + 0.1 % FA) started with a ratio of 90:10, reached after 30 min 50:50, after additional 15 min 90:10 and in the last ten minutes 98:2.

Interpretation of the Obtained Chromatograms
Basidiomycetes. The main peak of T. atrotomentosa ( Figure S3, tr = 23 .0 min) correlated with a mass over charge value (m/z) of 323.1 ( Figure S5) in the negative mode of the ESI-MS experiment and two absorption maxima at 266 and 371 nm, respectively ( Figure S5). Hence, the main pigment was identified as atromentin ( Figure S3), 5, 6 which was already described by Kögl et al. in 1924. 7 The HPLC-chromatogram of C. croceus was more complex than the one of T. atromentosa ( Figure S3). The major peak in the acetone fraction with a retention time of 43.64 min showed a m/z value of 565 for [M-H] -( Figure S6). This, in combination with the taxonomic description of Cortinarius subgenus Dermocybe 8 and the electron-absorption properties (i.e. λmax = 284s, 310sh, 440br nm, see Figure S6) led to the identification of 7,7'-biphyscion ( Figure S6) as mainpigment of these fruiting bodies. Interestingly, this dianthraquinone is not genuinely present in this fungus. Rather it is an oxidation product from a flavomannin-precursor. 9 The identification of the main pigment of the third basidiomycete, I. obliquus, was not possible with the current experimental setup. For this organism, the HPLC-chromatogram showed one main peak in the acetone fraction with a retention time of tR = 25.2 min, which was surrounded by several minor peaks (see Figure S3). The major peak was characterized with m/z = 379.2 [M-H]and the following electron-absorption properties, i.e. λmax = 412 nm. None of the pigments known for I. obliquus, i.e. phelligridins or innonoblins, 10, 11 correlated with these characteristics.
Ascomycetes. The chromatograms detected at 468 nm are depicted for all fungi in Figure S4. For B. brongniartii one peak was detected in the polar region at tR = 17.2 min. This peak was assigned to the red pigment of B. brongniartii (i.e. oosporein) via comparing its retention time, UV-Vis spectrum (λmax = 288, 501 nm) and MS-spectrum (m/z (-) = 305.1) with the reference compound oosporein ( Figure S7).
The orange-green extract of M. brunneum was characterized by a small peak eluting at tR = 11.8 min. The UV-Vis spectrum showed a λmax = 410 nm in the visual range (Figure S8 left). The ionization of the metabolite was not good enough to gain a reliable m/z value for this metabolite, most likely it is an unknown one.
The R. subterranea extract (see Figure S7) is due to its complex pigment pattern the most interesting one. The three main pigments absorbing at 468 nm elute at tR = 29.4 min, 34.8 min, and 47.9 min. The UV-Spectra are characterized by a broad absorbance band between 330 and 500 nm (see Figure S8 right), which is characteristic for phenalenone like structures and fits to the vast number of poly-hydroxylated phenalenones and phenalenone-like structure described for Roesleria and related fungi [12][13][14] .

Experimental Details of the Photo-cytotoxicity Test
Briefly, cells (2000/well) were seeded in Opti-MEM® (2.5% FCS, P/S) and after 24 h treated with each extract (50, 25, and 5 µg/mL, stock solution 10 mg/mL in DMSO, max DMSO 0.65%). After additional 24 h, the medium was aspirated and replaced by fresh Opti-MEM ® . Thereafter, the plate was irradiated for the indicated amount of time (i.e. 0, 2.5, 5, or 7.5 min; 0, 3.1, 6.2, or 9.3 J/cm² respectively). The cells were fixed with chilled trichloracetic acid (100 µL) after 72 h in total. After washing (4 times 200 µL, water) the cells were stained with sulforhodamine B (SRB, 0.4 % in acetic acid (1 %)) for 30 min. Then, the plates were washed again (4 times 200 µL, acetic acid 1 %) and tried under air. The dried dye was dissolved with a TRIS solution (10 mM in water, 100 µL) and the absorbance measured at 540 nm with a plate reader (Tecan, Spark, M10). The EC50 of berberine was calculated with Prism 5.0 employing the relative Hill-Slope equation and is given with its confidence interval (95 %).

Results of the (Photo)-cytotoxicity Experiment
The detailed results of the (photo)-cytotoxicity experiment are given in Table S2.

The Selectivity Index
The selectivity indices of the extracts were calculated as discussed in the experimental part and are presented in Figure S9 Table 2). For these extracts, the calculated SI-values are therefore likely an underestimation of the real SI.

Interpretation of the Observed Induced Cell Death
A systematic investigation of the morphological changes induced by the extracts and the irradiation is depicted in Figure 3 and Figure S10-S15. Lung cancer cells treated with the AC extract of C. croceus ( Figure 3A, Figure S14) did significantly shrink and do otherwise show all hallmarks of an apoptotic cells death indicating that the PS accumulated primarily within the organelles. In contrast, cell populations treated with I. obliquus PE extract showed dispatched cells, called balloon-cells, and a lot of cell debris ( Figure S13). The swelling of the cells is a typical observation for membrane modulating agents, which break down the cellular membrane and therewith yield the loss of the osmotic balance. The PE extract of T. atrotomentosa induced as well in the dark as in the light a programmed cell death as observable by the shrunken cells and the absence of debris. In addition, the irradiated cells treated with R. subterranea showed apoptosis-typical membrane blebbing, whereas the extract of B. brongniartii induced rather an antiproliferative effect as the number of shrunken cells does not out count the number of still normal cells in the irradiation experiment. M. brunneum does not show any distinct conspicuity.
In sum, the phototoxic evaluation of the fungal extract indicated the existence of promising fungal PS with impressive selectivity indices and a putative apoptotic mode-of-action. The latter is clearly medicinally favored, due to its lack of inflammatory processes.