Comparative analysis of the effects of chemically transformed polyene antibiotics pimaricin, nystatin, lucensomycin, amphotericin B, and levorin on biological objects in vivo and in vitro revealed the greatest biological activity of original amphotericin B and levorin with its derivatives. The study also examined the effects of alkyl derivatives of amphotericin B and levorin modified in certain parts of the lactone ring on the lipid and biological membranes. It is established that methylated levorin possesses larger biological activity than the original antibiotic. Examination of the effects of alkyl derivatives of levorin and amphotericin B on cell cultures C6 (rat glioma) and HeLa (human cervical carcinoma) in vitro revealed the antitumor action of methylated levorin and original amphotericin B.

Fungal keratitis is one of the most challenging types of microbial keratitis for the ophthalmologist to diagnose and treat. Fungi causing human keratitis take the form of either yeasts or mold. Candida, the major pathogenic yeasts, can be detected in the normal ocular surface flora. Preceding ocular surface disorder, the wearing of contact lenses and the use of antibiotic/steroid eye drops may lead to candida keratitis. Infectious focus caused by Candida tends to melt the corneal stroma. Keratitis caused by mold often develops after an injury caused by soil and/or a plant. Mold can reach the anterior chamber without destroying the stromal layer of the cornea, which results in distinctive clinical features such as endothelial plaque and hyphate ulcer. Fungal keratitis needs to be managed by antifungal agents, most of which must be prepared by ourselves to apply to the ocular surface. Candida keratitis should be managed with azoles. If the infection seems to be caused by mold, several antifungal drugs including pimaricin, which is the only agent officially applicable to the eye, should be used. Some cases of mold keratitis need to have therapeutic penetrating keratoplasty because of their lack of response to intensive medication. Mold causing keratitis is variegated. Fusarium and Aspergillus can reach the intraocular space rapidly. Alternaria and some other unclassified molds remain in the superficial layer of the cornea for a long time. Our experiments indicate that the progress of focus in the cornea is regulated by the receptiveness of mold against temperature.

Cysts of 2 isolates of Acanthamoeba from the cornea of 2 patients with confirmed Acanthamoeba keratitis were tested in vitro for sensitivity to antimycotic agents such as fluconazole, miconazole, amphotericin-B, pimaricin, antiprotozoal agents such as pentamidine isetionate and antiseptics which could be use in the ophthamological region. Pimaricin was the most successful cysticidal agent against the two strains. Sensitivity to pentamidine isetionate showed variation. Fluconazole, miconazole and amphotericin-B were resistant against cysts with concentration of eye drops that have been used in the treatment of Acanthamoeba keratitis. It was supposed that 5% pimaricin eye drops could be use in the treatment of Acanthamoeba keratitis in addition to keratomycosis. Pentamidine isetionate which belong to the diamidine family, is not yet clear as to the side effects to corneal epithelium cell, but we believe that this drug could be expected as a new therapeutic agent for Acanthamoeba keratitis.

Phytophthora nicotianae Breda de Haan was isolated from turning tomato fruit (Solanum lycopersicum L.) in August 2010 from a garden in central New Mexico. Symptoms typical of buckeye rot including brown, water-soaked, necrotic lesions with concentric rings were observed on three tomato fruit. Tissue from each fruit was surface sterilized and plated onto water agar and incubated at room temperature. After 72 h, colonies of Phytophthora (identified by the presence of coenocytic hyphae and papillate sporangia) were found and subcultured by hyphal tips to V8 agar amended with ampicillin (250 mg/liter), rifampicin (10 mg/liter), and pimaricin (0.2% wt/vol). The isolates of Phytophthora were identified as P. nicotianae based on morphological characteristics and DNA analysis. Sporangia were sharply papillate, noncaducous, and ovoid to spherical. The average sporangium size was 44.5 × 35.5 μm with a length-to-width ratio of 1.26. Chlamydospores, both terminal and intercalary, were spherical to ovoid and averaged 38.9 × 37.5 μm. PCR amplification and sequence analysis on three isolates from the infected tomato tissue was performed using primers ITS4 and ITS6 that amplify the 5.8S rDNA and ITSI and ITSII internal transcribed spacers (1,2). A band of approximately 890 bp was amplified and directly sequenced (GenBank Accession No. HQ711620). A BLAST search of the NCBI total nucleotide collection revealed a 100% similarity to multiple P. nicotianae isolates previously sequenced. Pathogenicity tests with sequenced P. nicotianae isolates were performed to confirm virulence on tomato fruit. Tomatoes were surface sterilized with 95% ethanol and 0.1 ml of a P. nicotianae zoospore suspension (10,000 zoospores/ml) or sterile water was pipetted onto the surface of the tomato fruit. After 5 days in a humidity chamber, all three inoculated tomatoes had expanding water-soaked, circular lesions and the negative control showed no disease symptoms. P. nicotianae was successfully reisolated from the inoculated tomato tissue and the ITS region was sequenced to confirm its identity. Although the disease has been reported in many other states since the early 1900s, to our knowledge, this is the first report of P. nicotianae causing disease on tomato in New Mexico. References: (1) D. E. L. Cooke and J. M. Duncan. Mycol. Res. 101:667, 1997. (2) T. J. White et al. Page 315 in: PCR Protocols: A Guide to Methods and Applications. M. A. Innis et al., eds. Academic Press, San Diego, 1990.

During July and August 2020, symptoms of leaf yellowing and browning, sudden wilting, and death were observed on industrial hemp plants (Cannabis sativa L.) in several drip-irrigated fields in Yuma and Graham county, Arizona. About 85% of plants showed severe crown and root rot symptoms. A high percentage of affected plants collapsed under intensive heat stress. Shriveled stem tissue with necrotic lesions can often be seen at the base of the plant, extending upwards more than 5 cm. Internal tissue of main stem and branches was darkened or pinkish brown. Outer cortex of root bark was often completely rotten, exposing the white core. Cottony aerial mycelium was visible on the surface of stalk of some of the infected plants in two fields in Yuma. To identify the causal agent, a total of twenty symptomatic plants were collected from several fields across the state. Crown and root tissues from affected plants were harvested and rinsed in tap water to remove soils. Approximately 2 to 4 mm tissue fragments were excised from the margins of the affected stem and root lesions, surface sterilized in 0.6% sodium hypochlorite for 1 min, rinsed copiously in sterile distilled water, blotted dry, and plated on potato dextrose agar (PDA), and on oomycete-selective clarified V8 medium containing pimaricin, ampicillin, rifampicin, and pentachloronitrobenzene (PARP). Plates were incubated at room temperature for 2 days. Sixteen isolates were recovered and their mycelial colonies resembled the morphology of Pythium. Based on the culture morphology on V8 medium, all isolates were tentatively identified as P. aphanidermatum with fast-growing, aseptate hyphae ranging from 3 to 7 μm in width, globose oogonia ranging from 25 to 31 μm in diameter, barrel-shaped antheridia, globose oospores ranging from 15 to 21 μm in diameter (10 measurements) (Watanabe, 2002). Genomic DNA was extracted from mycelial mats of three isolates using DNeasy Plant Pro Kit (Qiagen Inc., Valencia, CA) according to the manufacturer's instructions. The internal transcribed spacer (ITS) region of rDNA was amplified with primers ITS1/ITS4 and three nucleotide sequences were obtained. All three sequences were identical and deposited under accession number MW380253 in GenBank. A BLASTn search revealed that MW380253 had a 100% query coverage and 100% match with sequences MK611609.1, KJ162355.1, and AY598622.2, obtained from isolates of P. aphanidermatum. To fulfill Koch's postulates, pathogenicity tests were conducted with 2 isolates using 12 seeds of a hemp line 14 sown in 12 1.9-liter pots filled with a steam-disinfested potting mix. Pots were placed in a plastic container and watered three times a week by flooding, to create waterlogged conditions. Plants were maintained in a greenhouse supplemented with artificial lighting of 14 h/10 h day/night light cycle. Plants were fertilized weekly with a 20-20-20 fertilizer at 1mg/ml. Three weeks after sowing, four plants were inoculated with each isolate by drenching each plant with 200 ml of a 1×105 zoospore/ml suspension. Four plants, serving as control, received each 200 ml of distilled water. Symptoms of leaf chlorosis, crown and root rot, and wilting were observed 3 weeks afterwards, while control plants remained asymptomatic. P. aphanidermatum were re-isolated from necrotic roots of inoculated plants, but not from control plants. P. aphanidermatum was previously detected on industrial hemp in a research plot in Indiana (Beckerman et al., 2017) and is also known to affect other crops in Arizona during the summer months as well (Olsen & Nischwitz, 2011). This report is the first publication documenting P. aphanidermatum on field grown hemp in Arizona. Industrial hemp (Cannabis sativa) is an emerging crop in Arizona. The first plantings of hemp were in June of 2019, where 5,430 acres of hemp was planted in thirteen counties in Arizona before the end of the year. The Arizona Department of Agriculture Industrial Hemp Program, 2019 Year End Report confirms that nearly one-quarter of all hemp planted in 2019 did not receive a final state inspection due to crop loss. This disease is a potential constraint to hemp production in hot, arid climates, where copious water is used in combination with plastic mulch and/or drainage is poor.

Keywords: Cannabidiol; Damping off; Seedling disease; waterlogging.

During August and September 2020, symptoms of leaf chlorosis, stunting, and wilting were observed on indoor hemp plants (Cannabis sativa L. cv. 'Wedding Cake') in a commercial indoor facility located in Coolidge, Arizona. Plants were grown in soilless coconut coir growing medium (Worm Factory COIR250G10), watered with 1.5 to 2.1 liters every 24 h through drip irrigation, and supplemented with 18 h of lighting. About 35% of plants displayed symptoms as described above and many symptomatic plants collapsed. To identify the causal agent, crown and root tissues from four symptomatic plants were harvested and rinsed with tap water. Tissue fragments (approx. 2 to 4 mm in size) were excised from the margins of the stem and root lesions, surface sterilized in 0.6% sodium hypochlorite for 1 min, rinsed well in sterile distilled water, blotted dry, and plated on potato dextrose agar (PDA) and on oomycete-selective clarified V8 media containing pimaricin, ampicillin, rifampicin, and pentachloronitrobenzene (PARP). Plates were incubated at room temperature (21-24 oC). Five isolates resembling Pythium were transferred after 3 days and maintained on clarified V8 media. Morphological characteristics were observed on grass blade cultures (Waterhouse 1967). Grass blades were placed on CV8 inoculated with the isolate. After a 1-day incubation at 25°C, the colonized blades were transferred to 8 ml of soil water extract in a Petri dish. Ten sporangia and oogonia were selected randomly and their diameters were measured under the microscope. Sporangia were mostly filamentous, undifferentiated or inflated lobulate, ranging from 7 to 17 µm in diameter. Knob-like appressoria were observed on branching clusters. Bulbous-like antheridia were formed on branched stalk with 1-8 antheridia per oogonium. Globose oogonia were terminal or intercalary and ranged from 21 to 33 µm in diameter. Globose oospores were mostly aplerotic and ranged from 15 to 21 μm in diameter. Based on these morphological characteristics, isolates were tentatively identified as Pythium myriotylum (Watanabe, 2002). Genomic DNA was extracted from mycelial mats of two isolates using DNeasy Plant Pro Kit (Qiagen Inc., Valencia, CA) according to the manufacturer's instructions. The internal transcribed spacer (ITS) region of rDNA was amplified with primers ITS1/ITS4 and two identical nucleotide sequences were obtained and deposited under accession number MW380925. A BLASTn search revealed ≥ 98% query coverage and 100% match with sequences HQ237488.1, KY019264.1, and KM434129, which were isolates of P. myriotylum from palm, tobacco, and ginger, respectively. To fulfill Koch's postulates, pathogenicity tests were conducted with 2 isolates using plants of 'Wedding Cake' grown in 12 1.9-liter pots filled with a steam-disinfested potting mix (Sungro Professional Growing Mix). Pots were placed in a plastic container and watered to flooding three times a week. Plants were maintained in a greenhouse with 18 h/10 h day/night supplemental light cycle (15-28 oC). Plants were fertilized weekly with Peters Professional fertilizer at 1mg/ml. Four plants were inoculated with each isolate at three weeks after seed sowing by placing two 5-mm mycelial plugs from active growing 4 days-old cultures on PDA media adjacent to the main root mass at an approximately 3 cm depth. Four plants were inoculated with blank PDA plugs as controls. Symptoms of leaf chlorosis, crown rot and wilting were observed after four weeks while control plants remained symptomless. P. myriotylum was re-isolated from necrotic roots of inoculated plants after surface-sterilization, but not from control plants. The pathogenicity test was repeated once. While P. myriotylum often occurs in warmer regions and has a wide host range of >100 host plant species including numerous economically important crops (Wang et al., 2003), there are only two reports of this pathogen on indoor hemp plants in a greenhouse in Connecticut (McGehee et al., 2019) and in Canada (Punja et al., 2019). This is the first report of P. myriotylum causing root and crown rot of indoor hemp in Arizona. A more careful water management in soilless growth medium to reduce periods of saturation would minimize the risk of Pythium root rot in indoor hemp production.

Keywords: Damping-off; marijuana; overwatering; seedling diseases.

Industrial hemp (Cannabis sativa) is a newly legal crop in California that is grown for cannabidiol oil, fiber and seed. In August 2019, whole plant decline and root rot were observed affecting <5% of plants in two industrial fields in Fresno County, CA. Symptoms included chlorotic, collapsed foliage, stem vascular discoloration, and root rot with abundant mycelial growth. Stem and root segments (1-2 cm) from three to five diseased plants were agitated in 0.1% tween-20 and soaked in 70% ethanol for 30 s and 1% NaOCl for 2 min. After incubating for 5 to 7 days on 1:10 potato dextrose agar (PDA) amended with tetracycline, Fusarium selective medium (FSM), and PARP (pimaricin + ampicillin + rifampicin + pentachloronitrobenzene [PCNB] agar) medium, white to pale cream aerial mycelium emerged from tissue of all plants on PDA and FSM but not PARP. Isolates cultured on 0.1% potassium chloride agar formed heads of microconidia on long monophialides consistent with the Fusarium solani species complex (FSSC) (Leslie and Summerell 2008). To obtain pure cultures of two isolates (CS529 and CS530), a single-hyphal tip was excised and grown on PDA. DNA was extracted from actively growing mycelium (PrepMan Ultra kit). The translation elongation factor gene (EF-1α) was amplified via PCR using EF1/EF2 primers (O'Donnell et al. 1998). Sequences of the two isolates were identical and deposited under accession number MW892973 in GenBank. The 599 bp sequence was 99.33% identical to FSSC 3 + 4 (Fusarium falciforme) accessions FD_01443_EF-1a based on FUSARIUM-ID BLAST analysis. To evaluate pathogenicity, stems of hemp plants (cv. 'Berry Blossom'; n=8 plants per isolate) were wounded by penetrating the epidermis in an area about 0.5-cm square by 1-mm deep and 8-inches above the soil line. A 0.5 cm-diameter plug of 7-day old F. falciforme-colonized PDA was placed against the wound. Inoculation sites were loosely wrapped with parafilm for 2 days. A negative control consisted of a sterile PDA plug (n=3). Treatments were arranged in a completely randomized design in a greenhouse. The experiment was conducted once, due to regulatory restrictions at campus facilities. At 61 days post-inoculation, external stem lesions were significantly larger in diameter (P < 0.05; Tukey's HSD) in plants inoculated with CS529 (8 ± 1 mm) compared to the control (2 ± 0 mm), and larger but not significant for CS530 (6 ± 1 mm). Internal stem lesions (i.e., rot in stele) were observed in plants inoculated with CS529 (9 ± 3 mm); stem rot was very minor in plants treated with CS530 (1 ± 1 mm) and nonexistent for control plants. No other disease symptoms were observed. F. falciforme was isolated from stems of CS529- and C530-inoculated plants. Sequences of re-isolates matched 100% with accession MW892973. These results suggest that F. falciforme causes rot in hemp in California. These studies specifically confirm stem rot abilities; field observations of root rot indicate root rotting abilities, but further tests are needed for confirmation. This is the first report of F. falciforme causing disease in industrial hemp. FSSC was described as causing foot rot in hemp in Italy (Sorrentino et al. 2019), but these isolates belonged to phylogenetic species 5 (F. solani) not F. falciforme. In addition, F. falciforme was reported as causing root rot in hydroponically grown cannabis (Punja and Rodriguez 2018). These studies provide the foundation for development of management tools for hemp disease.

Keywords: Causal Agent; Crop Type; Etiology; Field crops; Fungi; Subject Areas; other.

Collards (Brassica oleracea var. acephala) are grown throughout the United States. Hydroponic greens are more common now due to technological advances lowering the cost and increasing the efficacy of production. In January 2021, a 325 m2 indoor hydroponic farm opened to provide fresh produce for a school in Los Angeles County, CA. Three week old collard seedlings were purchased from a local nursery, rinsed of their rooting media, and transplanted into deep water culture beds (1.2 m x 2.5 m x 0.3 m). Two weeks later, symptoms including plant stunting, chlorosis, leaf curling and wilting, and brown necrotic roots appeared. By and by 80-100% of usable plants were lost to disease. Symptomatic roots were plated on corn meal agar (CMA) amended with 2 ml of 25% lactic acid and CMA amended with pimaricin, ampicillin, rifampicin, and pentachloronitrobenzene (PARP) (Kannwischer et al. 1978). After 2 days a single colony type emerged on PARP but no growth occurred on acidified CMA. Representative isolates were transferred to CMA and to filtered (0.02 µm) soil extract solution with boiled grass blades (Martin 1992), both of which were incubated at 22 C and ambient light conditions. On CMA, isolates produced coenocytic mycelium with minimal aerial hyphae. After 24 h in soil extract, isolates developed filamentous sporangia, elongated discharge tubes with slightly inflated tips, and zoospores. Oospores were not observed. Pathogenicity was confirmed by soaking the roots of five day old collard seedlings in beakers containing zoospores (1 x 102 zoospores/ml) in filtered soil extract. Four isolates were tested on 15 seedlings each. After 24 h at 22 C in ambient light conditions, plants were transferred to new beakers with roots placed on filter paper at the bottom and saturated with sterile distilled water. Three days after this transfer, leaves on all plants turned chlorotic and roots developed brown lesions from which morphologically identical colonies were isolated. Control plants, soaked in filtered soil extract, developed no root or foliar symptoms. To molecularly identify the collard isolates, DNA was extracted from mycelial original and re-isolated isolates and was amplified by PCR using mitochondrial primers for the cytochrome oxidase I (COI) gene (Robideau et al. 2011) and the cytochrome oxidase II (COX2) gene (Martin 2000). The only species that matched both loci from the original and re-isolated isolates with a high percent identity was Pythium dissotocum. The COI locus from the original isolate (MZ027311) matched P. dissotocum with 99% identity and with 332/334 base pairs matching the isolate with Sequence ID MT981134.1. From the re-isolated isolate (MZ027313), the COIequence perfectly matched 657/657 base pairs of P. dissotocum (Sequence ID MT981147.1). The COX2 locus from the original isolate (MZ027312) matched P. dissotocum (Sequence ID MG719859.1) with a 99% identity and 517/518 matching base pairs and the re-isolated isolate (MZ027314) perfectly matched P. dissotocum (Sequence ID MG719859.1) with 515/515 matching base pairs. Based on these molecular and morphological data, the isolates were identified as Pythium dissotocum. To our knowledge, this is the first report of P. dissotocum causing root rot on collards. At this same facility, P. dissotocum was also confirmed as the cause of declining bean (Phaseolus vulgaris) plants. As hydroponics will be necessary to feed a growing population - especially in urban areas -- and because leafy greens are a main crop of the hydroponics industry, we anticipate this issue may become common. Hydroponic systems are highly conducive to the persistence of Oomycetes and a record of infection and plan of action will be necessary to preserve crop health.

Keywords: deep water culture; hydroponics; plant disease; root rot; waterborne pathogens.

In Aug 2019, approximately 10% of mung bean plants at the experimental farm of the Jiangsu Academy of Agricultural Science (32.03 N; 118.88 E) showed symptoms of stunting and wilting. Brown and water-soaked stem lesions were often observed at the base of the diseased plants. In severe cases, the plants collapsed and cumulous aerial mycelia were visible on the basal stem surface (Figure S1 A). To identify the causal agent, a total of 20 tissue fragments (5 mm long) were excised from roots and basal stems of five symptomatic plants. The fragments were surface sterilized in 2% sodium hypochlorite solution then plated on 2.5% potato dextrose agar (PDA) plates containing 10 μg/mL pimaricin, 100 μg/mL ampicillin, 10 μg/mL rifampicin, and 10 μg/mL pentachloronitrobenzene (PARP; Beckerman et al. 2017). After 3-4 days incubation at 25^oC in dark, 14 colonies with white and cumulous mycelia grew from the tissue pieces (named as JS19-1 to JS19-14). JS19-1 and JS19-2 were purified by hyphal tipping, then grown on PDA medium for 7 days for morphological observation using a compound microscope (Figure S1 B, C). Width of coenocytic hyphae ranged from 3.7 to 8.9 (avg. 6.1, n=20) μm. Terminal oogonia were globose and with a diameter of 13.8 to 25.8 (avg. 22, n=20) μm. Antheridia were barrel-shaped, and mostly intercalary, sometimes terminal. Most of antheridia were diclinous, with 6.2 to 12.5 (avg. 9.3, n=20) μm in width and 7.6 to 15.3 (avg. 12.8, n=20) μm in length. Oogonia were fertilized with one or two (rare) antheridia. Oospores were aplerotic, 10.1 to 23.5 (avg. 20.4, n=20) μm in diameter. Sporangia had terminal inflated hyphal branches (Figure S1 D, E). The two isolates were preliminary identified as Pythium aphanidermatum. For molecular identification, the sequences of internal transcribed spacer (ITS) rDNA, cytochrome oxidase subunit I (CoxI) (Robideau et al. 2011), and β-tubulin (Kroon et al. 2004) of JS19-1 were detected, and deposited in GenBank (MT949538, MT949539 and MT949540). The ITS and CoxI sequences were identical with P. aphanidermatum CBS28779 ITS (759/759 bp, HQ643439.1) and PYT01 CoxI (640/640 bp, MH760243.1) respectively, the β-tubulin sequence showed 99% (830/840 bp) similarity of P. aphanidermatum P2 (AY564048.1). Thus, JS19-1 was confirmed as P. aphanidermatum. To fulfill Koch's postulates, the pathogenicity of JS19-1 was tested using the procedure of Kiyoshi et al. (2021) with some modifications. Barley grains infested with JS19-1 were as inoculum and thoroughly mixed with potting mixture at a rate of 10% in volume. Six mung bean seeds were sown per pot and then grown in a greenhouse. Potting mixture with no inoculum was used as control. Three pots of replicate plants used for inoculation and control. After 3 weeks, emergence in the inoculated pots was 33% and symptoms of stunting and root rot similar to those in field were observed, while control plants were asymptomatic (FigureS1 F, G). P. aphanidermatum was successfully reisolated from symptomatic plants of both methods. The pathogenicity tests were repeated twice. P. aphanidermatum causes seed rot, pre- and postemergence damping-off, or stem/root rot of a wide range of industrial crops and vegetables (Liu et al, 2018). To our knowledge, this is the first report of P. aphanidermatum causing disease on mung bean in China. Since Phytophthora vignae (Sun et al, 2020) and P. myriotylum (Yan et al, 2021) have been reported causing mung bean root rot, integrated disease management should be adopted to reduce damage.

Keywords:Pythium aphanidermatum; mung bean; root rot.

AmphL is a cytochrome P450 enzyme that catalyzes the C8 oxidation of 8-deoxyamphotericin B to the polyene macrolide antibiotic, amphotericin B. To understand this substrate selectivity, we solved the crystal structure of AmphL to a resolution of 2.0Å in complex with amphotericin B and performed molecular dynamics simulations. A detailed comparison with the closely related P450, PimD, which catalyzes the epoxidation of 4,5-desepoxypimaricin to the macrolide antibiotic, pimaricin, reveals key catalytic structural features responsible for stereo- and regio-selective oxidation. Both P450s have a similar access channel that runs parallel to the active site I helix over the surface of the heme. Molecular dynamics simulations of substrate binding reveal PimD can "pull" substrates further into the P450 access channel owing to additional electrostatic interactions between the protein and the carboxyl group attached to the hemiketal ring of 4,5-desepoxypimaricin. This substrate interaction is absent in AmphL although the additional substrate -OH groups in 8-deoxyamphotericin B help to correctly position the substrate for C8 oxidation. Simulations of the oxy-complex indicates that these -OH groups may also participate in a proton relay network required for O₂ activation as has been suggested for two other macrolide P450s, PimD and P450eryF. These findings provide experimentally testable models that can potentially contribute to a new generation of novel macrolide antibiotics with enhanced antifungal and/or antiprotozoal efficacy.

Keywords: Cytochrome P450; antibiotics; crystal structure; molecular dynamics; substrate specificity.