Paclitaxel (6 mg/mL) was purchased from Hikma-specialized pharmaceuticals, Egypt. Aluminum thin layer chromatography (TLC) plates were purchased from Sigma Aldrich, USA. Other chemicals were of analytical grade.

Isolation of associated fungi with Artemisia judaica’s roots

Samples of healthy Artemisia judaica roots were collected from Wadi Abu Shih area region at Red Sea governorate (latitudes 26° 30′ and 26° 44′ N and longitudes 33° 20′ and 33° 30′ E). Three types of fungi were isolated: rhizosphere (i.e., fungi that closely surround plant roots and support their growth), rhizoplane (i.e., fungi interact with plant roots’ surface), and endophytic fungi (i.e., fungi that colonize within the plant tissue and exchanges mutualistic effect on each other). Two culturing media were used: Czapek’s agar (CZA) (glucose 20 g, NaNO3 2.0 g, KH2PO4 0.5 g, KCl 0.5 g, MgSO4·7H2O 2.0 g L−1; pH 8.5 ± 0.02) and its modified form Cellulose- Czapek’s agar (CCZA) which contains cellulose instead of glucose as the sole carbon source. First, rhizosphere fungi (RSF) were isolated by adding 5 g root sample into 100 mL sterile purified water and mixed thoroughly. The produced suspension (10 mL) was transferred into 40 mL. Finally, the plates were inoculated with 1 mL suspension. Second, Artemisia root was washed with sterile purified water and cut into small pieces for rhizoplane fungi (RPF) isolation, and 4 segments were placed per plate. Third, for endophytic fungi (EPF) isolation, Artemisia roots were first cut into small fragments (~ 5 × 5 mm) using sterile blade. Then, sterilized by successive soaking in 70% ethanol for 1 min and then 2.5% NaClO for 3 min, rinsed thrice with sterile purified water and dried using sterile paper tissues. All plates were incubated for 5 days at 28 ± 2.0 °C.

Screening and extraction of paclitaxel produced by endophytic fungi

Isolated fungi were tested for their ability to produce paclitaxel (PTX) naturally by cultivation in 50 mL freshly prepared potato dextrose broth (PDB) medium (potato extract infusion of 200 g potatoes/L; prepared by boiling for 30 min and then filtrated through cheese cloth; 20 g/L dextrose was added at pH 6.5). Cultures were incubated at 28 ± 2.0 °C for 18 days statically. After incubation, mycelia were separated by filtration. For TLC detection, the mycelia were thoroughly crushed in a mortar, after that the fermented broth and ground mycelia were extracted by 50 mL chloroform. For HPLC and GC/MS analysis, 1 g of each culture was dried at 60 °C for 5 h and then immersed in chloroform for 24 h. Next day, chloroform was separated by filtration through anhydrous sodium sulfate. Finally, chloroform was evaporated to obtain the dried crude extracts. The crude extracts were redissolved in less amount of chloroform for the subsequent separation. Paclitaxel production was identified and detected from fungal crude extracts by spotting 50 μL on TLC plates comparing to authentic sample, silica gel coated with fluorescent indicator F254. The TLC plates were developed using chloroform: methanol (7:1, v/v) solvent system (Strobel et al. 1996). After the plates were air-dried, they were sprayed with a vanillin-H2SO4 reagent (2 g of vanillin dissolved in a mixture of 100 mL methanol and 1 mL H2SO4) and then incubated at 110 °C for 15 min. Also, the isolates showing intense bands of PTX on TLC were further confirmed using high-performance liquid chromatography (HPLC) analysis. The HPLC analysis (HPLC Chromass Young, Korea) was performed by Series equipped with a quaternary pump, a kinetex evo-C18 column 100 mm HPLC × 4.6 mm (Phenomenex®, USA), and operated at 35 °C. The injected volume was 20 μL. The UV detector was set at a wavelength of 205 nm, and the humidity was 38% rH.

Molecular identification of Aspergillus fumigatiaffinis

A culture of the interested fungal isolate (AA17) was sent to Macrogen (Seoul, South Korea) for molecular identification and sequencing analysis using the universal primers: ITS1 (5′-TCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′). The obtained data files contained a consensus sequence of ITS1 and ITS4 primers, which was used to run a BLAST search on NCBI website for matching identification with deposited related species.

Impact of various media types on paclitaxel production by Aspergillus fumigatiaffinis PP235788

A 7-day culture was used (1 mL spore suspension) to inoculate 49 mL medium of potato dextrose broth (PDB), Czapekʼs-Dox broth (CZB), M1D, and malt yeast extract broth (MYE). Each type of medium was inoculated in triplicates. Negative controls of each medium without inoculation were also considered. All conical flasks were incubated at 28 ± 2.0 °C for 2 weeks statically. Following incubation, microbial cultures were separated by filtration, and the previously described extraction and determination of PTX were carried out.

Quantification of paclitaxel produced by Aspergillus fumigatiaffinis PP235788

Aspergillus fumigatiaffinis PP235788 extract was analyzed by GC/MS (unknown concentration of PTX) where the concentration was measured using a calibration curve of standard PTX. The calibration curve was performed by injecting five concentrations of standard PTX (30, 50, 80, 100, and 150 ppb ≈ μg L−1). An average of five independent trials was used to assess the concentration of PTX produced from extracted fungal filtrate. A TSQ triple quadrupole GC/MS instrument coupled with a Thermo Scientific™ TRACE™ 1300 GC (Thermo Scientific, Austin, TX, USA) was used. Sample’s introduction was performed by Thermo Scientific™ AS3000 autosampler and chromatographic separation using a Thermo Scientific™ TraceGOLD TG-5MS 30 m × 0.25 mm I.D. × 0.25-μm-film capillary column.

Optimizing paclitaxel production by Aspergillus fumigatiaffinis PP235788

Based on the impact of various types of media on PTX production, potato dextrose broth medium was selected for performing optimization conditions in addition to adding a stimulator for PTX production, which is vitamin B complex (becozyme) at different concentrations per liter. Becozyme is a vitamin B complex which includes thiamine hydrochloride (B1, 10 mg), riboflavin sodium phosphate (B2, 4.57 mg), nicotinamide (B3, 40 mg), dexpanizole (B5, 6 mg), and pyridoxine hydrochloride (B6, 6 mg). Also, pH values and days of incubation were considered.

The design of experiment (DOE) statistical analysis was employed to assess the individual and interactive effects of 3 variables on PTX production using SigmaXL Version 10 (SigmaXL Inc., Ontario, Canada) (Jankovic et al. 2021). An experimental matrix of 2-level factorial design (8-run, full factorial) was designed upon 3 variables (becozyme, pH value, and incubation days). Each factor is represented by 2-levels of “low” and “high” to measure one response (PTX production μg L−1). The settings were “50, 150 μg L−1,” “5.0, 8.0,” and “14, 18” for becozyme, pH, and fermentation days, respectively (Table 1). Two biological replicates were used to maximize standard deviation value. The order of experimental runs was randomized to minimize the impact of any uncontrolled variables. All runs were performed in 250-mL-Erlenmeyer flasks with 50 mL PDB at 28 ± 2.0 °C with freshly prepared fungal spore suspension (1 mL). After growth, PTX was extracted and quantified by TLC and HPLC, as described earlier.

Table 1 Design of 3 factor, 8-run, 2**3, full-factorial for one response (paclitaxel production μg L−1). One block and one center point

After determining the optimum conditions for PTX production among the 3 desired variables, it was found that pH and becozyme are the most important factors for production. Thus, these two factors were selected for the response surface design (RSD) analysis. Response surface approach is a very useful tool for determining the best circumstances to boost microbial production and for building correlational mathematical models which predict the response variable based on different parameter combinations. Central composite design (CCD, 2 ctr pts) was employed using 2 replicates, and alpha axial value was set to face centered (alpha = 1.0) for 10-run. The used design is depicted in Table 2; each factor is represented by 3 center points (low, medium, and high).

Table 2 Response surface methodology of PTX production. Paclitaxel production, predicted values, and residuals using 2 variables through 20-run

The two generated models, full-factorial design, and response surface methodology were statistically analyzed using two-way ANOVA, for detecting the significance of two variables, pH value, and becozyme concentration on produced concentration of PTX.

Statistical analyses

All experiments were conducted with two or three biological replicates, as indicated, and the results were expressed by mean ± standard deviation. Statistical analyses and analysis of variance (ANOVA) for each model is generated using SigmaXL Version 10.

Deposition of fungal material

The isolate A. fumigatiaffinis AA17 was deposited at the Assiut University Mycological Centre (AUMC), Egypt, with deposit number AUMC16334 and accession number PP235788.1 at GenBank.