Application of Priming Strategy for Enhanced Paclitaxel Biosynthesis in Taxus × Media Hairy Root Cultures

Despite huge progress in biotechnological approaches to paclitaxel production, Taxus spp. in vitro culture productivity still remains a challenge. This could be solved by developing a new strategy engaging mechanisms of the primed defence response joined with subsequent elicitation treatment to circumvent limitations in paclitaxel biosynthesis. The hairy roots were primed by preincubation with β-aminobutyric acid (BABA) for 24 h or 1 week, and then elicited with methyl jasmonate (MeJA) or a mixture of MeJA, sodium nitroprusside and L-phenylalanine (MIX). The effect of priming was evaluated on a molecular level by examination of the expression profiles of the four genes involved in paclitaxel biosynthesis, i.e., TXS (taxadiene synthase), BAPT (baccatin III: 3-amino, 3-phenylpropanoyltransferase), DBTNBT (3′-N-debenzoyl-2-deoxytaxol-N-benzoyltransferase) and PAM (phenylalanine aminomutase), as well as rolC (cytokinin-β-glucosidase), originated from the T-DNA of Agrobacterium rhizogenes. The maximum paclitaxel yield was achieved in cultures primed with BABA for 1 week and elicited with MIX (3179.9 ± 212 µg/g dry weight), which corresponded to the highest expression levels of TXS and BAPT genes. Although BABA itself induced the investigated gene expression over control level, it was not translated into paclitaxel production. Nevertheless, preincubation with BABA essentially affected paclitaxel yield, and the duration of BABA pretreatment seemed to have the most pronounced impact on its productivity.


Introduction
Since its discovery in the 1960s, paclitaxel has become a well-established and extensively used anticancer drug in the treatment of a wide variety of malignancies, including breast and ovarian cancers, non-small-cell lung cancer, head and neck tumours, Kaposi's sarcoma and urologic cancers [1,2]. Paclitaxel is also considered as a chemotherapeutic in Alzheimer's disease [3], as well as in restenosis treatment [4]. The development of numerous new practical applications of paclitaxel continues to place this molecule in the centre of scientific interest, covering not only its chemistry and novel formulations, but also the aspects of paclitaxel supply. Although paclitaxel is now produced semisynthetically from its analogues, biotechnological methods for its production on a commercial scale have also been elaborated, and the FDA (Food and Drug Administration, USA) has approved plant cell cultures as a renewable source of paclitaxel [5]. Nevertheless, despite huge progress in biotechnological approaches to paclitaxel production, Taxus spp. cell culture productivity still remains a challenge. This could be solved by the identification of the limiting steps at subcultured every four weeks. The cultures were performed on a gyratory shaker at 105 rpm (TR 250 INFROS AG, Bottingen, Switzerland) in the dark at 23 ± 1 • C.
Twenty-eight-day-old roots were subjected to experiments. The following culture variants were carried out: (i) control culture-untreated 28-day-old roots (day "0"); (ii) treated with ETOH 35 µL per flask; (iii) elicited with 100 µM methyl jasmonate (MeJA); (iv) elicited with 100 µM MeJA, 100 µM L-phenylalanine and 10 µM sodium nitroprusside (all together denoted as MIX); (v) primed with 100 µM β-aminobutyric acid (BABA) for 1 h, 3 h, 6  The hairy root growth in control culture was determined based on fresh weight (FW) increase using the following equation [14]: where m 28d is the FW of hairy roots after 28 days of culture (FW 28d ), m 0d is FW at day of inoculation. Meanwhile, hairy root growth in treated cultures was determined based on the equation: where m (14d) is the FW of hairy roots after 14 days of treatment (FW 14d ), m 0d is FW at day of inoculation (day "0").

Quantitative Real-Time PCR (qRT-PCR)
The expression profiles of the four genes involved in paclitaxel biosynthesis, i.e., TXS, BAPT, DBTNBT and PAM, as well as rolC (originated from T-DNA of A. rhizogenes), were investigated. TXS commits first step of paclitaxel and other taxane biosynthesis, i.e., converts geranylgeranyl pyrophosphate to taxa-4(5), 11(12)-diene; PAM plays role in formation of β-phenylalanoyl-CoA side chain from L-phenylalanine, while BAPT and DBTNBT are engaged in late-pathway taxane biosynthesis steps. BAPT catalyses conjugation of the β-phenylalanoyl-CoA side chain to the C13 hydroxyl group of baccatin III to generate 3 -N-debenzoyl-2-deoxytaxol. DBTNBT participates in formation of 2 -deoxytaxol through ligation of benzoyl CoA group to 3 -N-debenzoyl-2-deoxytaxol [15,16]. For gene expression analysis, primed BABA24 or BABA1 and elicited culture samples were collected after 1 h, 3 h, 6 h, 12 h, 18 h, 24 h, 48 h, 72 h. The total RNA was isolated from hairy roots frozen in liquid nitrogen according to the method described by Sykłowska-Baranek et al. [17]. One µ of total RNA was used to reverse transcription carried out with cDNA RevertAid First Strand cDNA Synthesis Kit and random hexamer primer (Thermofisher Scientific, Waltham, MA, USA) according to the manufacturer's instruction.
Quantitative real-time PCR (qRT-PCR) was used to analyse baccatin III: 3-amino, 3phenylpropanoyltransferase (BAPT), 3 -N-debenzoyl-2-deoxytaxol-N-benzoyltransferase (DBTNBT), phenylalanine aminomutase (PAM), cytokinin-beta-glucosidase (rolC), taxadiene synthase (TXS) gene expression changes in investigated T. × media hairy roots during elicitation. The primer sequences were designed using Primer3, Version 0.4.0. Verification of primers' matching and specificity was performed by in silico validation, using Basic Local Alignment Search Tool (BLAST). The sequences of the primers are presented in Supplementary Table S1. The qPCR reactions for BAPT, DBTNBT, PAM, TXS genes were performed using 2X PowerUp™ SYBR™ Green Master Mix (Applied Biosystems by Life Technologies, Austin, TX, USA), whereas for rolC gene using 2X SYBR™ Green PCR Master Mix (Applied Biosystems by Life Technologies, Austin, TX, USA) in 10 µL final reaction volume according to the manufacturer's protocols. The cDNA templates were 20-fold diluted and the final concentration for each primer was 500 nM. Each qPCR reaction was performed in duplicate on 96-well FrameStar plates (4titude ® Ltd., Wotton, UK) using the LightCycler ® 480 II instrument (Roche Diagnostics GmbH, Mannheim, Germany). The temperature cycling conditions were for 2X PowerUp™ SYBR™ Green Master Mix as follow: UDG activation at 50 • C for 2 min, Dual-Lock™ DNA polymerase at 95 • C for 2 min, followed by 40 cycles of amplification at 95 • C for 15 s, 60 • C for 15 s and 72 • C for 1 min; whereas for 2X SYBR™ Green PCR Master Mix: AmpliTaq Gold ® polymerase activation at 95 • C for 10 min followed by 40 cycles of amplification at 95 • C for 15 s and 60 • C for 1 min. Specificity of the primer amplicons was confirmed by melting curve analysis and 2% agarose gel electrophoresis. No-template controls (NTCs) were included in every qPCR run. Raw qPCR data were analysed using LightCycler 480 Software, Version 1.5.0 SP3. Number of quantification cycles were determined using the second derivative maximum method. The qPCR data of the candidate reference genes 18S rRNA, 3,5-epimerase-4-reductase-like protein (TBC41), β-tubulin (TUBB) were analysed using the tool RefFinder (https://heartcure.com.au; accessed on 14 October 2020), available online. Based on the rankings from geNorm (Supplementary Figure S1), final gene expression data were normalised to the most stable reference genes TUBB and TBC41, the latter was also chosen as a reference gene in study by Sabater-Jara et al. [18]. Fold change of the gene expression, PCR efficiency-corrected, was calculated using REST 2009 Relative Expression Software Tool, Version 2.0.13 (Qiagen GmbH, Hilden, Germany).

Paclitaxel Quantitative Determination Using HPLC-UV-DAD Method
Paclitaxel concentration was determined in hairy roots cultivated in treated cultures and compared to its content in control. At set time points, roots were collected and gently pressed on the filter paper and lyophilised (Christ ALPHA 1-4 LSC; Osterode am Harz, Germany). Afterwards, roots were powdered (100 mg) and extracted with 1 mL MEOH by 15 min sonication (Sonorex; Bandelin, Berlin, Germany), and left overnight on gyratory shaker in the dark. Next, samples were centrifugated at 15,500× g (EBA 12R Hettich, Tuttlingen, Germany), extracts were transferred to fresh Eppendorf tubes and samples were re-extracted with 1 mL MEOH. Finally, extracts were collected, evaporated and cleaned using SPE method developed by Theodoridis [19]. Paclitaxel determination was performed according to Sykłowska-Baranek et. al. [20], applying the method described by Theodoridis [21].
All experiments were carried out in triplicate and the statistical significance between means was assessed by the Kruskal-Wallis one-way analysis of variance using STATISTICA 13.1 PL software (StatSoft Polska; Kraków, Poland). A probability of p < 0.05 was considered as significant. Pair-wise correlations were calculated by Pearson's correlation coefficient test.

Hairy Root Cultures and Paclitaxel Biosynthesis
The primed defence response approach, combined with subsequent elicitor treatment, was applied in the current study to enhance paclitaxel biosynthesis in hairy root cultures of T. × media. The examination of culture conditions on the biomass increase of the KT hairy root line showed that BABA applied alone significantly and almost 2-fold stimulated root growth in comparison to control ( Figure 1). Additionally, in the BABA1E variant, root growth was over two times higher than in the control. Among elicited variants, this supplemented with MIX resulted in a slight growth decrease, although not statistically significant (p < 0.05). the KT hairy root line showed that BABA applied alone significantly and almost 2-fold stimulated root growth in comparison to control ( Figure 1). Additionally, in the BABA1E variant, root growth was over two times higher than in the control. Among elicited variants, this supplemented with MIX resulted in a slight growth decrease, although not statistically significant (p < 0.05). In culture variants without BABA, the highest paclitaxel content was determined after a 14-day-long elicitation with MeJA (303.1 ± 42.5 µg/g dry weight, DW), which was over 58-fold more than in the control (5.18 ± 1.7 µg/g DW) ( Figure 2a). While under treatment with MIX, paclitaxel was detected only up to 72 h of culture, with its highest content noted after 24 h of elicitation. Nevertheless, it was a 2.4-fold lower concentration than after MeJA elicitation at the same time point (Figure 2a). The BABA applied alone, irrespective of duration of treatment, had no effect on paclitaxel biosynthesis. Medium supplementation with only ETOH did not induce paclitaxel production, in contrast to its application after 24 h of BABA treatment, where paclitaxel production was observed. Under those conditions, its highest yield was noted after 24 h, and decreased subsequently ( Figure 2b). Among other BABA24 variants, BABA24MIX was the most favourable for paclitaxel accumulation, although its presence was detected for the first time on the 7th day of treatment, while its highest content amounted to 21.5 ± 2.5 µg/g DW, noted on day 14. In the BABA24MeJA variant, paclitaxel biosynthesis started after 72 h of culture and its highest yield (17.1 ± 4.0 µg/g DW) was achieved on day 7, which was then of BABA on a significant decline in its concentration. In the BABA1 variants, paclitaxel biosynthesis was determined only after elicitor/s addition. In the BABA1MeJA cultures, paclitaxel content was detected for the first time after 7 days of elicitation. At the same time point, it reached its maximum concentration of 495.7 ± 86.3 µg/g DW. In the BA-BA1MIX variant, paclitaxel accumulation started after 72 h of culture and rose up gradually up to 3179.9 ± 211.9 µg/g DW on day 14. It was the highest paclitaxel content de- In culture variants without BABA, the highest paclitaxel content was determined after a 14-day-long elicitation with MeJA (303.1 ± 42.5 µg/g dry weight, DW), which was over 58-fold more than in the control (5.18 ± 1.7 µg/g DW) ( Figure 2a). While under treatment with MIX, paclitaxel was detected only up to 72 h of culture, with its highest content noted after 24 h of elicitation. Nevertheless, it was a 2.4-fold lower concentration than after MeJA elicitation at the same time point (Figure 2a). The BABA applied alone, irrespective of duration of treatment, had no effect on paclitaxel biosynthesis. Medium supplementation with only ETOH did not induce paclitaxel production, in contrast to its application after 24 h of BABA treatment, where paclitaxel production was observed. Under those conditions, its highest yield was noted after 24 h, and decreased subsequently (Figure 2b). Among other BABA24 variants, BABA24MIX was the most favourable for paclitaxel accumulation, although its presence was detected for the first time on the 7th day of treatment, while its highest content amounted to 21.5 ± 2.5 µg/g DW, noted on day 14. In the BABA24MeJA variant, paclitaxel biosynthesis started after 72 h of culture and its highest yield (17.1 ± 4.0 µg/g DW) was achieved on day 7, which was then followed by a significant decline in its concentration. In the BABA1 variants, paclitaxel biosynthesis was determined only after elicitor/s addition. In the BABA1MeJA cultures, paclitaxel content was detected for the first time after 7 days of elicitation. At the same time point, it reached its maximum concentration of 495.7 ± 86.3 µg/g DW. In the BABA1MIX variant, paclitaxel accumulation started after 72 h of culture and rose up gradually up to 3179.9 ± 211.9 µg/g DW on day 14. It was the highest paclitaxel content determined under the conditions of the current study, over 613-fold higher than in control and over 6-fold higher than in the BABA1MeJA variant at its peak (495.7 ± 86.3 µg/g DW) (Figure 2c). termined under the conditions of the current study, over 613-fold higher than in control and over 6-fold higher than in the BABA1MeJA variant at its peak (495.7 ± 86.3 µg/g DW) (Figure 2c).

Gene Expression Profiles
In the control culture, the expression of early-as well as late-pathway genes was at a parallel level, with the lowest expression showed by rolC (Figure 3a). BABA affected the levels of BAPT and rolC genes the most (Figure 3b), however, they did not follow the same expression pattern, and, in addition, the correlation between them was negative (r = -0.01). The highest BAPT transcript level was noted 72 h after BABA addition, while for rolC it was after 24 h. For further experiments with BABA, two time points were chosen for elicitor supplementation, i.e., 24 h and 1 week. Relative expression levels of genes of interest between these time points did not differ significantly (p < 0.05), and they were also not significantly different than at other examined time points.

Gene Expression Profiles
In the control culture, the expression of early-as well as late-pathway genes was at a parallel level, with the lowest expression showed by rolC (Figure 3a). BABA affected the levels of BAPT and rolC genes the most (Figure 3b), however, they did not follow the same expression pattern, and, in addition, the correlation between them was negative (r = −0.01). The highest BAPT transcript level was noted 72 h after BABA addition, while for rolC it was after 24 h. For further experiments with BABA, two time points were chosen for elicitor supplementation, i.e., 24 h and 1 week. Relative expression levels of genes of interest between these time points did not differ significantly (p < 0.05), and they were also not significantly different than at other examined time points.
In culture variants without BABA, ETOH added individually caused an increased in steady-state transcript levels. The most affected were TXS, DBTNBT and rolC after 24 h, and PAM, which peaked twice after 3 and 24 h (Figure 4a). However, these inductions of gene expression were not translated into paclitaxel production (Figure 2a). MeJA elicitation caused a rise in the gene expression levels of almost all genes investigated, except TXS. The coincidence in the highest transcript abundance was observed in the expression levels of PAM and rolC. BAPT peaked after 12 h of treatment, although the maximum expression was noted after 72 h. Meanwhile, the maximum level of DBTNBT expression was detected at 24 h post-elicitation (Figure 4b). These enhancements in transcript abundance reflected the capacity in paclitaxel accumulation, which was the highest among these culture variants (Figure 2a). In the hairy roots elicited with MIX, the significant induction of late-pathway genes was determined with the highest PAM transcript abundance 12 h after the addition of elicitors (Figure 4c). The expression of rolC was also remarkably affected; however, its levels were over 2-fold lower than under MeJA added as a single elicitor.  In culture variants without BABA, ETOH added individually caused an increased in steady-state transcript levels. The most affected were TXS, DBTNBT and rolC after 24 h, and PAM, which peaked twice after 3 and 24 h (Figure 4a). However, these inductions of gene expression were not translated into paclitaxel production (Figure 2a). MeJA elicitation caused a rise in the gene expression levels of almost all genes investigated, except TXS. The coincidence in the highest transcript abundance was observed in the expression levels of PAM and rolC. BAPT peaked after 12 h of treatment, although the maximum expression was noted after 72 h. Meanwhile, the maximum level of DBTNBT expression was detected at 24 h post-elicitation (Figure 4b). These enhancements in transcript abundance reflected the capacity in paclitaxel accumulation, which was the highest among these culture variants (Figure 2a). In the hairy roots elicited with MIX, the signif- After BABA24 treatment followed by ETOH, the most induced gene was BAPT, which peaked two times, i.e., after 1 and 48 h (Figure 5a). The paclitaxel production under this treatment was the highest after 24 h, followed by its decline. After the addition of MeJA to BABA24 treated hairy roots, the transcript abundance was higher than in the BABA24E variant, except the DBTNBT gene. TXS demonstrated the highest gene expression levels, with a peak after 12 h (Figure 5b). Under these conditions, paclitaxel accumulation was observed starting after 72 h of elicitation, with its maximum after 7 days, and was over 3-fold higher than in the BABA24E variant. In the BABA24MIX and BABA24MeJA variants, a similar pattern of gene expression was observed (Figure 5c). However, in opposition to the BABA24MeJA variant, the DBTNBT gene were considerably higher expressed. Nevertheless, this was not reflected in the yield of paclitaxel, which was higher than in the BABA24MeJA variant, but this difference was not statistically significant (p < 0.05). Further, the shift in the time required to achieve this compound maximum concentration was noted, from 7 days to 14 days in BABA24MeJA and BABA24MIX variants, respectively.
Cells 2022, 11, x FOR PEER REVIEW 9 of 19 abundance 12 h after the addition of elicitors (Figure 4c). The expression of rolC was also remarkably affected; however, its levels were over 2-fold lower than under MeJA added as a single elicitor.  After BABA24 treatment followed by ETOH, the most induced gene was BAPT, which peaked two times, i.e., after 1 and 48 h (Figure 5a). The paclitaxel production under this treatment was the highest after 24 h, followed by its decline. After the addition of MeJA to BABA24 treated hairy roots, the transcript abundance was higher than in the BABA24E variant, except the DBTNBT gene. TXS demonstrated the highest gene expression levels, with a peak after 12 h (Figure 5b). Under these conditions, paclitaxel accumulation was observed starting after 72 h of elicitation, with its maximum after 7 days, and was over 3-fold higher than in the BABA24E variant. In the BABA24MIX and BA-BA24MeJA variants, a similar pattern of gene expression was observed (Figure 5c). However, in opposition to the BABA24MeJA variant, the DBTNBT gene were considerably higher expressed. Nevertheless, this was not reflected in the yield of paclitaxel, which was higher than in the BABA24MeJA variant, but this difference was not statistically significant (p < 0.05). Further, the shift in the time required to achieve this compound maximum concentration was noted, from 7 days to 14 days in BABA24MeJA and BA-BA24MIX variants, respectively.  In the BABA1E variant, within the first hour the enhancement in steady-state transcript levels of almost all examined genes except DBTNBT (Figure 6a) was noted; however, this rise was not translated into paclitaxel production. In the BABA1MeJA variant, starting from the first hour, an increased expression of late-pathway BAPT, DBTNTB and PAM genes, as well the rolC gene, was observed with continued abundance extending through to 72 h (Figure 6b). The most inducible gene was PAM, which peaked after 6 h of elicitation. TXS expression, an early-pathway gene, was very low, lower even than in the BABA1E variant. Eventually, under these conditions, paclitaxel production was detected at 7 and 14 days after elicitor feeding. In the BABA1MIX variant, the TXS, BAPT and PAM abundance of transcript was elevated in comparison to previous BABA1 variants, In the BABA1E variant, within the first hour the enhancement in steady-state transcript levels of almost all examined genes except DBTNBT (Figure 6a) was noted; however, this rise was not translated into paclitaxel production. In the BABA1MeJA variant, starting from the first hour, an increased expression of late-pathway BAPT, DBTNTB and PAM genes, as well the rolC gene, was observed with continued abundance extending through to 72 h (Figure 6b). The most inducible gene was PAM, which peaked after 6 h of elicitation. TXS expression, an early-pathway gene, was very low, lower even than in the BABA1E variant. Eventually, under these conditions, paclitaxel production was detected at 7 and 14 days after elicitor feeding. In the BABA1MIX variant, the TXS, BAPT and PAM abundance of transcript was elevated in comparison to previous BABA1 variants, especially PAM, which peaked after 12 h of treatment (Figure 6c). It should be noticed that TXS transcript levels were the highest among all BABA1 variants. DBTNTB was also induced; however, to a remarkably lesser extent than in the BABA1MeJA variant. The rolC gene expression was also lower than in the BABA1MeJA variant, but higher than in the BABA1E variant ( Figure 6). Taken all together, paclitaxel yield determined in this variant was the highest among all tested variants, with its concentration being detected starting from 48 h post-treatment and reaching its maximum after 14 days.

Discussion
The concept of harnessing the primed response induction to enhance secon metabolite production in plant cell in vitro cultures has not been investigated to However, there are some reports describing the physiological effects of priming i aspect of phytoalexin production and secretion in reaction to priming. Until now, i reported that the biosynthesis of various classes of secondary metabolites, i.e., phe compounds, glucosinolates and volatile compounds, was induced, and that their p depended on plant species and priming inducer [22].
The current study aimed to develop a new elicitation strategy involving the pr defence response approach, combined with subsequent elicitor treatment. It ass significant improvements in the yield of paclitaxel as a result of the preconditioning × media hairy root cultures with BABA for enhanced response, combined with subseq elicitor (MeJA) or elicitors (MIX) treatment.
Among the many biotechnological approaches developed to improve paclitaxe other taxane production in Taxus spp. in vitro cultures, the most effective seemed elicitation with MeJA [5]. Numerous investigations were undertaken to elucidate t fect of MeJA on taxane biosynthesis on a molecular level. The 1-7-day delay was rep between a time course for mRNA accumulation of the known taxane biosynthetic g and taxane production [15,23] and upregulation, not only of genes involved in ta biosynthesis, but also genes engaged in the stimulation of plant hormone and nylpropanoid biosynthesis, MeJA signalling, taxane transport and degradation, as w transcriptional regulation [23][24][25][26][27]. The taxane biosynthetic pathway is not fully ch terised; according to current knowledge, it requires approximately 20 enzymatic ste and possible rate-limiting steps were indicated in the late pathway, i.e., those contr by BAPT and DBTNBT.
In previous studies dealing with T. × media hairy root cultures, biomass retard upon MeJA or MIX addition was reported [12,13,17,20,28,29]. In the present study growth was not negatively affected by culture conditions, and further significan 0.05) biomass increase was noted in cultures supplemented with BABA alone. I previous study on hairy root cultures of S. miltiorrhiza, it was described that BABA a

Discussion
The concept of harnessing the primed response induction to enhance secondary metabolite production in plant cell in vitro cultures has not been investigated to date. However, there are some reports describing the physiological effects of priming in the aspect of phytoalexin production and secretion in reaction to priming. Until now, it was reported that the biosynthesis of various classes of secondary metabolites, i.e., phenolic compounds, glucosinolates and volatile compounds, was induced, and that their profile depended on plant species and priming inducer [22].
The current study aimed to develop a new elicitation strategy involving the primed defence response approach, combined with subsequent elicitor treatment. It assumes significant improvements in the yield of paclitaxel as a result of the preconditioning of T. × media hairy root cultures with BABA for enhanced response, combined with subsequent elicitor (MeJA) or elicitors (MIX) treatment.
Among the many biotechnological approaches developed to improve paclitaxel and other taxane production in Taxus spp. in vitro cultures, the most effective seemed to be elicitation with MeJA [5]. Numerous investigations were undertaken to elucidate the effect of MeJA on taxane biosynthesis on a molecular level. The 1-7-day delay was reported between a time course for mRNA accumulation of the known taxane biosynthetic genes and taxane production [15,23] and upregulation, not only of genes involved in taxane biosynthesis, but also genes engaged in the stimulation of plant hormone and phenylpropanoid biosynthesis, MeJA signalling, taxane transport and degradation, as well as transcriptional regulation [23][24][25][26][27]. The taxane biosynthetic pathway is not fully characterised; according to current knowledge, it requires approximately 20 enzymatic steps [5] and possible rate-limiting steps were indicated in the late pathway, i.e., those controlled by BAPT and DBTNBT.
In previous studies dealing with T. × media hairy root cultures, biomass retardation upon MeJA or MIX addition was reported [12,13,17,20,28,29]. In the present study, root growth was not negatively affected by culture conditions, and further significant (p < 0.05) biomass increase was noted in cultures supplemented with BABA alone. In the previous study on hairy root cultures of S. miltiorrhiza, it was described that BABA added at higher dosages (1 mM) resulted in slight root growth suppression, although this effect was not observed when BABA was used in combination with MeJA, which is in accordance with the results of the present study [11]. In the other study, it was demonstrated that BABA stimulated the photosynthesis and growth of Medicago interexta sprouts, to a higher extent when applied together with selenium nanoparticles (SeNPs) [30].
Until now, the priming strategy combined with elicitation has not been applied nor analysed in relation to enhancing paclitaxel biosynthesis. The results of the present study have been compared with previous investigations reporting changes in gene expression involved in taxane biosynthesis (Table 1). T. baccata BABA application caused the induction of investigated gene expression, even over their levels in control; however, it was not connected with the induction of paclitaxel biosynthesis. This could be attributed to the possible effect of BABA on changes in DNA methylation pattern which was postulated to affect taxane production in vitro [34][35][36].
Paclitaxel production was observed when the medium was supplemented with elicitor/s (Figure 2a). If the influence of ETOH on paclitaxel accumulation was taken into consideration, it was revealed that its content was detected only in the BABA24E variant, and in low quantities (Figure 2b), so it could be assumed that the effect of ETOH on paclitaxel biosynthesis could be irrelevant. In variants supplemented with BABA, it was demonstrated that the duration of preconditioning had a significant impact on paclitaxel productivity. The highest paclitaxel concentration was detected after a two-week elicitation in BABA1MIX variants, which corresponded to the highest expression levels of TXS and BAPT genes. DBTNBT, PAM and rolC transcript abundance was higher in the BABA1MeJA variant, in which BAPT expression was on a parallel level with the BABA1MIX variant, while TXS expression was very low (Figures 5 and 6). The activity of the TXS gene, committing the first step of taxane biosynthesis, was claimed not to be rate-limiting on taxane biosynthesis [15,16], which also found confirmation under the conditions of the present study. The results of the current investigations seem to confirm that among all investigated genes herein, BAPT could play an essential role in taxane productivity (which was previously postulated [15,16,34]), but also, PAM upregulation could affect paclitaxel biosynthesis capacity.
The maximum paclitaxel yield in the present study was determined in the BABA1MIX variant; it was also the highest paclitaxel productivity reported for this hairy root line [37]. Previously, the importance of preincubation with BABA on the tanshinone total yield was described in S. miltiorrhiza hairy root cultures, and the authors pointed out that tanshinone productivity increased steadily with the prolongation of BABA treatment [38]. Further, it was demonstrated that BABA influenced results, especially in combination with SeNPs' potentiated accumulation of phenolics and flavonoids, and especially when they were applied together [30]. Next, the stronger activation of defence response was observed in BABA-primed mango fruit to subsequent inoculation with Colletotrichum gloeosporoides [39]. Upon treatment with only BABA, in comparison to control, the total content of phenolics, flavonoids and lignin was significantly higher, which is in opposition to the results of the present study. On the other hand, the BABA pretreatment followed by C. gloeosporoides inoculation caused an acceleration in their increase and accumulation, which corroborates the present findings.
In plant secondary metabolism, the stimulating function of ROL genes was reported, as well as the effect of single ROL gene relays on plant species and groups of secondary metabolites [40,41]; however, until now it was not examined in Taxus spp. hairy roots. In the present study, the rolC gene was highly expressed under MeJA elicitation and in the BABA1MeJA variant, which corresponds to a high concentration of paclitaxel under those conditions. Nevertheless, despite high amounts of paclitaxel detected in these variants, its yield was the highest in BABA1MIX, i.e., the treatment where a noticeable rolC expression was also correlated with the upregulation of the other investigated genes (Figures 2 and 6).

Conclusions
In T. × media KT hairy root line cultures, pretreatment with BABA for one week followed by elicitation with a mixture of MeJA, sodium nitroprusside and L-phenylalanine was proved to be the most effective in the stimulation of paclitaxel biosynthesis. This coincided with considerably higher TXS, BAPT and PAM gene expression levels among all investigated genes, which could indicate that the orchestration of gene expression of early and late biosynthesis steps is indispensable for enhanced paclitaxel biosynthesis. The duration of BABA pretreatment seems to have the most pronounced impact on its productivity. The yield of paclitaxel reported in the present study is the highest achieved in T. × media hairy root cultures to date.
Although the knowledge on the regulation of the paclitaxel biosynthesis pathway in the KT hairy root line was broadened, further investigation is necessary to identify its controlling and limiting steps, and on the levels of transcription factor activity, as well as DNA methylation.