Cytochrome P450 from Plants: Platforms for Valuable Phytopharmaceuticals

Cytochrome P450 enzymes are important for biotechnology due to their capacity to modify diverse secondary metabolites that may produce chemicals with pharmacological properties. Most terpenes, flavonoids and alkaloids require P450 catalytic functions to reach their biological activity. In the last ten years, several efforts have focused on the expression and production of these three main types of secondary metabolites in engineered microorganisms and plants using P450 of ethnobotanical origin. Despite this, several P450 coding sequences from plant sources are discovered yearly but only a few have been screened by functional genomics. Amongst them, only a few have shown potentials for use in sustainable production of novel drugs and highly valuable products. Cytochrome P450 involvement in the biosynthesis of these products is discussed in this work. Keywords : Biotechnological platforms, Cytochrome P450, Phytopharmaceuticals, Yield improvement, Terpenes, Flavonoids, Alkaloids, Microbial expression


INTRODUCTION
Plant cytochrome P450 enzymes are important in biotechnology due to their ability to biosynthesize diverse secondary metabolites with biological properties. Numerous reports describe the application of P450 monoxygenases, desaturases and methyl transferases from plants as candidates for creating novel varieties of modified vegetables and bio-products [1][2][3]. P450 DNA coding sequences are also selected to originate pharmaceutical products, because of their importance as target genes for microorganism's metabolic engineering. Organic compounds with therapeutic uses and phytopharmaceuticals directly or indirectly derived from these enzymes are now recognized as critical alternatives for treating different illnesses. This is the case of taxol and the periwinkle alkaloids, which are widely used in cancer therapy [4].
Bioactive secondary metabolites are routinely extracted from the plant itself or produced by synthetic-chemistry in order to scale-up their production. Nevertheless, the sustainable production by those methods is not always successful [5]. Sometimes it is compromised because of many features that distress plants like, the required phenological stage, the specific interactions with biotic and abiotic factors that modify their accumulation, as well as difficulties in the extraction procedures and all associated with high investments [6]. Due to these circumstances, many phytopharmaceuticals usually do not reach the clinical trials stage or the commercial attractive just decline. Currently, biotechnological platforms are gaining grounds in the sustainable production of pharmacologically active metabolites from plants. This alternative is not only in accordance with global ecological legislation, their implementation is also a promising and reliable tool for scaling plant metabolites production with complex chemical structures [7]. Despite the presence of other key enzymes actively involved in the biosynthesis of secondary metabolites, P450 are essential in almost all pathways of plant natural products. They are especially involved in the fine modification of the chemical skeletons originated by diverse synthases. In this review, we present relevant approaches in the production of promising pharmacological metabolites through molecular biotechnology using plant P450 enzymes in microorganisms and other plants.

GENETIC ENGINEERING OF P450 ENZYMES INVOLVED IN TERPENE BIOSYNTHESIS Sesquiterpenes
Up to now, six sesquiterpenes have been successfully produced in heterologous systems by using P450 coding sequences ( Figure 1). The anti-malarial drug, artemisinin (1) is a sesquiterpene lactone biosynthesized by Artemisia spp. It is used as a precursor of artemether which when combined with lumefantrine results in an exceptional drug against the tropical disease called Coartem ® by Novartis. As in many secondary metabolites, artemisinin is present in small quantities in the plant itself, thus limiting its extraction for therapeutic aims. Among first remarkable results on artemisinic acid production was by engineered Saccharomyces cerevisiae yielding over 100 mg L -1 [8].
A recent report described the optimization of artemisinic acid biosynthesis in S. cerevisiae by the insertion of distinct DNA coding sequences isolated from A. annua such as cytochrome b5 (CYB5), an artemisinic aldehyde dehydrogenase (ALDH1), an alcohol dehydrogenase (ADH) and the CYP71AV1 genes [9]. The newly engineered yeast strains generated in this work were able to synthetize artemisinic acid in the order of 25 g L -1 when GAL promoters were induced to express the recombinant proteins. Therefore, artemisinic acid yields were increased ~250 times compared to the first attempt. The same report also describes an easy method for the chemical synthesis of artemisinin [9]. Orthologs of CYP71AV1 gene were reported for several Artemisia species revealing polymorphisms that could influence the transformation of amorpha-4, 11-diene to artemisinic acid [10]. To date, engineered strains that actively biosynthesize artemisinic acid seem to be an attractive, cheaper and effective choice for industrializing the artemisinin production and also to decrease costs of the anti-malarial medicines in the short term.
Nootkatone is a sesquiterpene and the main natural and expensive substance of the smell and flavor of grapefruit. Nootkatone is also an effective and environmentally friendly repellent/insecticide against mosquitos, bed bugs, head lice and other insects [11]. It is nontoxic to humans and is an approved food additive commonly used in foods, cosmetics, and pharmaceuticals [11]. Currently, flavor and fragrances are extracted from the oil of fruit peels to be used in drinks and perfumes. However and because supply of these fruits is limited and have a low amount of these substances, the price of nootkatone is around $ 2,000 a pound and the valencene (4) sells for $ 600 a pound or more [12]. Allylix (San Diego, CA) developed a fermentation process to produce nootkatone after engineering baker's yeast to produce a variety of smells and flavors [12].
The current knowledge of metabolic pathways of valencene and other sesquiterpenes like capsidiol (3), a bicyclic sesquiterpenic phytoalexin from Capsicum and Nicotiana plants [13]; is now used. The genes of 5-epiaristolochene synthase and the valencene synthase enzymes (both with N-terminal thioredoxin modification) in combination with the 5-epi-aristolochene dihydroxylase (CYP71D20); all were inserted and differentially expressed in a specific yeast strain (EPY300). The transgenic organism was able to produce ~250 mg L-1 capsidiol and small quantities of valencene [14]. This improved the production of the phytoalexin in a heterologous system without the need of adding exogenous precursors. Solavetivone (2) is an antifungal phytoalexin derived from a vetispirane-type sesquiterpene premnaspirodiene synthesized by Hyoscyamus muticus. Premnaspirodiene oxidase (HPO-CYP71D55) enzyme seems to be involved in the double oxidation of solavetivone and other vetispirane precursors, showing a multisusbtrate but very valuable activity [15].

Costunolide
(5) is a germacranolide sesquiterpene lactone that shows potent antiproliferative properties in different types of cancer [16]. Costunolide is also considered the precursor of many sesquiterpene lactones with significant biological activities [17]. Expressed sequence tag (EST) sets from Nicotiana benthamiana led to the isolation of germacrene A synthase (GAS), germacrene A oxidase (GAO) and the recently characterized chicory costunolide synthase (CiCos, CYP71BL3) genes. These were co-expressed in S. cerevisiae WAT 11 strain to produce low yields of costunolide [17]. Costic acid is a related sesquiterpene carboxylic acid that shows an interesting cytotoxic activity against several phytopathogenic fungi [18]. The functional characterization and expression of germacrene A synthase/ germacrene A oxidase (TcGAS/TcGAO, CYP71AV2) and costunolide synthase (TcCOS) genes from Tanacetum cinerariifolium in yeast (WAT11), resulted in the production of α (6) and γ (7) costic acid isomers [19].
Zerumbone (8) is a humulene derivative with anti-inflammatory, anti-HIV and potent antitumoral properties that is abundant in Zingiber zerumbet. Zerumbone biosynthesis includes the generation of the 8-hydroxy-α-humulene intermediate. Recent studies revealed that CYP71BA1 is the key enzyme involved in the biosynthesis of this molecule in accordance with its heterologous expression in yeast [20]. Coexpression of four genes from the mevalonate pathway including CYP71BA1 and ZSS1 in Escherichia coli, were able to produce small yields of 8-hydroxy-α-humulene [20].
These novel methods promise a major control on the biochemical synthesis of natural substances avoiding complex protocols for its induction and direct extraction from plant tissues. Table 1 describes the use of selected P450 for producing six sesquiterpenes with relevant biological activity.

Diterpenes
Some tricyclic and tetracyclic diterpenes produced by P450 enzyme activity are shown in Figure 2. Kaurenoic acid (9) is one of the most studied diterpenes with a wide range of biological activities [21]. Beyond the known participation of ent-kaurene oxidases (KO´s) The non-sugar sweeteners global market was calculated to rise from $ 9.2 billion in 2010 up to US$ 9.9 billion in 2016 [22]. Considering the commercial importance of this market, the source for new procedures to get these products was also elicited. Stevia is a sweetener and sugar substitute extracted from the leaves of Stevia rebaudiana. The reconstruction of the steviol metabolic pathway in yeast, including the expression of a fungal KO and the ent-kaurenoic acid 13-hydroxylase (KA-13-H) genes is currently protected by a patent due to the commercial interest in the controlled production of natural glucosides [23]. Genetic engineering for steviol glycosides in yeast is considered a reliable option that may contribute to the optimization of this non-caloric sweetener for industrial aims. This new technology will change the perspectives on stevioside production because of their ecological advantages.
Novel KO-cDNA sequences from medicinal or agronomic plants, such as Montanoa tomentosa (MtKO), Momordica charantia (McEKO), Cucumis sativus (CKO) and the moss Physcomitrella patens (CYP701B1); were recently reported [24][25][26][27]. However, only CYP701B1 and MtKO (CYP701A-type) were expressed in yeast and characterized at biochemical level, the respective enzymes showed resistance to relative high concentrations of azolic compounds [26,28]. This characteristic could further be used in metabolic engineering procedures. The apparent kinetic parameters of MtKO enzyme are also interesting considering that the high value of Km app supports aputative multisubstrate activity [28].
Genetic engineering of ent-kaurene and entkaurenoic acid in yeast was improved by using known KO enzymes from Gibberella fujikuroi and Arabidopsis thaliana [29]. The engineered yeasts containing these recombinant enzymes yield at least 0.5 grams L -1 of ent-kaurenoic acid in fed batch fermentation conditions. Deletions, peptide additions and codon optimization at the 5' end (N-terminal end) of the AtKO1 gene from A. thaliana confer favorable changes in its specific activity [30]. These modifications have contributed to the expression of this P450 in relatively low-cost sustainable microorganisms such as Escherichia coli. In addition, these studies also show that KO possesses multifunctional activity on ent-beyerene and isokaurene skeletons. Further studies on the biochemistry of KO´s from medicinal plants that actively biosynthesize tetracyclic diterpenoids, could reveal interesting enzymatic properties for biotechnology use.
Abietadiene diterpenes present mainly in resins from conifers commonly show insecticide, allelopathic and antifungal effects [31]. Expression of the multifunctional and multisubstrate CYP720B1 gene from Pinus taeda in yeast, demonstrates its capability for metabolizing abietadiene (10) and levopimaradiene (11) skeletons to originate oxygenated diterpenes associated with chemical defense [32]. The expression of CYP720B4 gene from Picea sitchensis in the same system, also demonstrated its effectiveness in the oxidation of C-18 in several abietadienes, levopimaridienes and pimaranes (12),  Ferruginol (13), an abietane diterpene (meroterpene type) with gastro-protective and anti-tumoral activities, has been produced at a yield of 10.5 mg L -1 in transformed WAT11 strain (S. cerevisiae) containing CYP76AH1 from Salvia miltiorrhiza and phyto-CYP reductase genes [33]. The discovery of this enzyme and its participation in tanshinone biosynthesis could be a relevant contribution to the elucidation of the biosynthetic pathway of carnosic acid, a pharmacologically active metabolite and also a chemotaxonomic marker of the Salvia genus. Yeast expression of P450 involved in abietane and piamarane type diterpenes represents the first approach for the generation of novel platforms for the diterpene resin acids controlled production. Table 2 shows recent P450 enzymes used for producing diterpenes with biological activity.

Triterpenes
Production of triterpenes with potential use in pharmacology has been improved in the last years ( Figure 3). Initial attempts to clone P450 sequences involved in triterpene biosynthesis were carried out in Glycine max, with the identification of CYP93E1 and heterologous gene expression in yeast. According to these studies β-amyrin and sophoradiol were transformed into olean-12-ene-3β-24-diol (14) and soyasapogenol B (15). Co-expression of CYP93E1 and β-amyrin synthase genes from this plant in the S. cerevisiae system showed small yields of olean-12-ene-3β-24-diol [34].
(licorice plants). Expression in yeast of CYP88D6 gene from an EST bank of these plants showed its participation in the modification of β-amyrin into 11-oxo-β-amyrin, a putative intermediate in glycyrrhizin biosynthesis [35]. Subsequent studies revealed that CYP72A154 enzyme, was able to oxidize 11-oxo-β-amyrin into glycyrrhetinic acid, which is a glycyrrhizin aglycone [36].

Medicago truncatula is a Mediterranean medicinal plant that actively produces hemolytic saponins.
First biochemical studies of CYP716A12 gene from this plant revealed its capacity for transforming β-amyrin and erythrodiol into oleanolic acid (17), a bioactive triterpene with hepatoprotective, antitumor and antiviral properties [37]. Interestingly, simultaneous studies confirmed that CYP716A12 protein is a multifunctional enzyme able to produce ursolic acid (18) and betulinic acid (19) after expressed in yeast (INVSc1). Homologs of CYP716A12 such as CYP716A15 and CYP716A17 from Vitis vinifera exhibited similar activities like CYP716A12 [38].
Research on ginsenoside biosynthesis (Panax ginseng) reveal specific P450 enzymes involved in the addition of the aglycone portion of such molecules. The oxidation of dammarenediol-II at the C-12 position to protopanaxadiol (20) by CYP716A47 in the yeast strain WAT21 was reported [39]. Subsequent findings revealed that CYP716A53v2 had the protopanaxadiol 6hydroxylase enzyme activity. This one produces protopanaxatriol (21) from the final aglycone protopanaxadiol, which is posteriorly glycosylated to several ginsenosides [40]. Recently, CYP716A52v2 was functionally characterized as a β-amyrin 28-oxydase involved in the biosynthesis of oleanane type ginsenosides [41].

Ganoderma lucidum
Ganoderic acids [43] *P450 involved in the biosynthesis of the bioactive product but not necessarily in the final step. Superscripts correspond to their expression in yeast (Y) Ganoderma lucidum, the millenary oriental fungus with medicinal and nutritional properties, biosynthesize ganoderic acid A (22) that shows a potent antioxidant activity [42]. Recent studies reported a GLCYP450 as a P450 probably involved in the biosynthesis of ganoderic acids [43]. Table 3 shows recent P450 enzymes used for producing triterpenes with biological activity.

GENETIC ENGINERING OF P450 INVOLVED IN FLAVONOID BIOSYNTHESIS
Some flavonoids with nutraceutical or pharmacological properties were produced in heterologous systems by using P450 coding sequences ( Figure 4). Hydroxylation process is critical for flavonoids with dynamic activity. Significant efforts to produce flavanones in S. cerevisiae have been carried out [44]. Starting from natural precursors such as p-coumaric acid, cinnamic acid and caffeic acid, a yield of 28.3 mg L -1 naringerin (23), 16.3 mg L -1 pinocembrin (24) and 6.5 mg L -1 eriodictyol (25) respectively were obtained. In these findings naringerin and pinocembrin were produced 62 and 22 times more efficiently respect to previous attempts in prokaryotic cells.
The reconstruction of anthocyanin pathway inside E. coli has also been achieved [45]. The gene cluster used for this aim included the flavonoid 3' hydroxylase (F3H-P450) anthocyanidin synthase (ANS) from Malus domestica and the flavonoid 3-Oglucosyltransferase from Petunia hybrida. Similar attempts to yield phenylpropanoids by using exogenous precursors and basic cDNA sequences for enzymes involved in flavonoid biosynthesis were posteriorly carried out in both E. coli and S. cerevisiae [46,47]. In these works the participation of flavonoid-3-hydroxylase (F3'H) and flavonoid-3-5-dihydroxylase (F3'5'H) displayed an essential role in the generation of the novel strains specialized in the production of flavonoids. Today F3'H and F3'5'H are key enzymes used in bringing about pigmentation changes in flowers and fruits in order to enhance their nutraceutical properties [48].
Leonard et al reported an interesting approach in engineered E. coli for the production of kaempferol (26) and quercetin (27), which are potent antioxidants and anti-obesity phenylpropanoids [49,50]. This attempt was carried out by reconstructing the basic flavonoid pathway which included the hydroxylating activity of flavonoid 3-β-hydroxylase from Malus domestica (MdFTH) and F3'5'H from Catharanthus roseus. The authors subsequently reported a significant improvement in flavonoid production by the insertion and overexpression of the very active acetyl-CoA carboxylase (PlACC) from Photoharbdus luminicens [51].
Vannelli et al described the production of phydroxycinnamic acid (28) from glucose in yeast [52]. The novel strain was achieved by using both phenylalanine and tyrosine ammonia lyases (PAL, TAL) genes from the fungus Rhodotorula glutinis. Both were simultaneously expressed with cinnamate-4-hydroxylase (C4H-P450 monooxygenase) and P450 CYP reductase genes from Helianthus tuberosus. Improvements in the production of p-hydroxycinnamic acid (up to 700 mg L-1) were reached in Streptomyces lividans using TAL and an endoglucanase from Rhodobacter sphaeroides [53]. Despite the latter advances, current engineered microorganisms for producing hydroxylated flavonoids require exogenous precursors to generate them. Therefore, a great challenge for flavonoid genetic engineering is to get self-sufficient microorganisms able to produce those metabolites without the addition of external intermediates. This condition has only been successfully achieved for relatively simple stilbene flavonoids such as resveratrol (29) in transgenic E. coli strains [54]. Table 4 shows some conserved P450 monooxygenases used for scale the production of hydroxylated flavonoids.

Helianthus tuberosus
p-hydroxycinnamic acid [52] *P450 involved in the biosynthesis of the bioactive product but not necessarily in the final step. Superscripts correspond to their expression in yeast (Y) and bacteria (B)

GENETIC ENGINEERING OF P450 ENZYMES INVOLVED IN ALKALOID BIOSYNTHESIS
Advances in the synthesis of pharmacologically active alkaloids by genetic engineering have been reported ( Figure 5). Hyoscyamus niger commonly known as blackhenbane, is an European and North American medicinal plant that biosynthesizes high concentrations of tropane alkaloids such as hyoscyamine (30). Li et al demonstrated the participation of CYP80F1 in the rearrangement of (R)-littorine to (S)hyoscyamine aldehyde, an intermediary in the biosynthesis of hyoscyamine [55]. Novel insights describe that CYP80F1 catalyzes the isomerization and hydroxylation of littorine at the 3′-position [56]. CYP719B1 gene was isolated from Papaver somniferum and showed a salutaridine synthase identity. This enzyme catalyzes the conversion of (R)-reticuline to salutaridine (31), a crucial step in morphine biosynthesis [57]. Due to the medical importance of morphine (32), a specific transgenic opium poppy line was obtained by the insertion of CYP80B3 ((S)-N-methylcoclaurine 3'hydroxylase) gene, their over-expression resulted in > 400 % increase of the total alkaloid content in the plant [58].

CONCLUSION
Biotechnological platforms have shown a significant advance in the past few years, especially in the synthesis of terpenes in heterologous systems. This undoubtedly demonstrates the potential uses of P450 enzymes as key steps for creating novel microorganisms that actively produce high amounts of plant natural products in a short time. This condition should help to reduce the cost of pharmaceuticals for treating public health problems. Up to date, the artemisinin production in engineered yeasts is the most concrete example of success in the heterologous production of terpenes [9]. Yeast is the preferred model for the induction of microsomal P450. However, genetic engineering of the aminoterminus have currently revealed promising results for expressing this group of enzymes in more reliable microorganisms as E. coli [30]. The number of P450 coding enzymes for triterpene biosynthesis has particularly increased in the last nine years. EST technology has significantly contributed to the availability of unique sequences from medicinal plant resources supporting the discovery of novel P450 with pharmacological potential.
Substantial advances have been achieved in the production of relatively simple flavonoids as the case of resveratrol (3,5,4'-trihydroxy-transstilbene) in E. coli [54]. Conversely, in many cases the microbial synthesis of hydroxylated flavonoids with a complex structure still depends on exogenous precursors to get enough yields.
Production of alkaloids and pseudo alkaloids in heterologous systems is another big challenge for synthetic biology, considering the exceptional reactions that they carry out and their diversity in the plant kingdom [3]. It is probable that functional genomics for alkaloid production is in full swing due to the recent studies validating their high scale production in genetically modified microorganisms as E. coli and S. cerevisiae [62]. Role of cytochrome P450 enzymes in alkaloids biosynthesis are quite specific compared with those involved in the biosynthesis of flavonoids and terpenes. Nonetheless, there are fascinating advances in the generation of transgenic lines of opium poppy for producing benzylisoquinoline derivatives with appreciated activity in the pharmaceutical market.