6.1.1. Bacterial cell factory as eukaryotic CYP production platform

Escherichia coli has always been the first-choice bacterial cell factory for heterologous CYP expression due to its unparalleled fast growth kinetics, inexpensive and rich complex media, high cell density cultures, and ease of genetic manipulation. With the extensive knowledge on genetics and physiology, many state-of-the-art molecular tools and protocols have successfully been developed for recombinant protein production and purification, facilitating functional and structural studies (Table 1). Besides, the absence of native CYP genes in E. coli makes it ideal for recombinant CYP production due to the advantage of lacking cross-interference. Owing to these advantages, a wide range of CYP-mediated biotransformations have been demonstrated in this bacterial cell factory leading to milligram- to gram-scale production of various compounds (Zelasko et al., 2013; Yun et al., 2006; Ichinose et al., 2015; Ajikumar et al., 2010). However, several factors concerning the prototypical elements of eukaryotic CYPs viz. the presence of N-terminal hydrophobic TMD comprising the conserved proline-rich residues prior to the CYP catalytic domain, demand for the addition of the precursor for heme prosthetic group, and dependency on ET proteins as well as NAD(P)H cofactor for electron supply complicate the heterologous CYP expression (Fig. 3). Remarkably, the major drawback with the bacterial cell factory is that this system often demands the membrane-bound eukaryotic CYPs to undergo significant modifications (mutagenesis/deletions) to obtain soluble fractions, possibly resulting in serious consequences. N-terminal CYP modifications though sporadically resolves the problems, the alterations could influence the native enzyme's interaction and fail the desired catalytic activity. Besides, this approach has turned out to be case-specific (not successful for all membrane-bound CYPs), since several N-terminal altered CYPs were still expressed as membrane fractions due to their indelible hydrophobic topologies. Though various strategies have been constructively developed (discussed in detail in Section 7), there is still no a universal guideline to enable successful eukaryotic CYP overexpression. The success rate varies upon a case-by-case basis, and therefore optimizations have to be performed discretely for every membrane-bound CYP protein.

1. Download : Download high-res image (457KB)
2. Download : Download full-size image

Fig. 3. Challenges associated with the prokaryotic and eukaryotic expression of membrane-bound eukaryotic CYPs.

6.1.2. Yeast cell factory as eukaryotic CYP production platform

Saccharomyces cerevisiae, the eukaryotic counterpart of E. coli serves as an ideal cell factory due to its salient features such as genetic accessibility, rapid growth, and convenient genetic and metabolic engineering approaches. The conventional yeast system is desirable for eukaryotic CYP expression as it offers rich ER environment in combination with the higher eukaryotic protein synthesis machinery, thereby enabling the recombinant expression of full-length membrane-bound CYPs without any genetic modifications/truncations (Table 1). S. cerevisiae offers an innate intracellular heme environment with the presence of three well characterized CYP enzymes including CYP51, CYP56 and CYP61, thereby providing an in vivo self-sufficient background for the heterologous CYPs ensuring complete saturation by hemin prosthetic group. The endogenous CYP environment of yeast cell factory is a significant attribute that lacks in E. coli, where supplementation of expensive heme precursors (5-aminolevuleic acid) is often demanded for the up-regulation of CYP activity. Besides, the added advantage of presence of native S. cerevisiae CPR can directly supply reducing equivalents to the heterologous CYPs, wherein an additional step of recombinant CPR co-expression is eliminated, thus avoiding ancillary stress to the host organism leading to lower CYP production. Another interesting attribute of yeast cell factory is its ability to regenerate the cofactor NADPH intracellularly since the effective coupling between NADPH and CYP is crucial for large-scale applications involving high substrate concentration(Park et al., 2020). Of note, the yeast cell factory allows biocatalytic investigations of recombinant CYPs through a one-pot biotransformation procedure achieved by either growing or resting (non-growing) cells (Durairaj et al., 2016; Lundemo and Woodley, 2015). Alternatively, the recombinant CYPs can be studied by yeast microsomes, favoring in vitro work concerning kinetics and reaction mechanism. Nevertheless, though yeast microsomal experiments are more precise, the factors such as sophisticated isolation process, culture-to-culture variations, and technical constraints in purification make it a complicated and cumbersome process to acquire active functional enzymes (Durairaj et al., 2016; Drăgan et al., 2011; Yan et al., 2017). As a result, yeast whole-cell biotransformation is preferred for CYP functional characterizations owing to its simplicity, straightforward analysis, and biocatalyst stability (Durairaj et al., 2016; Park et al., 2020; Lundemo and Woodley, 2015); though the kinetic and structural investigations are limited. The major drawback with the yeast biotransformation is the protein/product yield especially with the strains expressing heterologous CYPs (Fig. 3). Though gram-scale production is possible in yeast-based fermentations or whole-cell biotransformations, the range of production titer often varies and demands longer incubation (Park et al., 2020). Another limitation of yeast cell factory is that certain substrates and/or products may face some permeabilization issues, as the charged molecules are less likely to permeate multiple cell membranes (Yan et al., 2017).

6.1.3. Amalgamation of conventional bacterial and yeast cell factory platform

On the whole, our current knowledge on eukaryotic CYPs results from the composition of researches on both bacterial and yeast cell factories. On one hand, the bacterial cell factory aids higher yield of eukaryotic CYPs (14–20 nmol/mg protein), although the N-terminal modifications are inevitable. On the other hand, though the yield is relatively low (∼1 nmol/mg protein), yeast cell factory enables expression of membrane-bound CYPs in its native form without any modifications (Hausjell et al., 2018). The space-time yield analysis of several CYP biotransformations in bacterial and yeast cell factories has demonstrated that though the production titer can be comparable (1-2 g/L/h), the yeast system demands longer incubation (100-200 h reaction time) compared to that of bacterial system (Park et al., 2020). Of note, the CYP activity might differ depending upon the surrogate cell factory employed in terms of not only expression rate but also functionality. For instance, the recombinant human CYP2S1 expressed in E. coli and S. cerevisiae systems gave varied results as the enzyme's catalytic specificity was altered (Wu et al., 2006, Nishida et al., 2010). Thus, direct or combinational comparison of full-length membrane-bound CYPs in both the bacterial and yeast expression systems is required to enable a better functional and mechanistic understanding. Interestingly, microbial production of plant benzylisoquinoline alkaloids were successfully established in an combinatory co-culture system (Minami et al., 2008). The transgenic E. coli cells expressing the branchpoint intermediate (S-reticuline) biosynthetic gene were co-cultured with the S. cerevisiae cells expressing CYP80G2 along with other required biosynthetic genes (coclaurine-N-methyltransferase/berberine bridge enzyme). This co-culture system efficiently produced magnoflorine and scoulerine with a final yield of 7.2 and 8.3 mg/L, respectively. Likewise, metabolic pathway distribution within microbial consortium could solve the technical difficulties and significantly enhance the production of natural products. Incorporation of robust production of taxadiene in E. coli and the functional expression of CYP enzyme taxadiene 5α-hydroxylase in S. cerevisiae resulted in the highest production (33 mg/L yield) of oxygenated taxanes (Zhou et al., 2015). Amalgamation of both bacterial and yeast based CYP expression could therefore be a novel approach to solve the limitations and exploit the advantages of each expression system.

6.3.1. Fungal cell factory as eukaryotic CYP production platform

An evolutionarily closer organism, a filamentous fungus could serve as a preferred cell factory for the expression of homologous and heterologous membrane-bound eukaryotic CYPs (Durairaj et al., 2016; Zhang et al., 2021). Filamentous fungi could overcome several limitations faced in traditional expression systems (bacteria / yeast), including mRNA precursor maturation, higher splicing-rate, efficient protein secretion machinery and post-translational modifications (Tanaka et al., 2014; Nevalainen and Peterson, 2014). With the advent of modern biotechnology, several species of filamentous fungi have emerged as promising cell factories for the production of pharmaceutically relevant proteins with salient improvements. Of which, Aspergillus spp. (Meyer et al., 2011) dominate the scene as desirable expression hosts for eukaryotic CYP production as many studies have been reported with Aspergillus nidulans (Liu et al., 2019), Aspergillus oryzae (Cao et al., 2019), Aspergillus niger(Faber et al., 2001), Aspergillus sojae (Araki et al., 2019) and Aspergillus aculeatus(Thiele et al., 2020). Several fungal CYPs involved in the natural product biosynthetic pathways were functionally elucidated using the fungal cell factory with A. oryzae or A. nidulans as heterologous host (Zhang et al., 2021). For instance, the biosynthetic pathways of helvolic acid (Lv et al., 2017), fusidic acid(Cao et al., 2019), and cephalosporin P1 (Cao et al., 2019) from the ascomycetous fungi Acremonium fumigatus, A. fusidioides, and A. chrysogenum were studied using A. oryzae as an expression host system. Another example demonstrated the functional investigation of biosynthetic genes of fungal meroterpenoid from the filamentous fungus Acremonium egyptiacum in Aspergillus spp., wherein the genes ascA-D involved in the biosynthesis of ascofuranone and ascochlorin were expressed in A. oryzae, while AscE-G proteins were expressed in A. sojae high-copy expression system (Araki et al., 2019). Of which, AscE is a soluble CYP/reductase fusion protein that catalyzes stereoselective epoxidation of the terminal double bond of the prenyl group, while AscF (terpene cyclase) and AscG (CYP) are membrane-bound proteins involved in the terpene cyclizationand oxidation of ilicicolin into ascochlorin. Based on the preliminary experimental evidence it is suggested that the heterologous production of fungal CYPs yields better results in fungal cell factory than CYPs of other origins. Recombinant production of the active form of mammalian CYP proteins in fungal cell factory is often limited by the differences in the mode of glycosylation of mammalian and fungal cells. Filamentous fungal system features the high-mannose type of glycosylation, but lacks the mammalian-style terminal sialylation of glycans which may affect the functionality, serum half-life and immunogenicity of recombinant proteins (Nevalainen and Peterson, 2014). In addition, factors such as DNA manipulation, incorrect processing / misfolding and secretory yields may also add up to the list of limitations associated with fungal cell factory for recombinant CYP production (Tanaka et al., 2014; Nevalainen and Peterson, 2014).

6.3.2. Plant cell factory as eukaryotic CYP production platform

Plant CYPs involved in secondary metabolism are sometimes difficult to express in a microbial cell factory, and therefore several plant-based hosts have been developed to procure sufficient protein expression and improve the yield of desired compounds. Plant-based heterologous expression offers several advantages as they permit defined mRNA and protein processing, protein subcellular localization and metabolic compartmentalization, and have essential metabolic precursors and coenzymes (Table 1). Though plant (N- and O-linked) and mammalian (terminal sialylation of glycans) cells feature different glycosylation patterns, the variances do not impair the recombinant protein production; and tremendous efforts have been made for humanization of protein N-glycosylation in plant cell factory (Gomord et al., 2010). A wild relative of tobacco, Nicotiana benthamiana, serving as an efficient production system of flu vaccines at industrial-scale (Marsian and Lomonossoff, 2016), has also acted as a competent cell factory for the heterologous expression of several plant CYPs (Reed et al., 2017). The multifunctional enzyme AsCYP51H10 involved in the modification of C and D rings of the pentacyclic triterpene scaffold into 12,13β-epoxy-3β,16β-dihydroxy-oleanane, was successfully expressed and studied using this transient plant expression system (Geisler et al., 2013). Moreover, co-expression of additional heterologous redox partner is not necessary since the N. benthamiana CPR provides sufficient electron equivalents to mediate CYP catalysis. Recently, the members of CYP79C family belonging to the glucosinolate (GLS) biosynthetic pathway in Arabidopsis thaliana were functionally characterized through Agrobacterium-mediated transient expression in N. benthamiana (Wang et al., 2020). Functional investigation of CYP79C1 and CYP79C2 in the GLS pathway engineered N. benthamiana facilitated simultaneous testing of substrate specificity against multiple aliphatic and aromatic amino acids. The whole-genome sequenced non-vascular plant Physcomitrella patens offered efficient homologous recombination, and facilitated recombinant production of several commercially important pharmaceutical proteins (Khairul Ikram et al., 2017). Furthermore, the cinnamic acid 4-hydroxylase from the aquatic plant Anthoceros agrestis,which could not be expressed in yeast system possibly due to high GC contentand/or different codon usage, was successfully expressed in the haploid plant Physcomitrella patens and functionally characterized at biochemical and molecular levels (Wohl and Petersen, 2020). Specifically, A. thaliana serves as an excellent model for recombinant protein production and has been extensively studied for the developmental and molecular biology, as well as pharmaceutical applications (Von Schaewen et al., 2018). Remarkably, the recently developed Arabidopsis-based recombinant protein production platform (Jeong et al., 2018) is expected to serve as an efficient cell factory for the heterologous CYP production suitable for biochemical and structural studies. Interestingly, the plant cell factory system offers efficient endogenous redox partner system (e.g., N. benthamiana CPR) that can effectively pair up with the heterologous CYPs for direct ET. However, the rich intracellular heme environment with enormous native/endogenous CYPs may also have a downside due to the cross-link effect and/or interference with the desired recombinant CYP protein of interest. Though there are successful instances, functional studies in plant cell factories remain limited with early-stage objectives due to the restrictions including enzyme stability, relatively low turn-over, yield, tedious protein extraction and purification process, and high cost of downstream processing (Schillberg et al., 2019).

6.3.3. Mammalian cell-line factory as eukaryotic CYP production platform

Development of genetically modified mammalian cell lines to functionally express membrane-bound CYPs represents another successful strategy by facilitating practical in vitro approaches for enzymatic characterization, drug metabolism screening and early detection of drug toxicity (Xuan et al., 2016; Satoh et al., 2017) (Table 1). The cell-line factory allows recombinant production of larger and complex proteins including membrane-bound CYPs as it offers inherent transcriptional and translational environment along with appropriate chaperonin, secretory pathway and redox assembly coupled with efficient protein folding, and excellent post-translational modifications. Though the glycosylation pattern of some of the cell lines slightly varies to that of human-type glycosylations; the cell lines can be fine-tuned by codon optimization including glycoengineering which improves the efficacy and enhances the recombinant expression (Hunter et al., 2019). Alternatively, human cell line (e.g., HEK-293) systems represent an ideal source for CYP mediated drug metabolism studies, and allow post-translational modifications of membrane-bound eukaryotic proteins for functional production at high levels. Cell lines can be readily transfected or virally transduced, and the recombinant CYP proteins can be produced either transiently or by stable expression (Hunter et al., 2019). Primary human hepatocytes, hepatic cell lines, and stem cell derived hepatocytes are primarily utilized as in vitro models for the recombinant CYP production and functional studies. Nevertheless, high donor to donor variability, scarcity, limited lifespan, low expression and yield restrict the efficient utilization of major cell lines for enzymatic analysis involving CYP mediated drug metabolism. Currently, several alternative cell line platforms have been developed, and tremendous efforts have been put forward to determine CYP functions with the advances of synthetic biology and next-generation engineering to achieve gram-scale productivity (Hunter et al., 2019; Gutiérrez-González et al., 2019; Boon et al., 2020). For instance, human hepatoma celllines can demonstrate increased stability, ample life-span and accessibility compared to the traditional cell-based assays using primary human hepatocytes. Transchromosomic HepG2 cell lines have facilitated expression of four major CYPs (CYP2C9, 2C19, 2D6, and 3A4) and CPR using the mammalian-derived artificial chromosome vector (Satoh et al., 2017). The expression levels were significantly higher than that of the parental HepG2 cells demonstrating a highly versatile model for the evaluation of drug-drug interactions and screening of hepatotoxicity. Another interesting study demonstrated expression of 14 human CYPs in HepG2-derived cell lines individually with the aid of lentiviral expression system (Xuan et al., 2016). In order to determine the most suitable expression platform for the in vitro CYP enzymatic activities, four different mammalian cell lines (COS-7, HepG2, 293T and 293FT) were functionally investigated with the typical variants of CYP2C9, CYP2C19 and CYP2D6 (Dai et al., 2015). The results indicated that the fast-growing variant of 293 cell line 293FT demonstrated higher levels of expression as well as in vitroactivities. Recently, another interesting study successfully developed an ideal heterologous expression system with three CYP isoforms (CYP1A2, CYP2C9, and CYP3A4) using mammalian 293FT cells by optimizing high-precision conditions (Kumondai et al., 2020). Therein, the highest CYP expression which can be quantifiable by CO-difference spectroscopy was achieved in a lost-cost manner by replacing expensive transfection reagent with cost-effective and efficient substitute PEI-Max, and demonstrated significantly higher enzymatic activity by co-expressing CPR and Cyt B5. Though the mammalian cell line factory serves as an excellent system for recombinant human CYP production and in vitro functional studies, some issues concerning expression, reduced activity and low CYP inducibility remain to be addressed. Besides, factors such as demand for expensive culture media, complex growth requirements, technical demand, time-consuming procedures, lengthy expression phase, lot-to-lot heterogeneity, and scalability are some of the major bottlenecks to be overcome.