Engineering yeast for sustainable production of sterols, steroids, and bile acids: advances, strategies, and future directions

1 Introduction

Sterols, steroids, and bile acids are structurally related lipid molecules essential for membrane organization, signal transduction, and metabolic regulation. While traditionally studied as isolated physiological mediators or pharmaceutical agents, emerging evidence reveals they are not merely metabolic end products, but dynamic regulators of homeostasis and disease. This paradigm shift emphasizes the need to understand these molecules as integral components of an interconnected, systems-level metabolic network (Figure 1).

Despite sharing structural and enzymatic origins, sterols and steroids are distinct classes of high-value bioactive molecules (Miller and Auchus, 2011). While sterols are structural essentials for membrane integrity and cellular homeostasis (Hu et al., 2017), steroids act as potent physiological effectors and key therapeutics, active even at remarkably low concentrations (Dumas et al., 2006).

Beyond these classical roles, bile acids are increasingly recognized as dynamic signaling molecules. They coordinate diverse processes–from energy metabolism and inflammation to host-microbiome interactions (Chiang, 2013; Hofmann and Hagey, 2014). Based on these diverse functions, bile acids have become highly attractive therapeutic candidates for metabolic and immune diseases, representing some of the most critical nodes in the sterol metabolic network.

Saccharomyces cerevisiae is a highly effective platform for studying and engineering complex lipid metabolism (Nielsen and Keasling, 2016). Combining genetic tractability, industrial robustness, and essential features, it serves as an excellent model for reconstructing and dissecting complex mammalian metabolic pathways (Botstein and Fink, 2011). Because its native sterol networks and regulatory circuits are exceptionally well characterized, S. cerevisiae provides a definitive chassis for investigating intricate lipid homeostasis (Espenshade and Hughes, 2007).

In this review, we summarize the applications of sterols, steroids, and bile acids, critically assessing the limitations of S. cerevisiae as a production chassis and the engineering strategies used to overcome them. Rather than cataloging individual target molecules, we highlight how yeast platforms have evolved toward the rational production of functional lipid derivatives. Finally, we explore the potential of S. cerevisiae as an expandable chassis for bile acid biosynthesis.

2 Sterol

Sterols are essential membrane components that regulate fluidity and permeability. Based on their biological origin, sterols are primarily classified as cholesterol in animals, phytosterols in plants, and ergosterol in fungi (Hu et al., 2017). We focus here on two key sterols (Figure 2A). Ergosterol, yeast’s native sterol, establishes the foundational metabolic network. Conversely, while cholesterol is not naturally synthesized in yeast, its biosynthetic pathway can be reconstructed to serve as a key branching precursor for steroid and bile acid biosynthesis. For both sterols, we explore their biological and industrial relevance, characterize the metabolic constraints limiting high-level production, and highlight the engineering strategies designed to overcome these bottlenecks.

2.1 Ergosterol

2.1.1 Biological significance and production potential

Ergosterol is a major structural component of fungal membranes, dictating fluidity and permeability. Functioning analogously to cholesterol in animals, it serves as a key lipid moiety required for the proper activity of membrane proteins and transporters (Parks and Casey, 1995). Beyond its physiological significance, ergosterol exhibits broad industrial and pharmaceutical potential. It is a precursor for vitamin D₂ synthesis, and its pathway intermediates offer starting points for diverse steroid-based pharmaceuticals (Tan et al., 2003). Thus, ergosterol acts as a critical metabolic hub, highlighting the value of yeast as an engineering platform for lipid-derived compounds.

2.1.2 Native pathway and structural constraints

In yeast, ergosterol biosynthesis is a complex cascade requiring over 25 enzymes to convert acetyl-CoA into the mature sterol (Bhattacharya et al., 2018). Because ergosterol is essential for membrane integrity and cell viability, its production is not a simple linear pathway. Instead, it operates as a highly coordinated metabolic network, strictly regulated by cellular energy status, oxygen availability, and intracellular sterol levels (Gu et al., 2021).

Functionally, the ergosterol biosynthetic pathway is divided into three modules (Figure 2B): mevalonate (MVA) synthesis, farnesyl pyrophosphate (FPP) synthesis, and the downstream ergosterol synthesis (Hu et al., 2017). The initial module converts acetyl-CoA to mevalonate. Two molecules of acetyl-CoA condense to form acetoacetyl-CoA, which then reacts with a third acetyl-CoA to generate 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA). HMG-CoA is then reduced to mevalonate by HMG-CoA reductase (HMGR), which is driven by two yeast isoenzymes, Hmg1p and Hmg2p (Miziorko, 2011). As the primary rate-limiting enzyme, HMGR acts as a key regulatory node controlling overall sterol biosynthetic flux (Bhattacharya et al., 2018).

The second module converts mevalonate into FPP via isoprenoid intermediates. Beyond fueling sterol biosynthesis, FPP is a central metabolic hub required for vital processes including protein prenylation. Because depleting FPP compromises cell viability, it acts as a critical branch point linking sterol production to fundamental cellular survival.

The final module encompasses the post-squalene pathway, where ergosterol is synthesized from FPP. This module begins with the condensation of two FPP molecules to form squalene and proceeds through a series of sequential enzymatic reactions, representing the most complex segment of the pathway. Notably, multiple oxidative steps are required to remodel the sterol backbone. These reactions are strongly dependent on oxygen availability and metal cofactors. Key enzymes, including squalene monooxygenase (Erg1p), lanosterol 14α-demethylase (Erg11p), C-4 sterol methyl oxidase (Erg25p), sterol Δ5-desaturase (Erg3p), and sterol Δ22-desaturase (Erg5p), catalyze oxygen-dependent oxidation reactions and are therefore highly sensitive to changes in oxygen levels (Gu et al., 2021). Iron also plays an essential role in multiple reactions within the post-squalene pathway. Erg11p and Erg5p belong to the cytochrome P450 family and catalyze oxidative reactions using a heme cofactor. Erg25p and Erg3p are members of the fatty acid hydroxylase/sterol desaturase family and function as oxo-diiron enzymes, catalyzing oxidation at active sites that coordinate two iron ions (Shakoury-Elizeh et al., 2010). As a result, the post-squalene module is highly sensitive to oxygen and iron availability and frequently contributes to a major bottleneck in ergosterol biosynthesis.

Ergosterol biosynthesis is an energy-intensive process demanding substantial ATP and NADPH. As membrane homeostasis is critical for cell viability, this pathway operates under strict feedback control (Hu et al., 2017). When sterol levels exceed a threshold, metabolic flux is suppressed, and excess free sterols are converted into inert steryl esters for storage in lipid droplets (LDs). This buffering system stabilizes the intracellular sterol pool and limits the overaccumulation of membrane-active free sterols (Garbarino et al., 2009). While these multi-layered regulatory networks are essential for yeast survival, they impose significant bottlenecks on metabolic engineering efforts aimed at maximizing ergosterol production.

2.1.3 Engineering strategy

Early strategies to boost yeast ergosterol production relied on overexpressing native pathway enzymes (Table 1). A prime target was HMGR, the rate-limiting step of the mevalonate pathway. Because overexpressing full-length, ER-anchored Hmg1p triggers massive ER expansion (karmellae) and cellular stress, Polakowski et al. (1998) engineered a truncated, cytosolic form of Hmg1p. This soluble enzyme lacked the ER-anchoring domain, effectively bypassing sterol-mediated feedback inhibition while driving robust mevalonate flux. Nevertheless, the accumulation of squalene indicated that major downstream bottlenecks remained unresolved. To resolve this squalene bottleneck, Veen et al. (2003) systematically overexpressed post-squalene ERG genes in a tHMG1-overexpressing strain. Co-overexpressing ERG1 and ERG11 successfully drove squalene conversion into lanosterol, tripling overall sterol flux compared to wild-type levels. However, final ergosterol yields increased only marginally, indicating further downstream limitations.

TABLE 1. Stepwise engineering strategies for ergosterol pathway rewiring in S. cerevisiae.

indicates overexpression.

Collectively, these studies demonstrate that overexpression of individual enzymes cannot substantially elevate ergosterol yields. The core limitation is not a single enzymatic bottleneck, but the rigid regulatory control governing the entire pathway. Veen et al. (2003) highlighted this by showing that ERG1 and ERG11 overexpression primarily drives the accumulation of esterified sterols in LDs, rather than increasing the active free sterol pool. This forced esterification demonstrates that membrane-compatible sterol levels are tightly constrained by feedback regulation. Consequently, LD formation creates a structural barrier that effectively decouples pathway flux from ergosterol accumulation.

To bypass strict intracellular sterol homeostasis, He et al. (2007) co-overexpressed ERG4 and ARE2, encoding sterol Δ24 (28)-reductase and sterol acyltransferase, respectively. By rapidly converting newly synthesized ergosterol into steryl esters, Are2p acts as a metabolic sink that prevents free sterol accumulation. This forced esterification effectively alleviates feedback repression, pulling metabolic flux forward. As a result, the engineered strain yielded 49.2 mg/g DCW of total ergosterol (sum of free and esterified ergosterol), a 1.46-fold increase over the control in flask fermentations.

As targeting single enzymes failed to rebalance ergosterol metabolism, metabolic engineers pivoted to global regulatory strategies (Wang et al., 2018). In yeast, the Zn(II)₂Cys₆ transcription factors Upc2p and Ecm22p act as master regulators of this pathway. By binding to a conserved 7-bp sterol regulatory element (SRE) across multiple ERG promoters (e.g., ERG2 encoding sterol Δ8-isomerase, ERG3, and ERG11), they drive the coordinated, network-wide upregulation of sterol biosynthesis (Yang et al., 2015). Validating this strategy, overexpressing ECM22 increased ergosterol yields by 20% (32.7 mg/g DCW) in 5-L bioreactors (Wang et al., 2018), while overexpressing UPC2 achieved a 24.5% improvement (11.91 mg/g DCW) in flask cultures (Ma et al., 2018).

More recently, engineering strategies have evolved beyond merely boosting biosynthesis to actively expanding cellular storage capacity. For instance, overexpressing ACC1 encoding acetyl-CoA carboxylase—a key enzyme in fatty acid biosynthesis—promotes LD formation, creating expanded reservoirs for steryl esters. Leveraging this, Sun et al. (2021) integrated pathway activation (Upc2p), esterification (Are2p), and expanded storage (Acc1p). This synergistic approach yielded 40.6 mg/g DCW of ergosterol in flask fermentations, a 3.6-fold increase compared with the control strain. Ultimately, this success underscores that maximizing ergosterol production demands systems-level coordination rather than isolated enzyme tweaks.

2.2 Cholesterol

2.2.1 Biological significance and production potential

Cholesterol is a major structural component of eukaryotic membranes, dictating fluidity, microarchitecture, and the spatial organization of signaling proteins (Burger et al., 2000). Crucially, the dysregulation of cholesterol-rich microdomains—known as lipid rafts—drives the pathogenesis of diverse conditions, ranging from neurodegenerative disorders and cardiovascular diseases to systemic lupus erythematosus and acquired immunodeficiency syndrome (AIDS) (Michel and Bakovic, 2007). Therefore, cholesterol is no longer viewed merely as a passive structural scaffold, but as a dynamic bioactive molecule central to cellular signaling and disease pathology.

Initially, cholesterol-producing yeast were engineered merely as model systems to study how sterol structure dictates membrane and protein function (Shivapurkar et al., 2011; Souza et al., 2011). Yet, these foundational models provided the essential conceptual and technical frameworks for advanced metabolic engineering. Today, constructing a robust cholesterol-producing yeast chassis is a strict prerequisite for extending sterol engineering downstream toward complex bile acid biosynthesis.

2.2.2 Native pathway and structural constraints

Cholesterol production in yeast relies on rewiring the native S. cerevisiae pathway by eliminating ergosterol-specific branches and expressing mammalian sterol reductases (Souza et al., 2011). Yeast, mammals, and plants share an identical upstream pathway terminating at zymosterol, meaning species-specific differences emerge only during downstream modifications (Nes, 2011). Therefore, the specific combination of post-zymosterol enzymes strictly determines the final sterol profile (Figure 2C).

In wild-type yeast, zymosterol is converted to ergosterol through sequential reactions driven by sterol Erg6p, Erg5p, and Erg4p. Because this native fungal branch yields sterols structurally distinct from cholesterol, it effectively blocks the functional integration of mammalian enzymes and downstream metabolic pathways that strictly require cholesterol as a substrate.

Sterol uptake presents an additional constraint. Although yeast expresses sterol transporters such as ATP-binding cassette sterol transporter (Aus1p) and ATP-binding cassette multidrug transporter (Pdr11p), their activity is strictly restricted to anaerobic or specific stress conditions (Kohut et al., 2011). Consequently, exogenous cholesterol cannot be reliably imported to fuel downstream metabolic pathways. To bypass this transport limitation, engineering strategies must drive de novo intracellular cholesterol biosynthesis rather than relying on external supplementation.

2.2.3 Engineering strategies in S. cerevisiae

2.2.3.1 Deletion of ergosterol biosynthetic genes

To enable cholesterol biosynthesis, the initial step requires eliminating the native ergosterol-specific branch (Figure 2C). Deletion of ERG5 and ERG6 strictly blocks ergosterol formation, driving instead the accumulation of 7-dehydrodesmosterol (cholesta-5,7,24-trienol) (Souza et al., 2011). Because the intermediate lacks the C24 methyl group and the C22 double bond, it serves as a suitable precursor for cholesterol biosynthesis. Ultimately, this approach fundamentally redirects the yeast sterol pathway, establishing the structural baseline for de novo cholesterol production.

2.2.3.2 Introduction of heterologous genes

With the native ergosterol branch eliminated, the next phase required introducing heterologous enzymes to reconstruct mammalian cholesterol biosynthesis. Early studies expressed 7-dehydrocholesterol reductase (Dhcr7) and 24-dehydrocholesterol reductase (Dhcr24) from Danio rerio to reduce specific double bonds (C7 = C8 in the B-ring and C24 = C25 in the side chain). This strategy successfully engineered a strain where cholesterol comprised over 68% of the total sterol pool. Importantly, this provided the first in vivo proof that cholesterol can fully replace ergosterol as the dominant functional sterol (Welscher et al., 2008).

However, plasmid-based systems often suffer from variable expression and genetic instability during prolonged cultivation, leading to inconsistent sterol profiles. To alleviate this issue, subsequent studies integrated dhcr7 and dhcr24 directly into the disrupted ERG5 and ERG6 loci. This genomic integration stabilized enzyme levels and balanced pathway flux, generating a robust yeast strain with over 95% cholesterol of the total sterols (Shivapurkar et al., 2011; Souza et al., 2011). Therefore, these results demonstrate that beyond simply selecting the right enzyme, stable chromosomal expression is a critical prerequisite for engineering high-value downstream steroidal pathways.

2.2.3.3 Fermentation engineering

Despite the successful construction of cholesterol-producing yeast, early studies yielded low titers (∼1 mg/g fresh cell weight), impeding their use as platforms for high-value steroids or bile acids. Addressing this without further genetic modifications, Wang et al. (2022) optimized fed-batch fermentation for the strain developed by Souza et al. (2011). They demonstrated that combining high glucose with corn steep liquor significantly boosted both cell mass and carbon flux into the mevalonate pathway. Complementing these nutritional optimizations, supplementing cultures with calcium and zinc ions enhanced cellular tolerance to fermentation-induced stress, thereby maximizing cholesterol accumulation (Cheng et al., 2017).

Notably, citric acid supplementation served as an effective process intervention. Rather than directly regulating specific enzymes, exogenous citric acid appears to alter intracellular metabolic states. Consistent with the observed decrease in citrate synthase (CIT1) transcription and the upregulation of mevalonate pathway genes, these changes may be associated with a redistribution of carbon flux toward sterol biosynthesis. Collectively, these findings highlight the potential of citric acid supplementation to modulate metabolic flux beyond its conventional role as a carbon source.

Furthermore, employing a fed-batch strategy with periodic glucose and citric acid feeding under controlled dissolved oxygen ensured stable metabolism during extended cultivation. This bioprocess optimization boosted cholesterol titers to ∼7.1 mg/g fresh cell weight in a 5-L bioreactor (Wang et al., 2022). This improvement represents a critical step in elevating cholesterol from a proof-of-concept sterol to a practical substrate for downstream bile acid and high-value steroid production.

3 Steroid (pregnenolone and progesterone)

Steroids are high-value compounds that exhibit diverse physiological activities through minor structural modifications, making them essential raw materials in the pharmaceutical industry. Despite advances in chemical synthesis, including the development of new catalytic methods, these approaches remain limited by the structural complexity of steroids, which hinders precise and site-selective modifications due to the presence of multiple similar reactive sites. Moreover, their reliance on complex, multi-step transformations—requiring intricate redox control and stereoselective assembly of polycyclic ring systems—reduces overall efficiency, scalability, and practical sustainability (Khatri et al., 2020; Paccoia et al., 2025; Wang and Gui, 2025; Wojtkielewicz et al., 2024). Consequently, microbial biosynthesis has emerged as a promising and scalable alternative, leveraging enzyme specificity to enable more selective and sustainable production of steroid compounds (Dumas et al., 2006; Gu et al., 2021).

3.1 Biological significance and production potential

Pregnenolone and progesterone occupy the critical entry point of steroid hormone biosynthesis, serving as master precursors for glucocorticoids, mineralocorticoids, androgens, and estrogens. In mammals, this cascade begins when the cholesterol side-chain cleavage enzyme (CYP11A1) converts cholesterol into pregnenolone, which is then oxidized to progesterone by 3β-hydroxysteroid dehydrogenase (3β-HSDH) (Miller and Bose, 2011).

Beyond serving as metabolic intermediates, pregnenolone and progesterone independently regulate neural activity, immunity, and homeostasis. Given these properties, both compounds are widely used as key materials in pharmaceutical development (Baulieu, 1998; Miller and Auchus, 2011). As traditional chemical synthesis is constrained by feedstock availability, high costs, and process complexity, microbial de novo biosynthesis has emerged as a more sustainable and scalable alternative. Developing engineered yeast strains to produce pregnenolone and progesterone is a significant advance for the steroid industry (Fernandes et al., 2003). Consequently, pregnenolone and progesterone are not only primary targets for microbial engineering but also critical milestones for synthesizing more complex steroidal therapeutics.

3.2 Native pathway and structural constraints

In mammals, pregnenolone biosynthesis relies on precise intracellular cholesterol trafficking and organelle-specific microenvironments. Specifically, cholesterol must be transported to the inner mitochondrial membrane, where CYP11A1 cooperates with dedicated electron transfer proteins to catalyze side-chain cleavage (Figure 3A). Thus, effective steroid production requires not only the specific enzymes but also the host cell’s proper spatial organization and electron transfer systems (Miller, 2013).

Conversely, yeast primarily synthesizes ergosterol, which differs from cholesterol by a C7 double bond, a C22 double bond, and a C24 methyl group (Figure 2A). These structural features dramatically limit its compatibility with mammalian enzymes (Duport et al., 1998). Moreover, yeast lacks the specialized mitochondrial transport machinery central to mammalian steroidogenesis and cannot efficiently import exogenous cholesterol under aerobic conditions (Jacquier and Schneiter, 2012). Together, these structural and systemic mismatches present significant challenges to producing pregnenolone and progesterone in native yeast.

3.3 Engineering strategy

To bypass these structural constraints, yeast must be engineered to synthesize de novo sterols recognizable by CYP11A1, rather than relying on exogenous cholesterol. Addressing this, rewiring the ergosterol pathway to generate cholesterol-like substrates is essential for compatibility with mammalian enzymes (Dumas et al., 2006; Duport et al., 1998). Duport et al. (1998) engineered yeast to accumulate the cholesterol analog ergosta-5-eneol by expressing Arabidopsis thaliana Δ7-reductase (DWF5) and disrupting Δ22-sterol desaturase (Erg5p), subsequently driving the successful conversion of the ergosta-5-eneol into steroid hormones through the introduction of mammalian CYP11A1, its electron transfer proteins adrenodoxin (ADX) and adrenodoxin reductase (ADR), and 3β-HSDH. This study provided the first demonstration of pregnenolone and progesterone biosynthesis in yeast without any external supplementation.

Duport et al. (2003) investigated the side-chain cleavage (SCC) mechanism in yeast under identical substrate–enzyme conditions. Notably, they uncovered a striking spatial mismatch: while mammalian CYP11A1 operates within the inner mitochondrial membrane, the enzyme predominantly localizes to the plasma membrane (PM) in engineered yeast. Furthermore, yeast SCC enzymes lack the coordinated colocalization found in mammals, residing instead in distinct compartments: ADR in the endoplasmic reticulum (ER), ADX in the cytosol, and CYP11A1 on the PM (Figure 3B). Despite this spatial fragmentation, yeast remains capable of synthesizing pregnenolone and progesterone. These findings indicate that SCC efficiency is determined by electron transfer kinetics rather than substrate availability, suggesting ADX as the primary rate-limiting factor in this system. Collectively, these results reveal that SCC efficiency in yeast is dictated by the functional integrity of the electron transfer chain rather than physical enzyme localization. This establishes the redox link from NADPH to CYP11A1—not spatial organization—as the primary bottleneck in heterologous steroid biosynthesis (Duport et al., 2003).

This insight reframed electron transfer from a supporting component into a primary engineering target. Chen et al. (2025) leveraged 7-dehydrocholesterol (7-DHC) as a cholesterol-like sterol intermediate, first enhancing cholesterol biosynthesis by redirecting electron flux from NADPH toward the Bos taurus 7-dehydrocholesterol reductase (DHCR7) active site. To optimize the subsequent conversion to pregnenolone, the authors replaced the conventional mammalian CYP11A1 with a plant-derived cytochrome P450 enzyme, RgCYP87A3 (Figure 3C), and its cognate reductase (ATR). To further improve efficiency, RgCYP87A3 and ATR were fused via a linker, reducing the physical distance between the proteins and enabling direct electron transfer rather than diffusion-mediated interactions. This approach systematically addressed the electron transfer limitation conceptually proposed by Duport et al. (2003), setting a new benchmark for pathway-level electron transfer efficiency (ETE).

Building on early work focused on pregnenolone and progesterone, recent studies have expanded the scope of steroid biosynthesis in S. cerevisiae. Complex corticosteroids such as hydrocortisone have been produced by reconstructing fungal oxidation systems and optimizing transmembrane flux (Chen et al., 2020). In addition, plant-derived steroids such as diosgenin have been produced at the gram scale by improving P450 electron transfer and fine-tuning sterol metabolism and transport (Xu et al., 2022). Recently, these advances have extended beyond the biosynthesis of naturally occurring steroids. Building on established platforms, researchers have combined enzyme engineering with metabolic network design to produce non-natural steroid derivatives, including 6-dehydroprogesterone (Li et al., 2025). This non-natural progestogen with therapeutic potential highlights the critical role of C6 modification in shaping steroid activity. These developments demonstrate that biosynthetic strategies can be integrated not only to reconstruct native pathways but also to enable the rational design of novel steroid structures. Collectively, these studies underscore the expanding versatility and industrial relevance of yeast as a platform for steroid biosynthesis.

4 Bile acids

4.1 Biological significance and production potential

In the human liver, cholesterol is converted into primary bile acids, which are secreted into the bile ducts, stored in the gallbladder, and released into the intestine to facilitate lipid digestion. Approximately 95% of bile acids are subsequently reabsorbed in the distal intestine and returned to the liver via enterohepatic circulation, allowing their repeated reuse (Chiang, 2013). A small fraction of bile acids escapes hepatic recirculation and enters the intestinal lumen, where they interact with the gut microbiota. Primary bile acids—cholic acid (CA) and chenodeoxycholic acid (CDCA)—are enzymatically transformed into secondary bile acids by microbial metabolism (Funabashi et al., 2020). Beyond their classical role as detergents that facilitate lipid digestion in humans (Hofmann and Hagey, 2014), bile acids have emerged as vital endocrine signaling molecules that regulate energy metabolism, immune responses, and intestinal homeostasis. Bile acids precisely regulate lipid and glucose metabolism, as well as inflammatory signaling in the liver and intestine, by activating receptors such as farnesoid X receptor (FXR) and the G protein–coupled receptor TGR5 (Chiang, 2013). Among bile acids, ursodeoxycholic acid (UDCA) is widely used clinically to treat cholestatic liver diseases and gallstones. In addition, diverse bile acid derivatives are emerging as promising therapeutics for broader metabolic and inflammatory disorders (Chen and Gong, 2025).

Despite their clinical value, bile acids are still primarily sourced through animal bile extraction or multi-step chemical synthesis. These traditional methods are severely hindered by supply instability, ethical concerns, and the risk of viral contamination. Thus, current manufacturing capability cannot sustainably meet the growing demand for these essential therapeutics. To overcome these bottlenecks, engineered microorganisms—particularly yeast—have rapidly emerged as a robust, scalable, and sustainable platform for bile acid synthesis (Feng et al., 2009).

4.2 Native pathway and structural constraints

4.2.1 Primary bile acid pathway

In human hepatocytes, bile acid biosynthesis is partitioned into two main routes: the classical pathway, responsible for ∼90% of production, and the alternative (acidic) pathway, accounting for the remaining ∼10% (Fleishman & Kumar, 2024) (Figure 4). The classical pathway is determined by its initiating and rate-limiting enzyme, cholesterol 7α-hydroxylase (CYP7A1), which catalyzes the 7α-hydroxylation of cholesterol. Subsequently, the enzyme 3β-hydroxy-Δ5-C27-steroid oxidoreductase (HSD3B7) converts the resulting 3β-hydroxy-Δ⁵ sterol intermediates (e.g., 5-cholesten-3β,7α-diol) into their active 3-oxo-Δ⁴ forms (e.g., 4-cholesten-7α-ol-3-one).

From this point, the pathway bifurcates depending on sterol 12α-hydroxylase (CYP8B1) activity. In the CA-producing branch, CYP8B1 mediates a critical 12α-hydroxylation. The steroid A-ring is then reduced by steroid 5β-reductase (AKR1D1) and 3α-hydroxysteroid dehydrogenase (AKR1C4), preparing the molecule for terminal oxidation. Finally, sterol 27-hydroxylase (CYP27A1) drives a cascade of sequential oxidations—yielding alcohol, aldehyde, and carboxylic acid intermediates—to form the direct precursor of CA. Conversely, the CDCA-producing branch completely bypasses CYP8B1-mediated 12α-hydroxylation. In this route, AKR1D1, AKR1C4, and CYP27A1 operate sequentially to yield dihydroxycholestanoic acid (DHCA), the direct precursor for ultimate conversion into CDCA (Ferdinandusse and Houten, 2006).

The alternative pathway is initiated by CYP27A1, which oxidizes cholesterol at the C27 position to progressively form alcohol, aldehyde, and carboxylic acid intermediates. Next, oxysterol 7α-hydroxylase (CYP7B1) executes 7α-hydroxylation, functionally mirroring CYP7A1 in the classical route. The pathway then funnels into the same enzymatic cascade as the classical CDCA branch, utilizing HSD3B7, AKR1D1, AKR1C4, and CYP27A1. Notably, because the C27 position is already pre-oxidized during initiation, CYP27A1 requires only one or two final oxidations to complete the terminal carboxylic acid conversion, distinguishing its workload from the classical pathway (Schwarz et al., 1996).

4.2.2 Secondary bile acid pathway

UDCA is generated from the primary bile acid CDCA via C7 epimerization, a structural shift that converts the 7α-hydroxyl group into a 7β configuration. Mechanistically, this transformation relies on a two-step microbial redox cascade: 7α-hydroxysteroid dehydrogenase (7α-HSDH) first oxidizes CDCA into a 7-oxo intermediate, which is subsequently reduced by 7β-hydroxysteroid dehydrogenase (7β-HSDH) to yield UDCA (Song et al., 2017). Notably, this pathway is primarily mediated by microbiota-derived HSDH enzymes rather than host enzymes. Thus, UDCA represents a key metabolite at the host–microbiome interface, with microbial composition and enzymatic capacity serving as critical determinants of its production.

4.3 Engineering strategy

4.3.1 Optimization of precursor (cholesterol) supply

To date, the de novo biosynthesis of primary bile acids in yeast has not been fully achieved. As all bile acids originate from cholesterol, establishing efficient cholesterol supply is essential. However, yeast cannot efficiently import exogenous cholesterol across its PM (Crowley et al., 1998). Consequently, engineering strategies have focused on enabling intracellular de novo cholesterol biosynthesis. Furthermore, as native S. cerevisiae synthesizes ergosterol rather than cholesterol, rewiring endogenous sterol metabolism from an ergosterol-centered pathway to a cholesterol-producing network has been proposed as a key prerequisite for successful bile acid engineering.

4.3.2 Engineering yeast with the primary bile acid biosynthetic pathway

Using the engineered cholesterol-producing strain, the entire 11-gene CDCA biosynthetic pathway was integrated into the yeast genome. Although all introduced genes were actively transcribed, the final CDCA product remained undetectable. Stepwise verification was initially precluded by the unavailability of commercial intermediate standards. However, subsequent stepwise pathway reconstruction and LC-MS/MS analysis uncovered a critical roadblock. While partial conversion was observed during early steps, intermediates beyond the third enzyme, AKR1D1, were completely absent, highlighting a major metabolic bottleneck that subsequent studies have yet to resolve (Moon, 2025).

4.3.3 Production of secondary and conjugated bile acids

Although de novo synthesis of primary bile acids from cholesterol has not yet been achieved in yeast, the conversion of CDCA to UDCA has been successfully implemented in S. cerevisiae. In this strategy, gut microbiota–derived 7α-HSDH and 7β-HSDH were introduced into yeast to catalyze C7 epimerization of CDCA. The 7α-hydroxyl group was converted to the 7β configuration through a sequential oxidation–reduction reaction. Importantly, optimizing the cofactor preferences of these enzymes to align with the intracellular redox balance of yeast was critical for achieving high conversion efficiency.

Furthermore, additional introduction of genes including bile acid–CoA:amino acid N-acyltransferase (BAAT) enabled the biosynthesis of the taurine-conjugated bile acid tauroursodeoxycholic acid (TUDCA) from UDCA. TUDCA production increased significantly when the taurine biosynthetic pathway was engineered within yeast, compared with external taurine supplementation. This improvement is likely due to limited membrane permeability of exogenous taurine (Moon and Smanski, 2025). Conjugated bile acids exhibit greater aqueous solubility than their unconjugated counterparts, resulting in improved bioavailability. They are also associated with reduced cytotoxicity and enhanced antioxidant and cytoprotective effects (Sabat et al., 2021). These findings suggest that yeast may serve as a more suitable production platform for downstream conversion and modification steps, particularly for the synthesis of secondary and conjugated bile acids from primary bile acids, rather than for the early stages of bile acid biosynthesis.

5 Conclusion and future perspectives: expanding the horizon with bile acids

5.1 Summary

Sterol metabolism entails a complex progression from a basic core structure to highly oxidized bile acids—a pathway entirely absent in native yeast. However, recent advances in synthetic biology have transformed yeast into a versatile platform for diverse sterol derivatives. Successful engineering efforts require more than simply expressing heterologous genes—they fundamentally depend on identifying and resolving structural and metabolic constraints within the yeast. Ultimately, these advanced engineering strategies unlock the scalable production of high-value therapeutics, laying a robust foundation for next-generation biomanufacturing and engineered microbial therapies.

5.2 Future perspectives

Although specific gut microbes clearly benefit host health, their exact molecular mechanisms remain poorly defined. Most studies highlight broad metabolomic shifts rather than demonstrating how distinct microbial metabolites regulate host pathways. Addressing this limitation requires engineered microorganisms that produce defined metabolites to causally drive host responses. As proposed here, bile acid-producing yeast serves as an ideal platform for this precise interrogation. For instance, treating gut organoids with these engineered yeasts allows researchers to directly track how defined bile acids reshape host gene expression, inflammatory responses, and epithelial barrier integrity through FXR signaling. This approach decisively shifts the focus from simple correlations to direct, mechanistic host–microbe interactions. In conclusion, engineering yeast to produce bile acids extends far beyond biomanufacturing, serving as a vital biological tool for decoding host physiology. By merging synthetic biology with mechanistic microbiome research, it seamlessly bridges bioprocess engineering and biomedical science, establishing a powerful technological foundation for next-generation microbe-based therapeutics.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The authors used ChatGPT (OpenAI, GPT-5.3 model) to assist with language editing and improving the clarity of the manuscript. All content was reviewed and verified by the authors, who take full responsibility for the accuracy and integrity of the manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fceng.2026.1833599/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Vasudeo Zambare, Balaji Enzyme and Chemical Pvt Ltd., India

Reviewed by: Xiunan Yi, University of California, Berkeley, United States

Lucero Romero-Aguilar, National Autonomous University of Mexico, Mexico