Overcoming Barriers to Medium Chain Fatty Alcohol Production

Abstract

Medium chain fatty alcohols (^mcFaOHs) are aliphatic primary alcohols containing six to twelve carbons that are widely used in materials, pharmaceuticals, and cosmetics. Microbial biosynthesis has been touted as a route to less-abundant chain length molecules and as a sustainable alternative to current petrochemical-processes. Several metabolic engineering strategies for producing ^mcFaOHs have been demonstrated in the literature, yet processes continue to suffer from poor selectivity and ^mcFaOH toxicity, leading to reduced titers, rates, and yields of the desired compounds. This opinion examines the current state of microbial ^mcFaOH biosynthesis, summarizing engineering efforts to tailor selectivity and improve product tolerance by implementing engineering strategies that circumvent or overcome ^mcFaOH toxicity.

Graphical abstract

Introduction

Oleochemicals are a class of aliphatic molecules derived from plant oils, animal fats, and petrochemical feedstocks. In 2021, global revenue in the oleochemical market reached USD 33.1 billion, with growth projected to continue to USD 54.4 billion by 2029 [1]. The diverse chemical properties of different oleochemicals have led to applications in pharmaceuticals, cosmetics, fuels, agriculture, and others [2]. This opinion will focus on a subset of oleochemicals referred to as medium chain (i.e. C₆-C₁₂) fatty alcohols (^mcFaOHs). Compared to short- or long-chain FaOHs, the amphipathic, stability, and anti-foaming properties of ^mcFaOHs have popularized their use in many personal care products, lubricants, surfactants, and plasticizers (Fig. 1, bottom) [2–4]. The global ^mcFaOH market exceeded USD 0.9 billion in 2019 and is predicted to reach USD 1.3 billion by 2027 [5]. Given the growing market demand and limited natural supply, ^mcFaOHs and their derivatives are sold at 2–3 times the price of other oleochemicals such as fatty acid methyl-esters (FAMEs), free fatty acids, and fatty alcohols of other chain lengths [6].

Figure 1. Overview of FaOH Synthesis.

(Green/Top-left) Microbial fermentation offers the potential to generate FaOHs from sustainable feedstocks. Microorganisms turn carbon and energy sources into acetyl-CoA, ATP, and reducing power which are in turn used to synthesize acyl-thioesters via fatty acid biosynthesis (FAB) or reversed β-oxidation (RβOx). Acyl-thioesters are converted to fatty alcohols by enzymes described in Figure 2. (Grey/Top-right) Abiotic reactions can also convert non-sustainable feedstocks such as plant oils, animal fats, and fossil fuels into FaOHs. Fatty acids in lipid sources are released by hydrolysis or esterification to FAMEs and subsequently hydrogenated to produce FaOHs. FaOHs can also be made by oligomerization of ethylene and subsequent oxidation. This iterative synthesis generates a mixture of different chain-length products and requires downstream processing. (Orange/Bottom) FaOHs are converted to fatty alcohol sulfates, ethoxylates, esters and other compounds that are used in cosmetics, detergents, personal care products, agrochemicals, plasticizers, and lubricants.

Currently, ^mcFaOHs are either derived from natural lipids or synthesized from petrochemical building blocks (Fig. 1, grey). Plant oils can be processed into fatty alcohols, but the vast majority of harvested oils only contain long chain fatty acids (≥ C₁₆). On the other hand, chemical FaOH synthesis involves oligomerization of ethylene derived from petroleum or natural gas. The resulting longer chain olefins are subsequently oxidized into alcohols [2,4,7]. These iterative synthesis methods (e.g. Ziegler process [8]) are facilitated by aluminum-based catalysts [9] and generate a Poisson distribution of products ranging from C₂ to beyond C₂₆, requiring further separation [10]. The limitations of the above processes have spurred the pursuit of alternative approaches that are economically viable, sustainable, and capable of yielding highly selective ^mcFaOH products.

Biotechnology offers an alternative means of synthesizing ^mcFaOHs, one that could leverage renewable resources and potentially become sustainable (Fig. 1, green) [11]. Biochemically, fatty alcohols are made by reducing fatty aldehydes, compounds produced by reductive cleavage of acyl-thioesters. Different enzymes act on thioesters created by fatty acid biosynthesis (acyl-ACP), β-oxidation (acyl-CoA), or acyl-cysteine intermediates in the active site of a carboxylic acid reductase (CAR). Unlike short-chain FaOHs (e.g., ethanol, butanol), which are natively produced by many microbes as part of energy metabolism, ^mcFaOH are often used as intermediates in synthesizing biomass components and secondary metabolites. The small amounts of ^mcFaOH produced by wild microbes prevents their industrial use and necessitates engineering efforts to optimize pathways, strains, and fermentation strategies [12]. Metabolic pathways for using these enzymes to produce ^mcFaOHs have been demonstrated in Escherichia coli [3,13,14], yeast [15], cyanobacteria [16], and other non-model microbes [17–19]. Yet, industrial production of ^mcFaOHs via fermentation remains elusive [2,4,12,20]. Fatty alcohols can be toxic to microorganisms, reducing growth rates, cell viability, and alcohol productivity below economically viable levels [21]. Downstream separation of FaOH mixtures remains a significant process cost and therefore economically viable strategies must both narrow selectivity and elevate titer [2]. In this opinion, we evaluate current progress in optimizing ^mcFaOH selectivity, the current understanding of ^mcFaOH toxicity, and strategies to overcome it.

Increasing selectivity towards ^mcFaOH

To produce ^mcFaOHs, the host organisms must first accumulate acyl-thioesters or free fatty acids of desired acyl chain lengths. These compounds are made by two pathways: Fatty Acid Biosynthesis (FAB) or Reversed β-Oxidation (RβOx) pathway (Fig. 2). Each pathway functions with intermediates linked via thioester bonds to carriers, acyl-carrier proteins (ACP) in FAB and coenzyme A (CoA) in RβOx. Both pathways use an iterative elongation and reduction cycle that make arrays of different chain length acyl-thioesters. Termination enzymes then convert acyl-thioesters into lipids or other oleochemicals (Fig. 2, purple), including ^mcFaOHs [12,22,23]. FAB leverages a decarboxylative Claisen condensation to produce long chain C₁₆ and greater lipids for cell membrane synthesis (Fig. 2, yellow and blue) [4]. Activation of acetyl-CoA, at the expense of ATP, drives elongation of the acyl chain. Conversely, RβOx is a functionally reversed degradation pathway whose driving force comes from thermodynamically favorable trans-enoyl-CoA reductase (TER) reactions that use abundant NADH as a reductant (Fig. 2, pink) [12]. RβOx pathways use reversible β-ketoacyl-CoA thiolases (hereafter thiolases) to catalyze chain elongation directly with acetyl-CoA. This difference reduces the cost of each elongation cycle by one ATP, thereby increasing theoretical oleochemical yield at the expense of reduced thermodynamic driving force [4,24,25]. For example, calculations based on a modified E.coli genome-scale metabolic model showed that the theoretical anaerobic yield of octanol from glucose is 0.5 mol/mol via the RβOx pathway and 0.32 mol/mol via the FAB pathway [26]. The intense ATP requirement of FAB also necessitates aerobic cultivation, often resulting in significant carbon-loss through uncontrolled respiration or production of fermentation products [24]. Strategies for producing ^mcFaOHs via FAB or RβOx have been demonstrated at over 10 g/L and 0.2 g/g yield from glycerol in a rich media [13].

Figure 2. Metabolic pathways for fatty alcohol biosynthesis.

Both fatty acid biosynthesis (FAB) and reversed β-oxidation (RβOx) pathways are initiated by acetyl-CoA derived from central metabolism. (Yellow/Left) Malonyl-ACP, the extension unit of FAB, is generated via ATP-consuming reactions catalyzed by acetyl-CoA carboxylase (ACC) and malonyl-CoA:ACP transacylase (MCAT). (Blue/Middle) FAB begins with the condensation of acetyl-CoA and malonyl-ACP in a reaction catalyzed by β-ketoacyl-ACP synthase (KS). The resulting β-ketobutanoyl-ACP then undergoes keto-reduction, dehydration, and enoyl-reduction catalyzed by β-ketoacyl-ACP reductase (KR^ACP), β-hydroxyacyl-ACP dehydratase (DH^ACP), and enoyl-ACP reductase (ER^ACP), respectively. The acyl-ACP product can be further elongated following similar catalytic reactions with malonyl-ACP as additive molecules, or enter the termination module. (Pink/Middle) RβOx condenses acyl-CoAs with acetyl-CoA via a non-decarboxylative Claisen condensation catalyzed by a β-ketoacyl thiolase. The resulting β-ketoacyl-CoA, undergoes keto-reduction, dehydration, and enoyl-reduction catalyzed by β-ketoacyl-CoA reductase (KR^CoA), β-hydroxyacyl-CoA dehydratase (DH^CoA), and trans-enoyl-ACP reductase (TER). The resulting saturated acyl-CoA can either enter another elongation cycle or enter the termination module. (Purple/Top-Right) FaOH are made by reducing acyl-thioesters and the intermediate fatty aldehydes, catalyzed by distinct fatty-acyl-ACP/CoA reductase (FAR^ACP/CoA) and aldehyde reductase (ADR), or by single enzyme complex (dual FAR/ADR). The FAB and RβOx pathways can also be terminated using thioesterases (TE), generating free fatty acids (FFAs). FFAs can either be reactivated into acyl-CoAs by acyl-CoA synthetase (ACS) and follow the FAR-mediated route, or be converted into fatty aldehydes by carboxylic acid reductase (CAR) with the cost of one ATP. The aldehyde products are then reduced to FaOHs by ADR. (Bottom-Right) Enzymes involved in oleochemical synthesis can have narrow or broad chain-length selectivity. Asterisks indicate range of engineered enzymes. Dashed border represents potential activity range that are not reported previously. Abbreviations: FabB: KS from E.coli. FabF*: engineered KS from E.coli with limited activity above C8 [27]. FabG: KR^ACP from E.coli. FabZ: monofunctional DH^ACP from E.coli. FabA: dual DH^ACP/isomerase from E.coli. FabI, FabV: ER^ACP from E.coli or Pseudomonas aeruginosa [28]. ^CupTE*: engineered TE from Cuphea palustris [29]. ^ClFatB3*: engineered TE from Cuphea lanceolata [23]. ^UcFatB1: TE from Umbellularia californica [30]. ^MmCAR*: engineered CAR from Mycobacterium marinum [22]. AtoB, PaaJ: thiolases from E.coli [31,32]. Pct: thiolases from Megasphaera elsdenii [32]. PhaA: thiolases from Cupriavidus necator [33]. ThlA: thiolases from Clostridium kluyveri [34]. BktB: thiolases from C. necator [13]. Hbd: KR^CoA from C. kluyveri [34]. Crt: DH^CoA from Clostridium acetobutylicum [34]. FadBA and FadIJ: trifunctional enzyme complex with thiolase, KR^CoA, and DH^CoA activities, from E.coli [3].

While both FAB and RβOx can produce varied acyl-chain precursors, terminating these precursors to a desired distribution of fatty alcohols remains a challenge. ^mcFaOH product profiles can be controlled by either narrowing the distribution of acyl-thioesters or by altering the activity of terminating enzymes towards a particular acyl-chain length [23,27]. A few studies have explored highly specific elongating enzymes in FAB. For example, a structural and mutagenic study of the E. coli β-ketoacyl-ACP synthase (FabF) successfully limited its activity to synthesis of eight-carbon or shorter β-ketoacyl-ACPs and enhanced synthesis of octanoic acid [27]. Analogously, thiolases have been the most frequent target for controlling RβOx flux, possibly due to their diverse specificity profiles in nature [12,33]. ^EcAtoB, which primarily converts acetoacetate into four carbon acetoacetyl-CoA [35], was engineered to enable hexanol production [32]. The native activity of other thiolases such as PhaA [33], BktB [13] and PaaJ [31] are capped at six or eight carbon acyl-chain substrates. High-throughput screening (HTS) has accelerated the identification of novel thiolases with medium-chain selectivity. For example, a cell-free system was used to express and screen for thiolase homologs with preference toward C₄ or C₆ substrates [34]. Several RβOx elongating enzymes from Clostridium species were also found to be highly active on substrates up to C₁₀ chain-length when co-expressed with BktB [34]. Expressing BktB (with limited thiolase activity beyond C10 substrates) and an acyl-CoA reductase displaying maximal activity around C₁₂-C₁₄ substrates in E. coli led to a product distribution of >90% 1-decanol within the range of C₆-C₁₂ alcohols [13]. Successful engineering of elongation enzymes motivates additional mutagenesis campaigns targeting the selectivity of reductases and dehydratases in FAB and RβOx.

Compared to thiolases, most termination enzymes have activity on a wide range of chain-length substrates [3,12,36]. The exception to this rule comes from plants, where the distribution of fatty acids in oils is dictated in part by the specific activity of thioesterases (e.g. FatB1 from Umbellularia californica [30]). When expressed in microbes lacking active β-oxidation, thioesterases generate tailored pools of free fatty acids that can be converted to FaOHs [37]. The conversion proceeds either via CAR and an aldehyde reductase (ADR), or via an acyl-CoA synthase and fatty-acyl-CoA reductase (FAR^CoA), which could be engineering targets as well [4]. In one example, a structural-based mutagenesis study improved selectivity and activity on CAR from Mycobacterium marinum leading to 2.8-fold increase in ^mcFaOH production compared to the wildtype enzyme [22]. Bioprospecting and engineering campaigns have also created a collection of TEs with up to 75–90% medium-chain selectivity [23,29], which could be used in conjunction with other terminating enzymes to produce specific ^mcFaOHs (Fig. 2, bottom-right). Databases of terminating enzymes are now available [38] and machine learning is helping identify variants with desired activities [39].

To avoid the ATP cost of reactivating free fatty acids, FAR could be engineered to selectively reduce acyl-thioesters from mixed pools made by RβOx and FAB [3,12,40]. Unfortunately, the lack of crystal structures limits our understanding of how substrate binding is controlled [22,40]. Recent progress in machine learning (ML) and data science offers the potential to infer intrinsic relations between protein sequences and functions [41]. For instance, a ML-driven methodology was applied to enhance the activity of a fatty-acyl-ACP reductase (FAR^ACP) [40]. Nevertheless, advanced HTS technologies (e.g. MALDI-MS [23] and lipoic acid growth-based selection [14,29]) and in situ detection (e.g. FadR-mediated biosensors [42]) for ^mcFaOH are still necessary to efficiently characterize enzyme libraries with extreme medium-chain specificity. Besides shifting selectivity of individual enzymes, recent studies suggest that the relative expression ratio of enzymes may also exhibit substantial influence over the product profile in cyclic pathways. For example, by simply adjusting the ratio of the terminal ACP-thioesterase to elongation enzymes in the FAB pathway, up to 125-fold increase in medium-chain FFA could be induced in vivo [43]. A similar in vitro study on the RβOx pathway also illustrated that manipulation of enzyme ratios can lead to noteworthy shifts in product profile, transitioning from over 90% butanol and hexanol to over 80% dodecanol or longer [44].

Product toxicity as a major hurdle in scaling ^mcFaOH production

While progress in metabolic and protein engineering has increased microbial ^mcFaOH production, product toxicity remains a major bottleneck in scale up [49,50]. ^mcFaOHs are uniquely cytotoxic compared to their short and long chain counterparts, with lethal concentrations falling as low as 0.57 mM, for the case of nonanol for E. coli (Table 1) [21]. Quantification of ^mcFaOH toxicity has only been reported for E. coli, limiting knowledgeable host selection based on tolerance, especially considering that other popular organisms, such as S. cerevisiae, have comparatively higher tolerance to a wider range of molecules [51]. Despite the severity of their toxicity, our understanding of the specific mechanisms of ^mcFaOH inhibition is currently limited to its effects on cell membrane integrity. ^mcFaOH hydrophobicities span the range considered to be toxic to cells (1 > logP_{Octanol/Water} > 4), meaning they have an increased propensity to intercalate into the cell membrane and disrupt the hydrogen bonding between adjacent fatty acid tails (Table 1) [21,52–54]. This membrane disruption can potentially result in energy and cofactor leakage, altered membrane-bound protein function, and elicitation of intracellular stress responses (Fig. 3) [52,55]. Beyond the membrane, the negative effects that ^mcFaOHs may have on periplasmic and cytosolic components are poorly understood. Further, hydrophobic compound exposure is known to elicit a wide array of cellular responses, which makes identifying the most critical points of toxicity to target challenging [45,56].

Table 1.

Fuel, chemical, and toxicity properties of fatty alcohols

	Fuel and Chemical Properties					Toxicity Properties for E. coli
Alcohol Species (carbon no.)	Density (kg/m3 25°C)	Vapor Pressure (kPa25°C)	Solubility in Water (mg/L 25°C)	Cetane Number	ONMED	logPO/W	Lethal Concentration (mM)
Ethanol (C2)	790*	7.91	1.00×106	-	0.624	−0.310	1455.72
Butanol (C4)	810*	0.933	6.32×105	12.0	0.760	0.880	163.92
Hexanol (C6)	816	1.24×10−1	5900	23.3	0.820	2.03	23.90
Heptanol (C7)	822*	2.88×10−2	1670	29.5	0.835	2.62	7.01
Octanol (C8)	829	1.06×10−2	540	39.1	0.855	3.00	3.19
Nonanol (C9)	827*	3.03×10−3	140	46.2	0.867	3.77	0.57
Decanol (C10)	830	1.13×10−3	37.0	50.3	0.880	4.57	1.57
Undecanol (C11)	830*	3.96×10−4	19.0	53.2	0.886	4.72	n.r.
Dodecanol (C12)	830	1.13×10−4	4.00	63.6	0.898	5.13	n.r.
Tetradecanol (C14)	836*	1.47×10−5	0.300	80.8	0.900	6.03	non-toxic
Hexadecanol (C16)	815	8.00×10−7	insoluble	-	0.912	6.83	non-toxic

ONMED = Octane Normalized Mass Energy Density, calculated from [45] using heat of combustion values from NIST [46]. Density, vapor pressure, solubility, and logP values from PubChem [47]. Cetane numbers from [48]. Lethal concentrations from [21].

^*Value measured at 20 °C. n.r.=not reported

Figure 3. Strategies to improve ^mcFaOH tolerance.

Non-tolerant ^mcFaOH-producing strains are represented in blue and tolerant ^mcFaOH-producing strains are represented in green. Engineering strategies for conferring tolerance can be separated into those that isolate and/or remove ^mcFaOHs and those that confer cellular resistance to ^mcFaOHs. Toxin isolation/removal strategies include fermentation engineering, transporter engineering, and enzyme localization. (Column a) Fermentation engineering employs organic solvents and/or air stripping materials to partition ^mcFaOHs out of the aqueous media. (Column b) Transporter engineering applies protein engineering techniques to improve ^mcFaOH activity and/or ^mcFaOH selectivity of transporter proteins, with the goal of enhancing ^mcFaOH cellular export. (Column c) Enzyme localization uses signal peptides to target ^mcFaOH-producing pathway expression within organelles or proteinaceous microcompartments, isolating ^mcFaOH products and preventing potential disruption to other intracellular components, such as proteins and DNA. Toxin resistance strategies include membrane engineering and adaptive laboratory evolution (ALE). (Column d) Membrane engineering involves manipulation of fatty acid tails, phospholipid heads, and/or membrane hydrophobicity to diminish or prevent ^mcFaOH insertion, typically generating a more rigid membrane. (Column e) ALE is a whole-cell technique that applies ^mcFaOH pressure to select genotypes and phenotypes of cells that are adapted to the desired conditions of intracellular ^mcFaOH production.

To add to the complexity, we lack methodologies that focus on studying the effects of toxins that are endogenously produced, as is the case for ^mcFaOH strains. Often, toxicity studies are performed with exogenously fed toxins [21,57]. Cellular responses to endogenous toxins can be more severe than exogenous toxins, owing to differences in the area of the cell affected (i.e., inner versus outer membrane) [58] and how soluble a toxic product is in the cytosol versus in the media [52]. Designing and implementing techniques like in situ adaptive laboratory evolution for both endogenous and exogenous stresses may identify distinct strain phenotypes beneficial for ^mcFaOH -production [59].

Promising strategies for addressing ^mcFaOH product toxicity

To overcome the presence of toxic compounds (Fig. 3), cells must either isolate the toxin away from the points of inhibition or evolve resistance to the presence of the toxin. The solvent-like toxicity of ^mcFaOHs in E. coli, for example, suggests that product removal via fermentation or transporter engineering strategies may provide the best protection against ^mcFaOH toxicity. Ultimately, deciding which strategy or combination of strategies to implement depends on the specific method(s) of toxicity within host organisms. Below, we discuss examples of these strategies in more detail.

Transporter engineering

Swift export of ^mcFaOHs lessens toxicity and reduces product inhibition of terminal enzymes, potentially supporting ^mcFaOH pathway flux [53,60,61]. Transporters natively serve this role in cells and are a defining trait of many naturally solvent-tolerant species (Fig. 3b) [53,62]. Due to their promiscuous nature, many transporters originally annotated as having activity on other hydrophobic compounds (e.g. fatty acids, n-alkanes) have been shown to similarly act on fatty alcohols [19,63]. Proton motive force-dependent RND efflux pumps, specifically AcrAB-TolC, are of particular interest for the cellular export of ^mcFaOHs [64]. Deletion studies of AcrAB-TolC have demonstrated the necessity of all three subunits (inner membrane AcrB, periplasmic fusion AcrA, and outer membrane TolC), without which free fatty acid export and production are greatly impaired or nearly eliminated [65,66]. Clearly, complete toxin efflux from the cell is paramount to ensuring cell viability and oleochemical production, warranting further research into ^mcFaOH-specific transporters.

Enhancing product export is theoretically straightforward but has proven to be practically challenging. Databases linking export of desired compounds to protein candidates are sparsely populated and transporter overexpression can pose a large metabolic burden on the cell [67–69]. Further, it is unclear how overexpression of a single transporter impacts global membrane function. Membrane protein capacity, like global protein expression, is finite and any additions must be compensated by losses of other membrane protein activities (e.g. transport, ATP synthesis, electron transport flux) [69–71]. Additionally, transporter promiscuity may cause issues if it secretes or re-incorporates both ^mcFaOHs and related compounds (e.g. free fatty acids) used in their synthesis [72]. Therefore, deleting importers and engineering transporters with inherently better export activity and/or specificity towards ^mcFaOHs can bypass the need for overexpression and futile transport cycles across the membrane [70]. Relevant transporter engineering efforts have employed directed evolution to generate variants resulting in either improved tolerance or productivity [73–75]. However, the transporter variants in the aforementioned studies were selected under exogenous toxin conditions [73–75], meaning the mutations discovered may not necessarily benefit endogenous production.

Fermentation engineering

High extracellular concentrations of ^mcFaOHs following export can still have damaging effects on the cell and affect product yields in the same fashion as exogenous toxins [21,45]. In situ product removal (ISPR) strategies like bi-phasic fermentation and continuous product removal afford non-biological means of removing ^mcFaOHs from the local extracellular environment (Fig. 3a). Bi-phasic fermentations employ immiscible organic solvents, e.g., dodecane, to partition hydrophobic products away from the aqueous media [76–78] and have contributed to some of the highest-reported ^mcFaOH titers [13,14]. Alternatively, gas stripping has been used to remove volatile ^mcFaOH products from cultures but requires high volumes of gas that exacerbates foaming and complicates product recovery from dilute gas streams [77,78].

Enzyme localization

Analogous to export, toxins can be isolated from their points of impact by sequestration of products and/or pathways within eukaryotic organelles, lipid droplets, or proteinaceous bacterial microcompartments (BMC) (Fig. 3c) [79–81]. Enzyme localization to isolated compartments can also enhance pathway flux by increasing the local substrate concentration(s), promoting substrate channeling via enzyme proximity, and isolating substrates away from competing pathways [79,81]. However, compartmentalization can be limited by the rate of transport flux into the subcellular compartment, selectivity of which metabolites are compartmentalized, and whether proteins maintain function in the compartment environment [79,81]. Heterologous proteins can be targeted to organelles or BMCs by appending signaling sequences, a process that requires optimization to avoid loss of enzyme activity [82,83]. In yeasts, compartmentalization of FAR in the peroxisome, where β-oxidation provides the necessary fatty acyl thioester substrates, has proven to be advantageous for ^mcFaOH production compared to cytosolic pathway expression [82]. Alternatively, BMCs are an attractive compartmentalization strategy for prokaryotes because of their tunable size, configuration, and permeability [81,84]. However, successful demonstrations of heterologous pathway expression in BMCs beyond ethanol production have yet to be published [83].

Membrane engineering

Tolerance can also be conferred by restoring proper membrane function, as hydrophobic association of ^mcFaOHs with the membrane is a primary mode of toxicity (Fig. 3d) [52]. To counteract the fluidizing effects of ^mcFaOH intercalation, cells increase the ratio of straight chain saturated fatty acids to bent chain cis-unsaturated fatty acids present in the membrane [55,85]. The straight chain conformation of saturated fatty acids allows for tighter phospholipid packing and an overall more rigid membrane [55,85]. This adaptive response has inspired many methods of cell membrane engineering to increase hydrophobic compound tolerance. For example, heterologous expression of Pseudomonas putida’s cis-to-trans isomerase in E. coli has shown to improve both octanoic acid tolerance and productivity [86]. In another example, a system for tunable expression of key fatty acid synthesis enzymes allowed for tailored membrane composition that improved tolerance to multiple organic inhibitors [87]. Alternatively, membrane surface hydrophobicity can also be modulated to improve tolerance by altering the abundance and distribution of membrane-associated proteins and sugars [88,89]. In E. coli, modifying cell surface hydrophobicity through the expression of multiple-stress resistance proteins, or modifying the charge distribution of phospholipid heads has improved tolerance to octanoic acid but has not identified specific tolerance mechanisms [89,90]. However, direct engineering of the membrane compared to other strategies often only marginally improves toxin tolerance and production, possibly indicating that membrane engineering strategies either need to be improved, or that mechanisms beyond membrane damage play a larger role in ^mcFaOH toxicity [91,92].

Adaptive laboratory evolution

Adaptive laboratory evolution (ALE) simulates natural selection by driving a cell population to adapt to a particular stress, like ^mcFaOHs, and enriches for strains capable of faster growth (Fig. 3e) [93,94]. In conjunction with functional genomics and reverse engineering, ALE can improve our understanding of complex tolerance mechanisms, and potentially provide a platform for improved production. Through growth selection, ALE can filter either naturally evolved mutations or populations carrying sets of designed mutations (i.e., global transcription machinery engineering (gTME), mutagenized genome libraries, protein/enzyme variant libraries, CRISPR-Cas9- and homology-directed-repair-assisted genome-scale engineering (CHAnGE) etc.) [95,96]. In regard to oleochemical production, ALE has been successfully implemented to confer tolerance towards exogenous short, medium, and branched chain alcohols, as well as medium chain fatty acids [57,97–99]. However, traditional ALE only selects for growth in an exogenous stress and may not always benefit endogenous production [93]. The combination of endogenously produced ^mcFaOHs with new techniques for in situ ALE that couple growth to product accumulation may identify the mutations most advantageous for ^mcFaOH-producing strains [100].

Conclusions

Microbial fermentation offers a sustainable route to producing high value ^mcFaOHs with high selectivity [12,14,16], yet many barriers to commercial production remain [3,13]. Enzyme engineering has the potential to narrow the product portfolio to desired compounds and overcome poor specific activity. Pathway engineering strategies such as activity window overlapping [14] and enzyme ratio tuning [43,44] could also be useful in improving selective ^mcFaOH production. Product toxicity, however, limits the maximum titer of ^mcFaOHs [49,50]. While the modes of endogenous ^mcFaOH toxicity remain unclear, understanding both exogenous and endogenous stress may be key to improving ^mcFAOH tolerance [65,66]. Successful scale-up and commercialization of lab-scale engineering is heavily dependent on consistently achieving a high product titer, rate, and yield with exceptional specificity towards designated chain-length(s) [2]. A crucial step moving forward will be understanding the genotype, physiology, and functionality of ^mcFaOH-producing organisms in industrially relevant conditions.

Aside from the preceding efforts, we are also aware of hidden possibilities lying in other kingdoms of life. In nature, fatty alcohols are common components in insect pheromones and plant waxes, which might provide a huge reservoir of enzymes tailored for specific chain-length alcohol production [101]. In fact, bioprospecting enzymes in less-studied organisms have been proved feasible in very long chain fatty alcohol synthesis in yeasts [102]. Similar to how oleaginous yeasts serve as alternative cell factories for fatty acid production [103], non-model organisms exhibiting higher alcohol tolerance might be a promising route to alleviate ^mcFaOH toxicity. Several yeasts and thermophilic bacteria are also found to demonstrate significant tolerance over fatty alcohols [104,105]. Conveniently, FAB and/or βOx pathways are widely distributed across the kingdoms of life, meaning almost any organism can serve as a host for ^mcFaOH biosynthesis [106]. Ultimately, the ideal host would be able to efficiently generate necessary co-factors and supply high amounts of acetyl-CoA, while having inherent tolerance towards ^mcFaOHs.

Highlights.

• ^mcFaOHs are high value oleochemicals due to their use in industrial and consumer products.

• Substrate selectivity and product toxicity are major hurdles in microbial ^mcFaOH production.

• Highly selective thiolases and thioesterases serve as examples for future enzyme engineering.

• Limited understanding of ^mcFaOH toxicity mechanisms hinders tolerance engineering efforts.

• Alleviating ^mcFaOH toxicity relies on product removal and increasing cellular resistance.

Data Availability

No data were used for the research described in the article.