Recent Advances in Developing Aldehyde-Accumulating Microbes and Future Perspective of Biosynthetic Applications

In light of recent strides achieved in the fields of genome engineering and synthetic biology, the paradigm of de novo aldehyde biosynthesis in microorganisms emerges as a compelling alternative avenue when juxtaposed against traditional modalities, such as the customary physical (e.g., aldehyde extraction from botanical sources [21]) or chemical routes. Nonetheless, most microorganisms are unable to accumulate aldehydes naturally because aldehydes are readily converted to the corresponding alcohols or acids by the endogenous NADH-dependent alcohol dehydrogenase (ADH) family, NADPH-dependent aldo-keto reductase (AKR) family, or aldehyde dehydrogenase (ALDH) family. Therefore, the pivotal interventions necessitate the targeted elimination of ADHs, AKRs, and ALDHs, simultaneous with the orchestrated implementation of metabolic pathways dedicated to channeling the carbon fluxes in the direction of aldehyde biosynthesis. To facilitate a structured discussion, this review offers a preliminary classification of biogenic aldehydes into two primary groups based on molecular structure: aliphatic and aromatic aldehydes. A comprehensive compilation of representative examples of de novo aldehyde biosynthesis, as detailed in the subsequent sections, is concisely summarized in Table 1 for ease of reference. 2.1. Microbial Synthesis of Aliphatic Aldehydes and Their Derivatives Aliphatic aldehydes, produced through fermentation or fatty acid synthesis, have been strategically employed to synthesize alcohols [37,38] as well as alkanes [22,23], thereby catering to their utilities as both fuels and foundational chemical compounds. As aliphatic aldehydes are classified based on carbon chain length, they are categorized into short-chain (C ≤ 5), medium-chain (C6–C12), and long-chain (C > 12) fatty aldehydes. Extensive research has demonstrated the capacity of diverse engineered microorganisms, such as Escherichia coli, Pseudomonas aeruginosa, and Clostridium acetobutylicum, to synthesize these aldehydes, as summarized in Table 1. Notably, the microbial synthesis of short-chain fatty aldehydes has reached a comparatively advanced stage, with numerous studies reporting gram-scale product yields. Acetaldehyde, in particular, functions as an essential food additive and flavoring agent, and acts as a precursor for producing various commercial chemicals, including acetic anhydride, acetic acid, and butadiene.In 2002, Chaves et al. [24] reported that overexpressing the glyA gene (encoding serine hydroxymethyltransferase) showed an increase in threonine aldolase activity and acetaldehyde (~0.005 g/L) and folic acid formation when the engineered Streptococcus thermophilus were grown in LM17 medium. In 2005, efficient glucose to acetaldehyde (~0.94 g/L) conversion was achieved in Lactococcus lactis by nisibiotic-controlled overexpression of Zymomonas mobilis Pdc and L. lactis Nox [25]. In 2018, Balagurunathan et al. [26] introduced Pdc from Z. mobilis and NADH oxidase (Nox) from L. lactis into engineered E. coli. They knocked out genes related to competitive metabolic pathways (ΔadhE ΔldhA ΔfrdC ΔackA-pta), resulting in the production of acetaldehyde with a titer of 0.73 g/L from glucose. In 2019, Balodite et al. [27] overexpressed pyruvate decarboxylase (Pdc) from the cytosol into the periplasm, culminating in the synthesis of acetaldehyde at 1.66 g/L with glucose as the carbon source. Kallenieks et al. [28] used glucose as a carbon source to modify the motile fermentative bacterium, producing 4 g/L acetaldehyde by reducing the activity of ADHs and enhancing respiratory capacity (overexpression of respiratory NADH dehydrogenase (NDH) or mutations in other components metabolized by respiration). For the synthesis of butyraldehyde, in 1987, Rogers et al. [29] discovered that a spontaneous mutant of C. acetobutylicum NRRL B643 could produce 1.66 g/L butyraldehyde from glucose. In 2017, Ku et al. [30] overexpressed Coenzyme A-acylating ALDH from Clostridium beijerinckii in E. coli while knocking out the endogenous ethanol dehydrogenases YjgB, FucO, EutG, YbbO, AdhP, GldA, YahK, and YghA to synthesize 0.63 g/L butyraldehyde from glucose. Isobutyraldehyde, an isomer of butyraldehyde, was considered a promising diesel or jet fuel component. In 2009, Atsumi et al. [31] introduced the isobutyraldehyde synthesis pathway (ketoacid decarboxylase gene kivd from L. lactis, alsS from Bacillus subtilis, ilvC and ilvD from E. coli) into Synechococcus elongatus PCC7942. They overexpressed ribulose 1,5-diphosphocarboxylase/oxygenase (Rubisco) to improve productivity. The modified engineering strain can directly produce 1.1 g/L isobutyraldehyde from CO₂. Jazmin et al. [32] adjusted the carbon flux of the pyruvate kinase (PK) reaction step and its bypass pathway based on the study of engineering strains and improved the ability of engineering strains to produce isobutyraldehyde using CO₂ in small-scale experiments (0.05 g mmol/gDW/h). In E. coli, Rodriguez and Atsumi [33] identified and deleted 8 natural isobutyraldehyde reductases in the strain JCL260 that produces isobutanol. Combined with an in-situ product removal (ISPR) system, they achieved isobutyraldehyde production capacity of 35 g/L, which is the highest reported yield to date. However, due to incomplete knockout of the intracellularly related reductase, the strain still produced quite a lot of isobutanol after fermentation. In 2014, Rodriguez and Atsumi [39] reported that the simultaneous inactivation of 13 genes (adhE, yqhD, adhP, eutG, yiaY, ahr, betA, fucO, yahK, dkgA, gldA, ybbO, and yghA) in E. coli strain AL1728 led to a reduction of 90–99% in the strain’s endogenous aldehyde reductase activity for a variety of aliphatic aldehyde (C2 to C12). In comparison, fewer studies have focused on the production of medium-chain and long-chain fatty aldehydes using microorganisms. Santiago-Gómez et al. [40] achieved a yield of 600 mg/L hexanal by expressing hydrogen peroxide lyase (HPL) in green pepper fruit using Yarrowia lipolytica. Šalić et al. [41] employed NAD⁺ dependent ADH from Saccharomyces cerevisiae in a microreactor system to facilitate the oxidation of hexanol to hexanal. Akhtar et al. [37] demonstrated that carboxylic acid reductase (CAR) from Mycobacterium marinum is capable of converting various fatty acids (C6–C18) into their corresponding aldehydes. In 2020, Zhu et al. [42] realized an efficient synthesis of medium-chain fatty acids in S. cerevisiae employing multidimensional engineering techniques. Further introduction of CAR could potentially facilitate efficient synthesis of medium-chain fatty aldehydes. 2.2. Microbial Synthesis of Aromatic Aldehydes Aromatic aldehydes serve as essential precursors in the synthesis of fragrances, pharmaceuticals, agrochemicals, and polymers [43]. Owing to their high-value applications and substantial market demand, there has been a significant surge in commercial interest for microbial aldehyde synthesis in recent years. A compilation of recent advancements in the microbial accumulation of aromatic aldehydes is presented in Table 2. Among these, vanillin and benzaldehyde are particularly prominent, holding substantial market share in the flavor and fragrance industry. In 2014, Kunjapur et al. [44] demonstrated that knockout of ADHs (YqhD, YahK, and YjgB) and AKRs (DkgB, YeaE, DkgA) in E. coli (RARE strain) had a significant effect on improving the microbial accumulation of aldehydes, and achieved the accumulation of benzaldehyde and vanillin under specific conditions. After introducing the vanillin synthesis pathway, 119 ± 3 mg/L vanillin can be synthesized from glucose (more than a 55-fold increase over the wild-type strain). In 2022, Kim et al. [45] found that a single knockout of NCgl0324 in Corynebacterium glutamicum substantially enhanced the production of protocatechualdehyde (1.18 g/L) and vanillin (0.31 g/L). In 2023, Butler et al. [46] executed an extensive genetic manipulation study, systematically inactivating a suite of oxidoreductases (AldB, PuuC, BetB, PatD, FeaB, GabD, AldA, Sad, AstD, and PutA) within RARE strains. This strategic intervention led to the creation of ROAR strains, which, employed as whole-cell biocatalysts, exhibited a markedly improved capacity for accumulating eight aromatic aldehydes: piperonaldehyde, naphthaldehyde, anisaldehyde, m-hydroxybenzaldehyde, butyraldehyde, vanillin, benzaldehyde, and furfural. Particularly noteworthy, the retention rates of these aldehydes, assessed 4 hours following their introduction, revealed that six maintained levels above 50%. In 2024, Wu et al. [47] synthesized 481.2 mg/L vanillin in E. coli by optimizing the protocatechuic acid pathway, deleting three endogenous AKRs (encoded by DkgA, DkgB and YeaE) and three ADHs (YqhD, YahK and YjgB). Cinnamaldehyde, derived from the oil of cinnamon bark, is extensively utilized as a flavoring agent in a broad range of food and beverage products due to its substantial impact on enhancing taste and aroma. In 2023, Bang et al. [48] knocked out oxidoreductases (DkgA, DkgB, YahK, YeaE, YbbO, YjgB, YqhD, YqhC, GldA, and YqhA) in E. coli to achieve a higher level accumulation of cinnamaldehyde (3.8 g/L). However, E. coli and C. glutamicum are unsuitable for food-related fermentation applications, which severely limits their utility and potential uses. Compared to E. coli, the aldehyde accumulation platform in S. cerevisiae has been relatively slow to be studied in recent years. Table 2 presents an exemplary de novo vanillin biosynthesis from glucose via 3-dehydroshikimate in engineered yeasts, S. cerevisiae and Schizosaccharomyces pombe, facilitated by the introduction of four pivotal enzymes: 3-dehydroshikimate dehydratase, CAR, phosphopantetheinyl transferase, and catechol O-methyltransferase (COMT) [50]. However, endogenous enzyme activity posed a limitation on vanillin accumulation. Inactivation of ADH6 among 29 candidate ADHs, aryl-ADHs, and related aldehyde reductases (ALDRs) led to a 50% reduction in vanillin conversion to vanillyl alcohol, overcoming this obstacle [50]. More recently, our group deleted a total of 12 genes comprising Adh6, Adh7, Sfa1, Gre2, Hfd1, Gre3, Gcy1, Ydl124w, Ypr1, Ari1, Ydr541c, and Aad3 in S. cerevisiae, and the resulting minimal aromatic aldehyde reduction (MARE) yeast platform allowed improved accumulation of vanillin [51]. This collective effort of engineering S. cerevisiae produced an impressive 365.55 mg/L vanillin from glucose. This concerted genetic intervention yielded marked reductions in alcohol formation, concomitantly fostering the augmentation of vanillin accumulation, thereby reflecting a salient exemplification of the approach. 2.3. Microbial Synthesis of Aromatic Aldehyde Derived Complex Alkaloids The alkaloids are the largest class of natural products and include several of the most important human medicines [53]. Hence, an increasing number of studies are embracing biomass-derived aromatic aldehydes as feedstocks to expedite the biosynthesis of alkaloids (Table 3). (S)-Reticuline is a crucial intermediate for the production of alkaloids, which are applied in the manufacture of pharmaceuticals, green biological pesticides, and food additives [54]. The biosynthetic pathway of (S)-reticuline in plants starts with dopamine and 4-hydroxyphenylacetaldehyde (4-HPAA). 4-HPAA condenses with dopamine to form (S)-reticuline through enzyme-mediated Pictet-Spengler reaction, followed by 6OMT-catalyzed methylation of a 6-position hydroxyl group to form (S)-reticuline [55]. In 2020, Martin et al. [1] continued to implement >20 successive strain modifications to the yeast shikimate, Ehrlich, and L-tyrosine metabolic pathway based on the previous studies. The refinement of the shikimate pathway involved the amplification of branch acid synthase, prenylate dehydrogenase, and phenylpyruvate decarboxylase within the manganic acid pathway. Furthermore, the targeted elimination of genes responsible for seven related reductases and oxidases in the Ehrlich pathway was performed, effectively impeding the reduction and oxidative functions of 4-HPAA. Through feedback regulation of enzyme inhibition and optimization of the fermentation process, the yield of de novo synthesized (S)-reticuline by yeast increased 57,000-fold to 4.6 g/L [1]. The research results indicate that microbial synthesis of (S)-reticuline has reached the industrial production level of BIAs, highlighting the great potential of microbial systems for producing plant secondary metabolites. . A range of alkaloids, including berberine, scopolamine, tropine, vinblastine, and catharanthine, has been successfully demonstrated to be synthesized de novo in S. cerevisiae and Pichia pastoris (Table 3). Considering that yeast cells are more friendly to host diverse plant-derived metabolic pathways, it is like to explore further the MARE yeast platform [51] to synthesize other more complex natural alkaloids such as galanthamine (Figure 2). 2.4. Microbial Synthesis of Other Natural Aldehydes As depicted in Figure 1, a diverse array of aldehydes is utilized in the spice, pharmaceutical, and food industries. Given their potential for direct or indirect human exposure, the designation of “naturally derived” through microbial biosynthesis could expand the application spectrum of these compounds. To date, numerous aldehydes have been successfully produced via engineered microbial platforms, leveraging the capabilities of microbial chassis. With the ongoing refinement of microbial metabolic engineering, strain development, and evolutionary techniques, the de novo biosynthesis of aldehydes that are rare or possess intricate structures has become feasible. Retinal and its derivatives have been used in food, cosmetics, pharmaceuticals, nutritional supplements, and animal feed additives [62,63,64]. In 2022, we first reported the microbial production of retinal as the sole product [36]. Prior to this, there were reports of microbial synthesis of retinal, which resulted in a mixture of retinoids (including retinal, retinol, retinyl acetate, etc.) [65,66,67]. In our study [36], the co-inactivation of four ADHs (Adh6, Adh7, Sfa1, Gre2) and one ALDH (Hfd1) in S. cerevisiae enhanced the yeast’s capacity to accumulate retinal. By integrating crtE/YB/I genes from Xanthophyllomyces dendrorhous and β-carotene 15,15′-monooxygenase (BCMO) from the marine bacterium 66A03 into the yeast genome, we optimized β-carotene synthesis, yielding 69.13 ± 5.82 mg/L retinal with >99% purity in shake-flask cultures. Plants constitute a vital reservoir for natural aldehyde compounds. A case in point is safranal, the principal volatile component responsible for the distinct aroma of saffron. Beyond its characteristic spicy fragrance, safranal has also been recognized for its diverse pharmacological activities, including potential therapeutic benefits for anxiety, depression, and other neurological disorders [68]. Here, we predicted the possible synthesis modes of saffron aldehyde in microbial platforms (Figure 3). Zeaxanthin is the key compound in the synthesis pathways of safranal. In 2016, Shen et al. [69] introduced the dynamically regulated mevalonate (MVA) pathway from S. cerevisiae into E. coli, achieving a zeaxanthin titer of 722.46 mg/L via batch fermentation with feed supplementation. In 2023, Zhang et al. [70] performed metabolic engineering on Y. lipolytica, enabling the engineered strain to produce up to 775.3 mg/L of zeaxanthin in YPD shake flasks. The production of zeaxanthin, an intermediate in the safrole biosynthesis pathway, has been validated and optimized across diverse microbial hosts. Furthermore, prior studies have identified carotenoid cleavage dioxygenase 2 (CsCCD2) [71] for its role in catalyzing the conversion of zeaxanthin to hydroxy-β-cyclocitral and CsUGT709G1 [72] for catalyzing the formation of picrocrocin from hydroxy-β-cyclocitral in Crocus sativus. Integrating CsCCD2 and CsUGT709G1 into the zeaxanthin-producing microbial platform could facilitate the microbial synthesis of picrocrocin. Picrocrocin, the precursor of the volatile safranal, could be extracted from saffron stigmas and biotransformed to 4-hydroxysafranal with complete conversion (>99%) using β-glucosidase from Alicyclobacillus acidophilus, followed by a thermic treatment at 60 ℃. A one-pot 2-step chemo-enzymatic transformation is established to produce a marketable natural safranal [73]. To date, there are few higher plant genes to construct microbial cell factories for safranal accumulation. Thus, the introduction of higher plant genes into microorganisms to increase safranal preparation will be a significant trial. The results provide important insights into the engineering of safranal producer. Reconstitution of the safranal biosynthetic pathway in E. coli RARE and the yeast MARE platform might eventually give a sustainable production of safranal from economical fermentation processes.