Metabolic engineering of Streptomyces albulus for de novo serotonin production

Abstract

Background

Serotonin (5-hydroxytryptamine, 5-HT) is a pivotal neurotransmitter with broad clinical relevance. However, its industrial-scale microbial production remains a significant challenge. In this study, we engineered Streptomyces albulus to develop an efficient enzymatic cascade for the scalable and high-yield biosynthesis of 5-HT.

Results

To enhance 5-HT production, we first overexpressed the native 5-hydroxytryptophan decarboxylase (5-OHTDC) in S. albulus, which strengthened the final decarboxylation step. Using 5-hydroxytryptophan as the substrate, this modification achieved a 5-HT titer of 6.87 g/L. We then introduced Escherichia coli L-tryptophan permease (Mtr) to improve cellular uptake of L-tryptophan, along with Actinomadura luzonensis tryptophan 5-hydroxylase (Luz15) to catalyze the hydroxylation of L-tryptophan. This triple-enzyme expression system (5-OHTDC/Mtr/Luz15) enabled the direct conversion of L-tryptophan to 5-HT, yielding 3.46 g/L. Leveraging the innate L-tryptophan biosynthesis capability of S. albulus, we pursued direct fermentative production from low-cost carbon sources. Mannitol was identified as the optimal substrate, supporting a maximum 5-HT titer of 6.54 g/L in fed-batch fermentation. Finally, implementing an L-tryptophan feeding strategy further boosted the final 5-HT titer to 12.0 g/L.

Conclusions

This work establishes S. albulus as an efficient microbial platform for high-level 5-HT production. The engineered strain and optimized fermentation process demonstrate strong potential for industrial application and provide a versatile chassis for synthesizing diverse tryptophan-derived bioactive compounds.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12934-026-02995-y.

Background

Serotonin, also known as 5-hydroxytryptamine (5-HT), was initially discovered in the serum and is widely distributed throughout mammalian tissues, with particularly high concentrations in brain tissue [1, 2]. It serves as a crucial molecule regulating neural activity and functions as an antidepressant with vasoconstrictive properties, while also regulating emotions and preventing brain aging [3].

5-HT is an intermediate metabolite produced during melatonin synthesis [4]. Melatonin is present in plants, animals, and microorganisms, and its metabolic pathways vary among organisms [5]. However, all pathways are based on tryptophan and involve a four-step reaction leading to final melatonin synthesis (Fig. 1).

Fig. 1.

Biosynthetic pathway of 5-HT and its derivatives in nature. TDC, L-tryptophan decarboxylase; T5H, tryptamin-5-hydroxylase; TPH, tryptophan hydroxylase; bTPH (Luz15), bacterial tryptophan hydroxylase; 5-OHTDC, 5-hydroxytryptophan decarboxylase; SNAT, serotonin N-acetyltransferase; ASMT/COMT, N-acetylserotonin methyltransferase

In plants, melatonin biosynthesis initiates in the cytoplasm with TDC-mediated decarboxylation of L-tryptophan to tryptamine, followed by T5H-catalyzed hydroxylation to 5-HT—a step exclusive to plants and dependent on cytochrome P450. Animals, lacking endogenous L-tryptophan synthesis, instead begin with TPH-catalyzed hydroxylation to 5-hydroxytryptophan (5-HTP; the BH₄- and O₂-dependent rate-limiting step), then decarboxylate it to 5-HT via 5-OHTDC. From 5-HT onward, both kingdoms converge: SNAT acetylates 5-HT to N-acetylserotonin (NAS; chloroplast-localized in plants), and ASMT subsequently methylates NAS to melatonin using S-adenosylmethionine (SAM). This conserved yet divergent architecture—decarboxylation before hydroxylation in plants versus the reverse order in animals—highlights a fundamental evolutionary adaptation in melatonin biosynthesis [6].

The microbial melatonin synthesis pathway is similar to that in animals; however, TPH was not discovered in microorganisms until 2023, when a study unveiled a novel tryptophan hydroxylase (bTPH) derived from bacteria [7]. However, the use of bTPH for the production of tryptophan derivatives is limited.

Currently, 5-HT production methods primarily include extraction from animals and plants and chemical synthesis. Animal and plant extraction methods are characterized by low efficiency and high costs, whereas the chemical synthesis of 5-HT involves a lengthy process, requires a variety of raw materials, and has drawbacks such as harsh reaction conditions and increased waste generation [1, 2, 8].

The microbial biosynthesis of 5-HT represents an environmentally sustainable approach and has thus attracted increasing research interest [9]. In one study, Park et al. demonstrated the production of 5-HT from tryptophan using Escherichia coli engineered to heterologously express T5H from rice and TDC from Catharanthus roseus. This system achieved extracellular and intracellular 5-HT levels of 24.0 mg/L and 4.00 mg/L, respectively [10]. Shen et al. subsequently reported the development of an E. coli platform for 5-HT synthesis, employing a multi-plasmid system to coexpress the 5-HT pathway along with a coenzyme regeneration module. Following IPTG induction, biotransformation was carried out using L-tryptophan and BH₄ as substrates, yielding a final 5-HT titer of 1.68 g/L [11]. More recently, Liu et al. implemented a combination of metabolic engineering strategies in E. coli to construct an optimized 5-HT biosynthetic pathway coupled with a cofactor regeneration system. Furthermore, they identified a putative 5-HT transporter protein. Using glucose as the sole carbon source, their engineered strain achieved de novo synthesis of 5-HT, reaching a final titer of 15.5 g/L [8].

Streptomyces albulus is the primary industrial strain for the production of ε-poly-L-lysine [12, 13], and previous studies have confirmed that this strain harbors high-activity 5-OHTDC [14]. However, its application in the biosynthesis of tryptophan derivatives has not been reported to date. Therefore, the objective of this study was to construct a tryptophan-derived biosynthetic pathway for 5-HT in S. albulus, with the subsequent aim of optimizing biotransformation conditions to achieve efficient 5-HT production using the genetically engineered strain.

Methods

Strains, plasmids, and culture conditions

The strains and plasmids used are listed in Table 1. The host bacterium, S. albulus CICC11022, was purchased from the China Center of Industrial Culture Collection (CICC). The vectors pSET152 [15] and pTHS [16], both having genome-integration capabilities, were used for gene expression. The S. albulus strains were cultured at 30 °C following a previously reported method [17]. Meanwhile, E. coli strains were aerobically cultured at 37 °C in Luria–Bertani medium. The culture media and components used in this study are listed in Table S1. When necessary, antibiotics were used at the following concentrations: 80 µg/mL apramycin, 50 µg/mL hygromycin B, 25 µg/mL chloramphenicol, 50 µg/mL kanamycin, and 25 µg/mL nalidixic acid.

Table 1.

Strains and plasmids used in this study

Strains and plasmids	Relevant genotypea	Source or reference
Strains
Streptomyces albulus CICC 11,022	Control strain, transformation host	CICC
S. albulus Q-152	Control strain, S. albulus CICC 11,022 harboring pSET152	Lab collection
S. albulus Q-OHTDC	5-OHTDC expression strain, S. albulus CICC 11,022 harboring pSET152-OHTDC	This study
S. albulus Q-OHTDC-Luz15	5-OHTDC and Luz15 co-expression strain, S. albulus CICC 11,022 harboring pSET152-OHTDC and pTHS-Luz15
S. albulus Q-OHTDC-Mtr-Luz15	5-OHTDC, Mtr, and Luz15 co-expression strain, S. albulus CICC 11,022 harboring pSET152-OHTDC and pTHS-Mtr-Luz15	This study
Escherichia coli ET12567/pUZ8002	recE, dcm−, dam−, hsdS, Cmr, Tetr, Strr, Kmr, non-methylating plasmid donor strain for intergeneric conjugation	[18]
E. coli Trans	General cloning strain	Rui Biotech
Plasmids
pSET152	5.7 kb, Aprr, integrative plasmid, lacZα oripUC19 oriTRP4 int-attPφ31 aac (3) IV	[15]
pSET152-pls	Aprr, pSET152 carrying SP43 promoter, SR40 RBS, and pls gene	[12]
pSET152-OHTDC	Aprr, pSET152 carrying SP43 promoter, SR40 RBS, and 5-OHTDC-encoding gene	This study
pTHS	7.6 kb, Hygr, integrative plasmid, int-TG-1	[16]
pTHS-Luz15	Hygr, pTHS carrying Sp43 promoter, SR40 RBS, and Luz15-encoding gene	This study
pTHS-Mtr-Luz15	Hygr, pTHS carrying Sp43 promoter, SR40 RBS, and Mtr and Luz15-encoding genes	This study

Strain construction

To construct the 5-OHTDC-overexpression plasmid, the 5-ohtdc gene was amplified from the genomic DNA of S. albulus CICC11022 using 2 × Phanta Max Master Mix (Vazyme, China). The polymerase chain reaction (PCR) mixture consisted of 4 µL of genomic DNA template, 4 µL each primer (OHTDC-F and OHTDC-R) (Table S2), 50 µL of 2× Phanta Max Master Mix, and 38 µL of ddH₂O. The PCR program followed a two-step amplification method: denaturation at 95 °C for 15 s, followed by annealing and extension at 72 °C for 90 s for 35 cycles. The PCR product containing the 5-ohtdc gene was then ligated to the pSET152-Sp43-SR40 fragment obtained through digestion of pSET152-pls with the restriction enzymes NdeI and EcoRI. The ligation product was subsequently transformed into E. coli DH5α. Transformants exhibiting apramycin resistance were selected and validated using a comprehensive approach involving PCR, enzyme digestion, and DNA sequencing. The resulting plasmid was designated pSET152-OHTDC.

To generate the heterologous expression plasmid for the tryptophan 2,3-dioxygenase (Luz15) and the co-expression plasmid for Luz15 and tryptophan: H(+) symporter (Mtr), the codon-optimized luz15 gene (ACCESSION No.: UKU09928.1) from Actinomadura luzonensis DSM43766 and mtr gene (Gene ID: 947675) from E. coli MG1655, along with the SP43 promoter and SR41 RBS, were chemically synthesized and inserted into pTHS. The resulting expression plasmids were designated pTHS-Luz15 and pTHS-Mtr-Luz15. A schematic representation of the plasmids used in this study is shown in Fig. S1.

The obtained plasmids were initially transfected into E. coli ET12567/pUZ8002. Subsequently, plasmid pSET152-OHTDC was introduced into S. albulus CICC11022 via conjugation transfer [19]. The resulting strain carrying pSET152-OHTDC was designated S. albulus Q-OHTDC. Following this step, plasmid pTHS-Luz15 and pTHS-Mtr-Luz15 was introduced into S. albulus Q-OHTDC via conjugation transfer. The resulting strains carrying pSET152-OHTDC and pTHS-Luz15 were designated S. albulus Q-OHTDC-Luz15, and the strains carrying pSET152-OHTDC and pTHS-Mtr-Luz15 were designated S. albulus Q-OHTDC-Mtr-Luz15.

Quantitative real-time PCR (qRT-PCR)

The gene expression levels of 5-ohtdc, mtr, and luz15 in S. albulus strains were compared via qRT-PCR. The primers and their corresponding sequences are shown in Table S2. All primer names begin with the prefix “RT.” All samples were collected after 36 h of cultivation in the M3G medium [19]. qRT-PCR was performed as described previously [19].

Production of 5-HT by whole cell biotransformation

S. albulus strains were spread onto MS solid medium and incubated at 30 °C for 5–6 days to collect spores. A 1 mL spore suspension was inoculated into a 300 mL conical flask containing 50 mL of M3G medium at 30 °C and shaken at 220 rpm for 24 h to obtain the seed culture liquid. The seed culture liquid was then transferred to fresh M3G medium in shaking flasks or fermenters at a 10% inoculation volume to prepare the cells for transformation.

For shaking flask transformation, the seed culture liquid was expanded at 30 °C and shaken at 220 rpm for 48 h. The cells were collected for transformation. Unless otherwise specified, the optical density at 600 nm (OD₆₀₀) of the cells was adjusted to 20 using transformation medium prior to transformation. The transformation media used included M3G, M9Y, M9, Gause’s medium No.1, and Gause’s medium No.2; the specific formulations are listed in Table S1. The substrate concentrations of 5-HTP and tryptophan were maintained at 5 g/L. Transformation was conducted at 30 °C with shaking at 180 rpm. Product generation and substrate consumption were evaluated at 72 h post-transformation.

When cultivating strains in fermenters, the temperature was maintained at 30 °C and the ventilation rate was set to 2 vvm. In the early stage, pH control was not implemented, and ammonia water was used to maintain a pH of 5.0 when it dropped below this value. The initial rotation speed was set to 300 rpm and coupled with dissolved oxygen to regulate the levels at 30%. After approximately 36 h of cultivation, the cells were collected for whole-cell transformation using M9Y medium supplemented with the appropriate substrates and other components. Lysozyme was used to partially disrupt the cell wall of S. albulus, as the thick peptidoglycan layer of Streptomyces spp. can limit the transport of substrates into the cell and products out of the cell, thereby reducing biotransformation efficiency. The transformation liquid pH was adjusted according to experimental requirements, while maintaining a temperature of 30 °C and dissolved oxygen level of 30%. Samples were collected every 12 h for substrate and product analyses. The optimal conditions obtained from each optimization experiment were used for subsequent optimizations in this section.

De novo production of 5-HT by fermentation

Spores and seed cultures of S. albulus strains were prepared as described above. The fermentation medium was based on M3G medium with the carbon source replaced. Flask fermentations were carried out at 30 °C with shaking at 220 rpm, using an inoculum size of 10%. After 72 h of fermentation, the concentrations of 5-HT and NAS in the broth were determined.

Fed-batch fermentation was performed in a bioreactor using M3M medium (Table S1). The temperature was maintained at 30 °C, with an aeration rate of 2 vvm and dissolved oxygen level controlled at 30%. For 5-HT production, the pH was maintained at 6.0. When the mannitol concentration in the broth fell below 10 g/L, a concentrated mannitol solution was added to increase its concentration by 20 g/L. For L-tryptophan feeding, 5 g/L of L-tryptophan was added each time, and subsequent additions were made only after its concentration dropped below 1 g/L. Substrate and product concentrations in the fermentation broth were monitored at regular intervals throughout the process.

Analytical method

Cell growth was measured at 600 nm using a spectrophotometer [17]. The quantification of tryptophan, 5-HTP, and 5-HT was performed utilizing a Wooking K2025 high-performance liquid chromatograph equipped with a Kromasil 100-5-C18 column (4.6 × 250 mm, 5 μm) maintained at a column temperature of 25 °C. A diode array detector set at a detection wavelength of 276 nm was used for analysis. The mobile phase consisted of methanol/potassium phosphate buffer (10 mM, pH 6.5). To detect tryptophan, 5-HTP, and 5-HT, the ratio of methanol to phosphate buffer was adjusted to 12:88. For sample preparation, 1 mL fermentation broth collected at different time intervals was centrifuged at 13 824 RCF for 10 min. The supernatant was harvested, diluted appropriately with distilled water, and filtered through a 0.45 μm filter membrane before HPLC analysis. Standard solutions of tryptophan, 5‑HTP, and 5‑HT were prepared to establish calibration curves, which were then used to quantify the concentrations of the three compounds in the samples.

5-HT was analyzed via liquid chromatography-mass spectrometry (LC-MS) using an Agilent 1260 Infinity II Series LC System equipped with a single quadrupole mass spectrometer (Agilent 6120 B). An eluent comprising 20% acetonitrile and 80% water (with 0.1% formic acid) was used for isocratic separation for up to 4 min, followed by a change in the flow gradient ratio for 10 min (80% methanol and 20% water). A reversed-phase separation column (Phenomenex Synergi 4 µ Hydro-RP C18 column, dimensions: 250 × 4.6 mm, particle size: 4 μm) was utilized at a flow rate of 1 mL min^− 1 and maintained at a column temperature of T = 30 °C. The injection volume was set at 0.6 µL with a total sequence time of 10 min. Mass spectra were acquired using atmospheric pressure ionization electrospray mass spectroscopy in both positive and negative modes within a scan mass range of 50–960 m/z. The collision-induced dissociation voltage was set at 70 V, while the drying gas (N₂) flow rate was maintained at 12 L min^− 1 with the drying gas temperature set to 350 °C and nebulizer pressure adjusted to 35 psig.

Experimental replication and data presentation

All experiments (including whole cell biotransformation and de novo fermentation) were performed with three biological replicates (independent cultures) and two technical replicates (duplicate measurements for each biological sample). For shake-flask experiments, data are presented as the mean value; for fermenter experiments, a representative result from three independent trials is shown.GraphPad Prism 8.3 (GraphPad Software, USA) was utilized for data visualization and statistical analysis. Significant differences were determined by one-way analysis of variance (ANOVA) followed by Dunnett’s multiple comparisons test.

Results

Strain construction and verification

Microbial 5-HT synthesis has emerged as a green and sustainable alternative, garnering increasing attention [1, 2]. However, despite studies exploring the biosynthesis of these compounds using microorganisms such as E. coli, the yields and efficiency remain suboptimal. Notably, current research relies exclusively on plasmid-based systems for the heterologous expression of relevant genes, resulting in poor stability. Furthermore, biotransformation processes require the addition of inducers, cofactors (e.g., BH₄), and antibiotics, which complicate the workflow and increase production costs [11, 14]. Thus, the development of an efficient and stable 5-HT biosynthetic method has significant application potential.

In this study, we engineered an S. albulus strain to express the 5-HT biosynthetic pathway components, as illustrated in Fig. 2. Two genome-integrated vectors, pSET152 and pTHS, were used to express the endogenous and exogenous genes, respectively. The successful expression of 5-OHTDC, Luz15, and Mtr in S. albulus was confirmed via qRT-PCR analysis (Fig. S2), which enabled the quantification of their expression levels. Engineered strains with either individual gene expression or co-expression could convert multiple substrates into 5-HT. Subsequent experiments were conducted to optimize the synthesis conditions for each strain using different substrates to enhance 5-HT production.

Fig. 2.

The biosynthesis pathway (A) and strains (B) constructed in this study. Mtr, L-tryptophan-specific permease from E. coli MG1655; Luz15, tryptophan 5-hydroxylase from A. luzonensis DSM43766; 5-OHTDC, 5-hydroxytryptophan decarboxylase from S. albulus CICC11022; SNAT, 5-HT N-acetyltransferase from S. albulus CICC11022

5-HT production from 5-HTP by whole cell biotransformation

The abilities of the 5-OHTDC-overexpression S. albulus strain, Q-OHTDC, and the control S. albulus strain, Q-152, to convert 5-HTP into 5-HT in M9 medium were compared. As depicted in Fig. 3A, the titers were 2.13 g/L for Q-OHTDC and 0.56 g/L for Q-152. Subsequently, the conditions for this transformation using Q-OHTDC were optimized; specifically, these included an optimal OD value of cells at 20 (Fig. 3B), a temperature of 30 °C (Fig. 3C), and M9Y medium as the optimal transformation medium (Fig. 3D). Using M9Y medium, an initial pH of 6.0 was optimal (Fig. 3E), the addition 15 µg/mL lysozyme yielded optimum results (Fig. 3F), and 0.5 mM PLP resulted in maximum efficiency (Fig. 3G). In the shaking flasks transformation, a 60 h transformation period resulted in the consumption of 3.74 g/L of 5-HTP, leading to a maximum 5-HT titer of 3.74 g/L with an almost complete conversion rate (Fig. 3H). Similarly, in the fermenters, a 72 h transformation led to the consumption of 6.98 g/L of 5-HTP and achieved a maximum 5-HT titer of 6.87 g/L with a conversion rate of approximately 98.4% (Fig. 3I). The presence of 5-HT in the conversion solution was verified through LC-MS (Fig. S3).

Fig. 3.

5-HT production from 5-HTP by whole cells of S. albulus Q-OHTDC. A comparison of 5-HT production capacity between the 5-OHTDC-overexpressing and control strains; B effects of cell concentrations on 5-HT production; C effects of transformation temperature on 5-HT production; D effects of transformation medium on 5-HT production; E effects of transformation pH on 5-HT production; F effects of lysozyme on 5-HT production; G effects of pyridoxal phosphate (PLP) on 5-HT production; H curve showing 5-HT production from 5-HTP in shaking flasks over time; I curve showing 5-HT production from 5-HTP in fermenters over time. ns, no significance; *, P < 0.05; **, P < 0.01; ****, P < 0.0001

These findings indicated that 5-HTP can traverse the cell wall and membrane of S. albulus to enter the intracellular space. Lysozyme treatment partially degraded the cell wall and increased cell membrane permeability, thereby facilitating 5‑HTP uptake. Additionally, 5-OHTDC exhibited inherent activity in the wild-type strains, and its overexpression significantly enhanced the ability of the strain to convert 5-HTP into 5-HT. Notably, PLP serves as a key cofactor for 5‑OHTDC, and supplementation with an appropriate amount of PLP was beneficial for 5‑HT biosynthesis.

5-HT production from L-tryptophan by whole cell biotransformation

The S. albulus 5-OHTDC-overexpression strain, Q-OHTDC, facilitated the biosynthesis of 5-HT from 5-HTP. Subsequently, introduction of the tryptophan 2,3-dioxygenase Luz15 from A. luzonensis DSM43766 into strain Q-OHTDC resulted in the co-expression strain Q-OHTDC-Luz15; however, this strain was unable to synthesize 5-HT from tryptophan (Fig. 4A). The concentration of tryptophan in the culture medium remained unchanged throughout the transformation process, indicating that intracellular tryptophan transport may play a critical role in the overall metabolic pathway. Therefore, a new strain, Q-OHTDC-Mtr-Luz15, expressing 5-OHTDC, Luz15, and Mtr, was constructed. Fortunately, this strain successfully synthesized 5-HT from tryptophan (Fig. 4A). Through optimization experiments, the optimal transformation temperature was determined to be 30 °C (Fig. 5B), with an initial transformation medium pH value of 6.0 (Fig. 4C). The maximum concentration of 5-HT achieved through shaking-flask fermentation was 1.52 g/L with a cell density of 20 OD (Fig. 4D). Maintaining pH control at 6.0 during the fermenter operation mode resulted in a maximum 5-HT titer of up to 3.46 g/L with a conversion rate of 79% from tryptophan (Fig. 4E). In contrast, controlling the pH at 6.5 only yielded a maximum 5-HT titer of 2.31 g/L with a significantly lower conversion rate of 51% (Fig. 4F). Lysozyme and PLP were not supplemented in this section because no beneficial effect was observed upon their addition.

Fig. 4.

5-HT production from tryptophan by whole cells of S. albulus Q-OHTDC-Mtr-Luz15. A comparison of 5-HT production capacity between the 5-OHTDC, Mtr, and Luz15 co-overexpressing and control strains; B effects of transformation temperature on 5-HT production; C effects of transformation pH on 5-HT production; D curve showing 5-HT production from 5-HTP in shaking flasks over time; E curve showing 5-HT production from 5-HTP at pH 6.0 in fermenters over time. F curve showing 5-HT production from 5-HTP at pH 6.5 in fermenters over time. **, P < 0.01; ****, P < 0.0001

Fig. 5.

De novo production of 5-HT by S. albulus Q-OHTDC-Mtr-Luz15. A effects of carbon sources on 5-HT production; B curve showing de novo production of 5-HT from mannitol in fermenters over time; C effects of L-tryptophan addition in M3M medium on 5-HT production; D curve showing de novo production of 5-HT from mannitol and L-tryptophan in fermenters over time. ****, P < 0.0001

De novo production of 5-HT from mannitol by fermentation

Genome analysis of S. albulus CICC11022 revealed that this strain harbors a complete L-tryptophan biosynthetic pathway. Thus, an attempt was made to achieve de novo production of 5-HT via direct fermentation using S. albulus Q-OHTDC-Mtr-Luz15. Initially, the effect of different carbon sources on 5-HT yield was investigated in shake flasks. The results demonstrated that the highest 5-HT titer (1.13 g/L) was obtained when mannitol was used as the sole carbon source, which was significantly higher than that achieved with glucose (0.44 g/L) or glycerol (0.72 g/L) as the sole carbon source, as well as any combination of two mixed carbon sources (Fig. 5A).

Subsequently, the initial concentration of mannitol was optimized, and 50 g/L was determined as the optimal concentration. Fermentation was further scaled up in a fermenter using S. albulus Q-OHTDC-Mtr-Luz15 and M3M medium with mannitol as the sole carbon source. The pH and dissolved oxygen (DO) were maintained at 6.0 and 30%, respectively, and mannitol was intermittently fed during the fermentation process. After 108 h of fermentation, the maximum 5-HT titer reached 6.54 g/L (Fig. 5B).

Fed-batch production of 5-HT from mannitol and L-tryptophan

Given that S. albulus Q-OHTDC-Mtr-Luz15 possesses the dual capacity to convert L-tryptophan into 5-HT and to synthesize 5-HT de novo using mannitol as the substrate, two-substrate fermentation was performed. First, the effect of L-tryptophan supplementation in M3M medium on 5-HT production was investigated in shake flasks. The results indicated that the maximum 5-HT yield reached 2.41 g/L when L-tryptophan was added at a concentration of 4 g/L, representing a 159.1% increase compared to the control condition (0.93 g/L) (Fig. 5C).

Subsequently, fed-batch fermentation was conducted in a fermenter using M3M medium supplemented with 4 g/L L-tryptophan. During the fermentation process, both mannitol and L-tryptophan were fed intermittently. After 108 h of fermentation, the maximum 5-HT yield attained 12.0 g/L (Fig. 5D).

Discussion

To date, in research on 5-HT or melatonin production using genetically engineered strains, the tryptophan hydroxylation step has primarily relied on TPH of animal origin [1, 6, 20]. To enhance the conversion efficiency, it is often necessary to establish an intracellular regeneration system for the cofactor BH₄ [8, 11, 14, 21]. In contrast, bacterially derived bTPH (Luz15) utilizes histidine-linked heme cofactors in conjunction with molecular oxygen or hydrogen peroxide to selectively catalyze region-specific hydroxylation of the indole moiety of tryptophan, which distinguishes it mechanistically from animal-derived TPH [7]. To our knowledge, we successfully achieved the heterologous expression of bTPH in S. albulus for the first time. We also observed that S. albulus exhibited insufficient L-tryptophan transport capacity. Accordingly, Mtr from E. coli MG1655 was heterologously expressed in S. albulus. Fortunately, the co-expression of Mtr and Luz15 enabled the intracellular transport and hydroxylation of extracellular L-tryptophan, thereby providing a solid foundation for the efficient production of 5-HT.

TDCs from plants and 5‑OHTDCs from mammals have been widely used to convert 5-HTP to 5-HT [4, 11, 22]. However, TDC is a bifunctional enzyme that catalyzes the decarboxylation of both 5‑HTP to 5‑HT and tryptophan to tryptamine, leading to the formation of tryptamine as a byproduct [6]. 5‑OHTDC is mainly distributed in animal cells, with only a few bacterial sources reported to date. Moreover, mammalian 5‑OHTDCs usually show low catalytic activity when expressed in bacterial systems [1]. Notably, the S. albulus-derived 5-OHTDC used here overcomes the above limitations, demonstrating both high activity and strict specificity for 5-HTP.

Table 2 summarizes the major research advances in the biosynthesis of 5-HT in recent years. Prior to 2025, the highest reported titer of 5-HT was only 1.68 g/L [11]. To address this limitation, Liu et al. constructed a high-yield 5-HT engineering strain using E. coli as the chassis. Through multi-dimensional strain modification strategies, including biosynthetic pathway construction and optimization, precursor regulation, metabolic flux balance, cofactor enhancement, and transport capacity improvement, they finally obtained the high-yield strain WW15 [8]. This strain achieved a 5-HT titer of 15.5 g/L via two-stage fed-batch fermentation [8], which currently represents the highest level of 5-HT biosynthesis reported in the literature.

Table 2.

Overview of 5-HT production using biosynthesis methods

Strain	Substrate	Titer (g/L)	Reference
E. coli / pET28b-TDC	5-HTP	0.035	[23]
Saccharomyces cerevisiae INVSc1 / pYES-TDC	5-HTP	0.040	[23]
E. coli BL21 (DE3) / pCOLADuet-GSTΔ37T5H + TDC	tryptophan	0.024	[10]
E. coli BL21 (DE3) ΔtnaA / pCOLAJ23-TDC	5-HTP	0.154	[22]
In vitro, purified enzymes from E. coli NEBExpress	tryptophan	0.087	[24]
E. coli BL21(DE3)△tnaA/BH4/HaDDC-SmTPH	tryptophan	0.414	[25]
E. coli SE24	tryptophan	1.68	[11]
E. coli WW15	glucose	15.5	[8]
S. albulus Q-OHTDC-Mtr-Luz15	mannitol and tryptophan	12.0	This study

Although the 5-HT titer obtained in our work is slightly lower (12.0 g/L), our research possesses a distinct innovation: it is the first report of 5-HT production using Streptomyces as the host strain. S. albulus was chosen for two major advantages: (i) the endogenous 5‑OHTDC of this strain has been confirmed to exhibit high activity [14], which provides a solid foundation for this study; (ii) it is the dominant industrial strain for ε‑poly‑L‑lysine production and displays excellent growth performance under industrial fermentation conditions [12, 13], thus laying a groundwork for the potential large‑scale production of 5‑HT.

The high-efficiency 5-HT production in S. albulus can be attributed to several key factors. First, the overexpression of native 5-OHTDC and the precise heterologous expression of Mtr effectively promoted the biosynthesis of 5-HT. Notably, the heterologous introduction of Luz15—a bacterial BH₄-independent TPH—successfully overcame the rate-limiting step in tryptophan hydroxylation [7]. Second, the systematic optimization of culture conditions and fermentation parameters further contributed to the achievement of high 5-HT yields in our study. We found that mannitol is a superior carbon source compared to glucose. The reduced 5-HT yield observed when glucose is used as the carbon source is likely due to glucose transport mediated by the phosphotransferase system (PTS). This system consumes phosphoenolpyruvate (PEP), a critical precursor for the biosynthesis of 2-dehydro-3-deoxy-D-arabino-heptonate 7-phosphate (DAHP). As the first committed intermediate linking glycolysis and the L-tryptophan biosynthetic pathway, the reduced DAHP production resulting from PEP depletion consequently limits L-tryptophan availability, thereby impairing 5-HT biosynthesis [26]. Notably, mannitol transport in Streptomyces species generally proceeds independently of the PTS [27, 28]. In the genome of S. albulus CICC11022, no PTS-related mannitol transporters were identified; instead, we detected the smoEFGK gene cluster, which encodes an ATP-binding cassette (ABC) transporter associated with mannitol transport. This ABC transporter-mediated mannitol uptake may facilitate the redirection of metabolic flux toward L-tryptophan biosynthesis, providing an advantage over glucose-based carbon metabolism for 5-HT production.

Despite these advances, several aspects remain for further optimization. First, the de novo L‑tryptophan biosynthetic pathway has not been systematically optimized. Future work can focus on regulating key enzymes such as DAHP synthase, anthranilate synthase, and phosphoenolpyruvate synthase to enhance precursor flux [20]. Meanwhile, competing branch‑pathway enzymes including prephenate dehydrogenase (TyrA), tyrosine transaminase (TyrB), prephenate dehydratase (PheA), and phenylalanine transaminase (PheB) should be knocked out or attenuated to reduce flux loss [1]. Second, the expression and activity of core 5-HT biosynthetic enzymes (5-OHTDC, Luz15, Mtr) have not been fully optimized. Promoter engineering can be used to tune transcription efficiency [29], and protein engineering can improve catalytic activity, substrate affinity, and stability [30]. Third, intracellular cofactor balance is critical for efficient 5-HT biosynthesis but remains unstudied. 5-OHTDC requires PLP (vitamin B6 derivative) [2], while Luz15 depends on heme, Fe²⁺, and molecular oxygen [7]. Enhancing expression of key enzymes in PLP and heme synthesis pathways could elevate cofactor levels. Finally, the lack of efficient gene editing systems for S. albulus greatly limits iterative metabolic engineering. Developing tailored precise editing tools is therefore a key prerequisite for advancing S. albulus as a high‑efficiency 5-HT chassis.

Conclusions

This study constructed an integrated biosynthetic pathway for 5-HT in S. albulus via metabolic engineering. By sequentially implementing a series of optimization strategies—overexpression of 5-OHTDC, heterologous expression of Mtr and Luz15, as well as development of de novo synthesis and fed-batch fermentation processes—a maximum 5-HT titer of 12.0 g/L was achieved. This work establishes S. albulus as an efficient microbial platform for high-level 5-HT production and provide a versatile chassis for synthesizing diverse tryptophan-derived bioactive compounds.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

5-HT

5-Hydroxytryptamine (Serotonin)

5-HTP

5-Hydroxytryptophan

5-OHTDC

5-Hydroxytryptophan decarboxylase

ABC

ATP-binding cassette (transporter)

ASMT

NAS methyltransferase

Apr^r

Apramycin resistance

BH₄

Tetrahydrobiopterin

bTPH

Bacterial tryptophan hydroxylase

CICC

China Center of Industrial Culture Collection

Cm^r

Chloramphenicol resistance

CRISPR-Cas

Clustered Regularly Interspaced Short Palindromic Repeats-CRISPR associated proteins

DAHP

2-Dehydro-3-deoxy-D-arabino-heptonate 7-phosphate

DO

Dissolved oxygen

E. coli

Escherichia coli

IPTG

Isopropyl β-D-1-thiogalactopyranoside

Km^r

Kanamycin resistance

LC-MS

Liquid chromatography-mass spectrometry

Luz15

Tryptophan 5-hydroxylase from Actinomadura luzonensis DSM43766

Mtr

L-tryptophan permease (from Escherichia coli MG1655)

NAS

N-Acetylserotonin

OD₆₀₀

Optical density at 600 nm

PEP

Phosphoenolpyruvate

PCR

Polymerase chain reaction

PLP

Pyridoxal phosphate

PTS

Phosphotransferase system

qRT-PCR

Quantitative real-time PCR

RBS

Ribosome binding site

SAM

S-Adenosylmethionine

SAH

S-Adenosylhomocysteine

SNAT

5-HT N-acetyltransferase

Str^r

Streptomycin resistance

T5H

Tryptamin-5-hydroxylase

TDC

L-Tryptophan decarboxylase

Tet^r

Tetracycline resistance

TPH

Tryptophan hydroxylase

Hyg^r

Hygromycin B resistance

λ-Red

Lambda Red recombination system

ANOVA

One-way analysis of variance

Author contributions

JQ and XW designed the research. YZ and MZ constructed the strains. YZ, AL and MZ performed the biotransformation and batch fermentations. JS and QW cultivated the strains. YZ and MZ wrote the draft manuscript. YL, QY and LK analyzed the data. JQ wrote the article and acquired funding. All the authors read, revised and approved the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (32370094).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.