Available strategies for improving the biosynthesis of surfactin: a review

Abstract Surfactin is an excellent biosurfactant with a wide range of application prospects in many industrial fields. However, its low productivity and high cost have largely limited its commercial applications. In this review, the pathways for surfactin synthesis in Bacillus strains are summarized and discussed. Further, the latest strategies for improving surfactin production, including: medium optimization, genome engineering methods (rational genetic engineering, genome reduction, and genome shuffling), heterologous synthesis, and the use of synthetic biology combined with metabolic engineering approaches to construct high-quality artificial cells for surfactin production using xylose, are described. Finally, the prospects for improving surfactin synthesis are discussed in detail.


Introduction
Surfactin was first identified as a secondary metabolite in the culture supernatant of Bacillus subtilis in 1968 [1], and it remains one of the most effective biosurfactants based on its ability to reduce surface tension.It is mainly produced by Gram-positive Bacillus spp.[2][3][4][5].Surfactin has a lipopeptide structure consisting of: two acidic amino acid residues (L-glutamic acid and L-aspartic acid), five non-polar amino acid residues (L-leucine, D-leucine, L-valine, D-leucine, and L-leucine), and a 12-19-carbon b-hydroxy fatty acid chain [6].It is synthesized by a non-ribosomal peptide synthase (NRPS) via a multi-carrier sulfur template mechanism [7,8].Surfactin has received extensive attention because of its: excellent surface activity, antimicrobial properties, outstanding biodegradability, and a variety of notable biological activities [9][10][11].Surfactin has high market value and broad application prospects in various industrial fields, such as: oil extraction, bioremediation, pharmaceuticals, agriculture, food, and household chemicals [12,13].However, because of the high production costs and low yields of known production strains, the industrial manufacture and commercial applications of surfactin remain elusive [14].
To meet the increasing demand for surfactin, considerable efforts have been made in recent years.Many studies focused on enhancing surfactin biosynthesis through various methods, including traditional approaches such as optimizing the nutrient supply and culture conditions of microbial sources, through which the: yield, composition, and biological activity of surfactin have been greatly improved [15][16][17].However, many factors of the surfactin biosynthesis mechanism are unknown, hampering further improvement and industrialization.Therefore, a comprehensive understanding of this mechanism at the systems biotechnology level, especially analyses of the pathways and enzymes involved in surfactin synthesis, is crucial for achieving the commercial application of surfactin.
This paper reviews the important enzymes of the three biosynthetic pathways of: surfactin, branchedchain amino acids, and branched-chain fatty acids.We further review the production of surfactin to illustrate the current research progress regarding its efficient synthesis based on different substrates, thereby facilitating a more in-depth understanding of the biological and molecular mechanisms of surfactin biosynthesis.synthesis of surfactin are presented in Supplementary Tables 1-4.
Surfactin is biosynthesized by the surfactin synthase SrfA, which is regulated by quorum sensing signals and the phosphopantetheine transferase Sfp.Numerous studies have aimed to improve the activity of SrfA and promote the production of surfactin [8].SrfA is composed of four subunits, namely: SrfAA, SrfAB, SrfAC, and SrfAD [18,19].SrfAA and SrfAB assemble the first six amino acids, and the extension of the initial product is catalyzed by a set of simultaneous thioester bond cleavage and transpeptidation reactions.SrfAC controls the assembly of the seventh amino acid, and it includes the first thioesterase domain that terminates the peptide chain extension and releases the resulting lipopeptide intermediate.SrfAD contains a second thioesterase/acyltransferase domain that activates the peptide chain.In addition, the NRPS multi-enzyme complex domains (including the: adenylation, peptide carrier protein, condensation, and thioesterase domains) are important for the biosynthesis of surfactin and enhancement of enzymatic activity [20].NRPS can be modified to synthesize novel surfactin analogs, and the engineering of NRPS has become a focus of research in related fields [21].Chooi et al. [22] indicated that the initial condensation domain of SrfAA is important for the structural modification of fatty acyl.However, no modification of the further condensation domain has been reported.Many reported engineering attempts resulted in chimeric enzymes with low product yields or even a complete loss of activity.The complexity of the NRPS structure, integrity of the modules, and diversity of link regions complicate genetic manipulation.Numerous unexpected modifications and low yields of peptide products need to be overcome.
In addition to SrfA, amino acid metabolism and fatty acid metabolism are also important modules for surfactin synthesis [7,[23][24][25].Among them, the branched-chain amino acids: L-isoleucine, L-valine, and L-leucine are the main precursors of branched-chain fatty acids [26][27][28], which are converted into the corresponding fatty acyl-ACPs [7].These fatty acyl-ACPs then enter the prolonged cycle of fatty acid biosynthesis.Next, the corresponding fatty acid is hydroxylated by the cytochrome P450 enzyme YbdT to form 3-hydroxy fatty acid [29,30].The final CoA-activated long-chain fatty acid produced by acyl CoA ligases (LcfA, LcfB, YhfL, YhfT, and YngI) is considered the substrate for initiating surfactin synthesis [25].This results in the transfer and condensation of fatty acyl AMP intermediates to the PpaN cofactor at the Nterminus of the peptide carrier protein domain in SrfAA [31].However, the fatty acid component for surfactin production can be generated through other pathways, such as sulfur transfer from ACP to CoA.Additionally, the mid-chain acyl carrier protein thioesterase Bte can increase the amount of free fatty acids [32].
Furthermore, the key enzymes of central carbon metabolism (glycolysis, pentose phosphate pathway, and TCA cycle) play important roles in amino acid and fatty acid metabolism, which directly or indirectly provide more energy and crucial substrates for enhancing surfactin biosynthesis.For example, glucose can generate ATP in the glycolysis pathway to synthesize surfactin using raw material.The pentose phosphate pathway can provide NADPH for fatty acid biosynthesis, and ribose 5phosphate is produced for the synthesis of nucleotide coenzymes and nucleotides.Finally, the TCA cycle contains key nodes that connect it to secondary metabolic branches, such as methylmalonyl-CoA [33].Consequently, increasing the expression of enzymes that generate methylmalonyl-CoA may also enhance surfactin synthesis.

Enhancement of surfactin synthesis under different culture conditions
Strategies for improving the biosynthesis of surfactin are presented in Figure 2. Surfactin productivity is related to cell growth in Bacillus spp.Various culture conditions (e.g.pH, temperature, vibration speed) have been described in detail in literature to affect surfactin production by influencing cell growth [3,34].In addition, optimization of substances added to the medium (carbon sources, precursors, and other promoters) is one of the most common and effective strategies for promoting surfactin production (Table 1).

Addition of glucose as a substrate
Many studies aimed to improve the medium composition whereby surfactin production could be promoted via supplementation with different or more concentrated carbon sources [60].A large number of studies reported that glucose is the preferred carbon source for surfactin production by B. subtilis [35][36][37][61][62][63], and optimization of the glucose concentration and the use of glucose, in combination with other substances, have been attempted to improve surfactin production [38][39][40].Hmidet et al. [38] revealed that a medium containing glucose (30 g/L) and glutamic acid (5 g/L) had a significant effect on surfactin production.Yeh et al. [40] found that surfactin is most effectively produced in the presence of 40 g/L glucose, whereas higher concentrations (50-60 g/L) may lead to the accumulation of excess by-products, resulting in a low pH and a corresponding decrease of surfactin production.Willenbacher et al. [41] increased surfactin production by analyzing the influence of different glucose concentrations (8-40 g/L) and improving the composition of Cooper's medium.Their results differed from those of Ghribi et al. [39] and Yeh et al. [40].However, surfactin production can fluctuate greatly during the culture process, which may explain these discrepancies.Hoffmann et al. [64] investigated the effect of glucose concentration on the non-foaming anaerobic surfactin production process of B. subtilis 2JABs24.The effect of glucose was weaker than that of ammonium.However, at lower glucose concentrations, acetate production was not reduced.As reported by Espinosa-de-los-Monteros et al. [65], the glucose/nitrate ratio can affect the synthesis of metabolic by-products such as acetate and lactic acid.Hu et al. [42] reported that glucose can be a good carbon source for surfactin production by B. subtilis BSFX022.However, glucose fermentation in a non-buffered system will result in decreased pH because of the accumulation of acidic by-products, thereby inhibiting surfactin production.The metabolic by-products produced by glucose metabolism during surfactin production require further research.Additionally, the carbon fluxes of glucose degradation represent another interesting topic for future research [64].These studies illustrated that many challenges remain in the production of surfactin from glucose.

Addition of xylose mixture as substrate
The production of surfactin is characterized by a high substrate cost, and effective strategies for reducing the production cost have always represented a key focus in the field of surfactin production.Surfactin production is usually based on sucrose, glucose, or glycerol as the carbon source [66].Agricultural waste biomass is one of the most abundant and inexpensive renewable resources, among which xylose, as the main product of lignocellulose hydrolysis, is the most abundant sugar in nature after glucose.The development of cheaper carbon sources can reduce costs, avoid the unnecessary generation of environmentally harmful waste components, and result in a sustainable industrial process.
It was recently reported that xylose-rich mixed substrates such as lignocellulosic biomass can be used as substitutes for refined sugars such as glucose [67][68][69], whereas lignocellulose biomass is widely used in fermentation processes [70].Notably, B. subtilis intrinsically expresses xylose isomerase and xylulokinase, granting it the ability to metabolize xylose, indicating that B. subtilis is a powerful candidate for surfactin production from lignocellulosic biomass [71].Examples of surfactin synthesis using added xylose mixtures are listed in Table 1.Prado et al. [43] proposed that hemicellulose-containing corncob hydrolyzate produced by alkaline pretreatment can be used as an added xylose mixture for surfactin production.Chen et al. [44] produced surfactin from corncob hydrolysate (xylose mixtures) and monosodium glutamate wastewater (MGW) in B. velezensis BS-37, achieving a surfactin yield of 0.523 g/L.Hu et al. [42] used waste biomasses such as corncob hydrolysate (xylose mixtures) and MGW to produce 2.032 g/L surfactin in an optimal waste matrix.Jarrell et al. [45] used cellulose, xylose, and xylan, as well as their combinations, as carbon sources to produce surfactin in B. subtilis, and the tested sugars yielded similar concentrations of surfactin.These results demonstrated that the use of enriched xylose added mixtures can improve surfactin production.

Addition of xylose as a single substrate
At present, there are few reports on surfactin production by Bacillus strains using xylose as a single carbon source.Khan et al. [46] studied surfactin production via submerged fermentation in a xylose medium, and the highest yield of surfactin reached 2.7 g/L, representing a 2.85-fold increase compared to no xylose addition.Bartal et al. [47] found that in xylose addition increased surfactin production in the C 15 variant by more than 11% compared to the findings for no xylose addition.

Addition of hydrocarbons
When hydrocarbons are added to the culture medium, many production strains will increase biosurfactant production to emulsify and degrade the hydrocarbons as a source of nutrients [34].Cheap petroleum hydrocarbons have also been used as carbon sources in the production of surfactin [34,55,72].Among them, kerosene and diesel have proven more suitable for surfactin production than other hydrocarbons [39,55].Ghribi et al. [39] reported that supplementing the optimized medium with 2% kerosene and diesel significantly increased the production of surfactin in B. subtilis SPB1, with the outputs being 1.74 and 1.5 g/L, respectively, representing 1.4-and 1.08-fold increases versus the control (no kerosene or diesel), respectively.

Addition of amino acids
Relevant studies demonstrated that surfactin production can be further increased by the addition of exogenous amino acids [50].Transcriptome analysis of strains with increased surfactin production revealed that the transcription levels of genes involved in L-leu- cine synthesis were related to the limitation of surfactin production, corresponding to the results of transcriptome analysis of the surfactin producer B. amyloliquefaciens MT45 [73] and the addition of L-leucine to the medium further increased surfactin production (up to 16.7 g/L), in line with previous findings, indicating that surfactin production was increased by 30-fold in the presence of exogenous L-leucine [74].Similar results were reported for the high surfactin-producing strain B. velezensis BS-37 [51].The surfactin yield of B. velezensis BS-37 was increased by 2-fold in the presence of 10 mM L-leucine.In addition, the addition of amino acids to the medium had a notable effect on the ratio of surfactin variants [52].The addition of L-valine or L-isoleucine as the nitrogen source can selectively increase the yield of [Val 7 ]-surfactin [53].These results demonstrated that the addition of different amino acids significantly affected surfactin production, providing a reference for improving the production of specific surfactin variants and clarifying the relationship between the culture medium and the corresponding products of specific bacteria.

Addition of metal ions
In addition to metabolic precursors, a large number of studies successfully increased surfactin production by B. subtilis through the addition of metal ions (Mg 2þ , K þ , Mn 2þ , and Fe 2þ ) [56,57].Micronutrients and metal ions can act as major cofactors of multi-enzymes to affect surfactin biosynthesis [35,56,57,75,76].Wei et al. [17,56] found that adding Fe 2þ to the culture medium significantly increased cell counts and surfactin production.In another study, Wei et al. [57] revealed that adding 0.01 mM Mn 2þ to the culture medium significantly increased the surfactin yield from 0.33 to 2.6 g/L.In addition, their studies illustrated that Mg 2þ acts as a cofactor of the B. subtilis Sfp protein, which contributes to surfactin production, and K þ could stimulate the secretion of surfactin.Sheppard et al. [58] found that the addition of Mn 2þ shortened the period of continuous growth of B. subtilis and improved the surfactin productivity in batch culture.Huang et al. [59] revealed that Mn 2þ actively promoted surfactin biosynthesis in B. subtilis ATCC 21332.Their results demonstrated that surfactin production gradually improved with increasing Mn 2þ supplementation (0.001-0.1 mmol/L), and the maximal yield of 1.5 g/L was 6.2-fold higher than that in medium without added Mn 2þ .In addition, the amount of Mn 2þ added was positively correlated with nitrate utilization, which then stimulated secondary metabolic activities and surfactin synthesis.Moreover, Mn 2þ enhanced the activity of glutamate synthase, thereby increasing the absorption and conversion of nitrogen and providing more free amino acids for surfactin synthesis.

Addition of nanoparticles to the culture medium
In addition to nutrients and buffers, there has been extensive research on the use of nanoparticles composed of iron, silver, or iron oxide to increase and improve the production of biosurfactants [77,78].
Recently, Fe nanoparticles (FeNPs) have been extensively applied because of their superior magnetic properties and perfect biocompatibility [79].Modabber et al. [48] studied the effects of starch-coated Fe 3þ nanoparticles on the biomass and surfactin production of B. subtilis.In a 5-L bioreactor, medium containing Fe 3þ nanoparticles resulted in greater biomass and surfactin production (6.02 g/L) as well as correspondingly lower surface tension (10.30mN/m), indicating that Fe 3þ nanoparticles have a positive effect on biosurfactant production.Yang et al. [49] used FeNPs to increase the production of surfactin by enhancing its secretion.The addition of 5 g/L FeNPs to shake flasks and 7-L bioreactors increased the yield of surfactin to 7.15 and 9.18 g/L, respectively, representing increases of 45 and 54.5%, respectively.Meanwhile, transcriptomic analysis confirmed the effect of FeNPs on gene expression in B. amyloliquefaciens MT45.These works provided an effective strategy to increase surfactin production that can be applied to other metabolites, especially biosurfactants.The results illustrated that FeNPs have a positive effect on surfactin production not both shake flasks and large-scale applications.This will promote the industrial application of FeNPs in surfactin production.

Enhancement of surfactin biosynthesis by various genome engineering technologies
The genome engineering strategies used to improve surfactin strains include: rational genetic engineering, genome reduction, and genome shuffling approaches.Among them, rational genetic engineering, including: promoter engineering, enhanced amino acid and fatty acid precursor supply, enhanced surfactin efflux, and enhanced surfactin synthase expression, improve surfactin production by increasing the metabolic efficiency and robustness of the producing strain.Genome reduction can create a simplified editable unit to improve production strains by targeting the streamlined construction of stable organisms with reduced or core genomes.However, these methods require in-depth knowledge of the genetic background of the target strain and the necessary genetic tools, which limits the widespread use of rational methods.Therefore, the use of genome shuffling, which has been considered a novel genome-wide engineering approach for rapidly improving cell phenotypes and which is more conducive to creating and improving the complex phenotypes of wild-type cells through the diversity of genomic variants, to improve surfactin-producing strains will be discussed.

Enhancement of surfactin biosynthesis by rational genetic engineering
Because of the low yield of wild-type B. subtilis, commercial surfactin production has not been achieved.
Recent studies attempted to optimize the fermentation process to improve the yield of surfactin [15,80].However, these efforts failed to achieve commercially viable and profitable surfactin production.With the development of genetic engineering, increasing attention has been paid to the rational engineering of strains to increase surfactin production and create new structures.The combination of genome sequencing and global transcriptome analysis is an effective strategy for revealing the characteristics of surfactin biosynthesis and regulation in high-yield strains.An increasing number of powerful gene expression regulation tools can further facilitate the development of synthetic biology and metabolic engineering, including: novel promoters, riboswitches, and CRISPRa/I [81][82][83][84].These approaches can provide important information for rational strain improvement [73].At present, studies on genetic engineering for improving surfactin production have mainly focused on three aspects, namely: enhancing precursor fatty acid and amino acid supply, enhancing efflux/ resistance systems, and increasing surfactin synthase expression (Table 2).

Enhancing the supply of precursors
Recent studies demonstrated that overexpressing genes involved in fatty acid biosynthesis pathways to increase the supply of fatty acid precursors can greatly enhance surfactin production [8].By overexpressing the global transcription regulator CodY [85] and YbdT [30] to provide 3-hydroxy fatty acids, the composition of surfactin can be changed, and the final product titer can be improved.Wang et al. [86] overexpressed biotin carboxylase II encoded by yngH to enhance acetyl CoA carboxylase activity and improve surfactin synthesis.The production of surfactin was increased to 13.37 g/L, representing a 43% increase over that of the control strain B. subtilis TS1726.In addition to fatty acids, amino acids are also important precursors for surfactin biosynthesis.Therefore, increasing the supply of amino acids is beneficial to increasing the yield of surfactin, which can be achieved by reducing the metabolic flux of competing amino acid biosynthesis pathways.Wang et al. [87] used CRISPRi technology to inhibit the metabolism of: L-glutamic acid (yrpC, racE, or murC), L-leucine, and L-valine (bkdAA and bkdAB, two genes involved in the utilization and subsequent conversion of L-leucine and L-valine to branched-chain fatty acids), leading to 2.5to 627-fold reductions in the transcription levels, respectively, which in turn increased the production of surfactin and the proportion of C 14 subtypes.According to Wang et al. [54], the increased synthesis of branched-chain fatty acids, L-glutamic acid, and L-aspartic acid could explain the increase in surfactin production in B. subtilis TS1726DspoIVB.Similarly, transcriptome analysis revealed that most genes in the fatty acid synthesis and branched-chain amino acid pathways were significantly upregulated in B. amyloliquefaciens MT45 [73].

Enhancing the expression of surfactin export proteins
Enhancing the expression of surfactin export proteins is also an effective method for increasing surfactin production.Tsuge et al. [98] indicated that the production and resistance of surfactin can be mediated by protein transporters.Li et al. [88] further confirmed the hypothesis that proton motive force (PMF)-dependent transmembrane export can promote the efflux of surfactin in B. subtilis.For example, three putative lipopeptide transporters (KrsE, YcxA, and YerP) dependent on the PMF were overexpressed, and the production of surfactin was significantly increased by 52, 89, and 145%, respectively [7].The liaIHGFSR operon related to daptomycin resistance is also considered to participate in surfactin resistance.Zhi et al. [8] simultaneously overexpressed the SwrC, AcrB, and LiaIHGFSR proteins, which significantly increased the production of surfactin to 3.8 g/L, representing a 1.2-fold increase versus that of the original strain B. subtilis 168S7.These findings are consistent with the high-level expression of these proteins in the high surfactin-producing strain B.
amyloliquefaciens MT45 [73], which confirmed their effects on the efflux and auto-toxicity of surfactin, indicating that secretion is important for surfactin production.However, these characteristics need to be further confirmed given their great significance to the construction of high production strains.Although genetic manipulation effectively enhances metabolite secretion, it is usually difficult to perform in wild-type strains because of the uncertainty regarding many genetic components.Therefore, it is pressing to develop strategies to increase the production in wild-type strains and enhance the secretion of surfactin.Proteomics is a good method to analyze the differences between key proteins related to efflux in different host strains with high surfactin productivity.However, the exact mechanism of surfactin efflux and the proteins involved in transmembrane transport remain to be studied further in the future.

Increasing the expression of srfA to enhance surfactin biosynthesis
The typical factors affecting the expression of srfA are presented in Figure 3. Surfactin biosynthesis requires the key gene srfA, which is a 27-kb large operon controlled by the P srf promoter.Because it is difficult to express srfA heterologously, promoter exchange is an effective and preferred approach to increase surfactin production.For example, B. subtilis was designed to increase the production of surfactin mainly through promoter exchange of the srfA operon [89,90].To establish a high-yield surfactin production strain through promoter engineering, the transcriptome and production capacity of the strain need to be analyzed first.Jiao et al. [14] confirmed the weakness of the natural P srfA promoter through transcriptome analysis.Subsequently, the super-strong chimeric promoter P g3 was developed to drive the synthesis of surfactin.The production of surfactin in shake flasks reached 8.61 g/L, representing a 15.6-fold increase versus the wild-type strain B. subtilis THY-7.Cheng et al. [91] found that the promoter P 10 was stronger than P srfA concerning the expression of GFP and that P 10 could be used for autoinducible overexpression systems of heterologous proteins.However, it is necessary to explore effective promoters to enhance surfactin production, as the targeted construction of SrfA expression promoters for each strain may be more suitable.The number of uncertainties regarding the genetic components of wild-type strains greatly increases the difficulty of promoter engineering.
Expression of the srfA operon is also controlled by transcriptional regulatory factors.Two classical factors, namely ComX and competence and sporulation factor (CSF), mediate the quorum sensing system of B. subtilis and influence spore formation.ComX and ComP/ComA form a two-component system that regulates surfactin production.Phosphorylation (ComA-P) stimulates the activity of the transcription factor ComA through two independent pathways, thereby activating transcription of the srfA operon [99,100].ComQXP, SodA, and DegQ positively regulated ComA expression [92].Jung et al. [93] overexpressed the two signal factors encoded by comX and phrC in B. subtilis to increase the production of surfactin and stimulate transcription of the srfA operon.The results illustrated that the genetically modified B. subtilis exhibited 6.4-fold higher surfactin production than the wild-type strain B. subtilis (pHT43).Several typical Rap proteins have been found to be "anti-activators" of ComA [94].Deletion of Rap protein can increase the expression of ComA-P [95], which in turn increases surfactin production.In addition, spore formation affects the distribution of nutrients in the cell [101], thereby influencing the synthesis of secondary metabolites such as surfactin.Wang et al. [54] revealed that the non-spore-producing engineered strain spoIVB-null may be more suitable for surfactin synthesis.
In addition to ComX and CSF, the expression of the srfA operon is also regulated by several global regulator proteins [102].The activating proteins ComK and DegU, as well as the inhibitory proteins: CodY, SpX, AbrB, Rok, PerR, SinI, and PhoP, also affect srfA expression [102][103][104][105][106][107][108].comS, which encodes a small protein related to the formation of ComK, is embedded in the srfA operon [109].The codY-dependent inhibition of srfA transcription can be triggered by high external amino acid concentrations.Knocking out codY resulted in a 10-fold increase in surfactin production in B. subtilis 168 [74].Wang et al. [110] found that the protein OppA in the carbon catabolite repression (CCR) pathway was correlated with surfactin biosynthesis.Point mutations in oppA lead to changes in oligopeptide uptake in B. subtilis [105,111].The Rap/Phr system can also affect the production of surfactin in B. subtilis [94,112].Rap inhibits the binding of ComA to the promoter of the srfA operon, whereas Phr inhibits the activity of Rap, indicating that the Rap/Phr system controls the synthesis of surfactin through an interaction with ComA.In addition, the Rap/Phr system regulates the expression of the sporulation-related gene spoIIE at the transcriptional level.These studies revealed the relationship between the antagonistic properties of B. subtilis and the increase of surfactin production at the: gene, protein, and metabolite levels, which may help clarify the regulatory network of B. subtilis.

Genome reduction and genome shuffling can also improve surfactin biosynthesis
With the rapid development of systems biology, genome reduction has become an intensive research hotspot for constructing various promising cell chassis.Researchers used homologous recombination and traceless deletion techniques to construct genomereduced strains [113][114][115].Zhang et al. [89] found that B. amyloliquefaciens LL3 contains a complete srfA operon in its genome that could be used for surfactin biosynthesis.The genome-reduced strain B. amyloliquefaciens GR167 was constructed by deleting approximately 4.18% of non-essential genes from the B. amyloliquefaciens LL3 genome, and it outperformed the parent strain in terms of: growth rate, transformation efficiency, intracellular reduction ability, and heterologous protein expression.Consequently, its surfactin yield was increased by 9.7% versus that of B. amyloliquefaciens LL3.Genome reduction may help improve overall cellular metabolic activity, leading to more efficient surfactin production.These results demonstrated that genome reduction is a feasible strategy for developing cell chassis that can effectively produce bacterial secondary metabolites.
Recombinant production based on appropriate technologies in the fields of molecular genetics and bioinformatics is a promising method for surfactin production.As a new genome-wide engineering method, genome shuffling technology can quickly improve the yield of surfactin in B. subtilis.Genome shuffling through recursive protoplast fusion techniques with multi-parental strains, combined with highthroughput screening methods to select the desired strains, has proven effective.Strains with high product yields can be quickly obtained through genome reorganization without understanding the metabolic regulation mechanism.For example, Chen et al. [96] obtained the genetically stable and high-yield recombinant B. velezensis F34 strain after three rounds of genome shuffling, and its surfactin yield was 3.99-fold higher than of the initial strain.Moreover, the marked upregulation of key genes (srfAA, sfp, and swrC) was consistent with high surfactin production in B. velezensis F34, reflecting the increase in surfactin biosynthesis and cross membrane efflux capacity.This study suggested that genome reorganization is an effective strategy for improving the yield of surfactin.Zhao et al. [97] described the production of surfactin in recombinant B. amyloliquefaciens F2-38 (FMB38) after two rounds of genome shuffling.Surfactin production in shake flasks and bioreactors was increased by 3.5-and 10.3-fold, respectively.The expression of the surfactin synthase gene (srfA) was 15.7-fold higher in B. amyloliquefaciens F2-38 strain in the parental strain.In addition, among the genome-shuffled strains, the levels of key enzymes related to: metabolism, recombination and repair, translation, cell secretion, and energy production were increased.The increased surfactin production in the strain was therefore not the result of a single metabolic pathway or a small amount of protein but instead the comprehensive effect of changes in: gene transcription, translation, and energy metabolism, as well as other related factors.These findings indicate that genome shuffling improves the metabolic capacity of mutants, which will enable the detailed study of the expression and function of genes related to surfactin synthesis in the future.

Heterologous biosynthesis of surfactin
Heterologous expression of genes involved in the surfactin biosynthesis pathway has been challenging because of the large size of the srfA operon and the complex biosynthetic regulation by quorum sensing.
Wild-type B. subtilis 168 cannot synthesize surfactin because there is a stop codon in sfp, which plays an important part in surfactin synthesis.Wang et al. [87] integrated the complete sfp in B. subtilis 168, and the surfactin yield of the recombinant strain reached 0.45 g/L.
With the emergence of recombinant DNA technology and synthetic biotechnology, heterologous expression of biosynthetic gene clusters has become a feasible strategy to: construct, reactivate, improve, and modify natural product pathways [116].The heterologous production of surfactin may also help clarify the mechanism of biosynthesis [117].Heterologous expression of clustered surfactin biosynthetic genes in suitable expression hosts would also be a reasonable choice, but it has not been reported to date.The reason is that the cloning of large DNA fragments and the stable and efficient expression of heterologous biosynthetic genes in hosts remains challenging [118].Additionally, the production of bioactive substances and surfactin can decrease the stability of strains and affect the yield of products.The target biosynthetic gene cluster for the heterologous production of surfactin is relatively large.Therefore, an appropriate cell chassis and suitable molecular genetic tools are essential [106].Pseudomonas aeruginosa, P. putida, and Streptomyces spp., as well as B. subtilis and Escherichia coli, have all been used in biosurfactant production [119][120][121][122].These strains can provide biosynthetic precursors and exhibit sufficient tolerance to their respective surface-active end-products.Several synthetic biology tools have been developed to directly clone entire biosynthetic gene clusters from complex genomes, such as: RecETmediated linear-linear homologous recombination in E. coli [123], yeast transformation-related recombination tools (TARs) [124], the yTREX system based on yeast recombination [118], CRISPR/Cas9-mediated TARs [125], and site-specific recombination mediated by phiBT1 integrase [126].Such systems have increasingly been used to successfully clone and modify biosynthetic genes, including those for the synthesis of biosurfactants [116,124,127,128].In the future, Pseudomonas and Streptomyces spp.can be selected, and recombinant engineering tools can be used to: design more stable genetics, ensure high production performance, realize the effective expression of biosynthetic gene clusters, and replace the traditional surfactin production hosts.In addition, there is increasingly widespread applicability of standardized components for a variety of heterologous hosts, including Streptomyces spp.Once the biosynthetic gene clusters for surfactin are transferred to the selected host, the corresponding engineering strategies to improve the production (e.g.promoter engineering, improving the supply of precursors) are also applicable to the newly constructed strain.The metabolic flow of the heterologous biosynthetic pathway can be statically controlled by regulating: the strength of the promoter, mRNA stability, and RBS engineering [129].Dynamic regulation can also be adopted to: overcome the accumulation of toxic pathway intermediates and their negative effects on the growth of host cells, improve the availability of biosynthetic precursors for secondary metabolism, and maximize the production of target metabolites.The combination of synthetic biology and molecular genetics will improve the regulation of large gene clusters, enabling the optimized production of heterologous surfactin in the final industrial production hosts.

Construction of high-quality artificial cells with xylose-transformed surfactin
A new research direction in the future is the use of synthetic biotechnology and metabolic engineering to improve the ability of genetically engineered strains to surfactin from xylose as the sole carbon source [130].The conversion efficiency of surfactin produced from xylose is low.The pathway by which xylose is channeled into the surfactin pathway in Bacillus spp. is presented in Figure 4.The utilization of xylose is limited by the production performance of strains and enzymes, and the inhibition of AraE by the AraR protein is associated with poor growth of B. subtilis 168 when xylose is used as the single carbon source [131].Recombinant B. subtilis 168 prefers to a medium rich in glucose to produce surfactin [74].However, Hu et al. [42] found that B. subtilis 168 could effectively use xylose when associated with an organic nitrogen source to eliminate the inhibition of AraE by AraR.The recombinant strain BSFX022 produced 2.203 g/L surfactin from xylose.Meanwhile, the reduction in the yield of organic acids by-products (such as acetate and lactate) with xylose as a carbon source enables the effective production of surfactin (up to 2.074 g/L) in a non-buffered system.
In view of the aforementioned reports on the synthesis of surfactin using xylose as the sole carbon source, there were some deficiencies, such as the complex and rigorous CCR mechanism of B. subtilis and the insufficient metabolic regulation of xylose utilization strains.A new research direction is the use of synthetic biotechnology combined with classical metabolic engineering, which has gradually enhanced xylose metabolism in the host and improved the ability of the strain to produce surfactin.Applying synthetic biology design principles (learn, design, build, and test), the pathway "optimization" approach is established through global transcriptome optimization of target bacterial gene clusters, combining: optimized libraries of promoters, RBSs, coding sequences, terminators, and other pathway elements in a metabolic engineering context, as well as screening for the highest level of microbial strains in the combined library by high-throughput techniques [132].
It has been reported that Caulobacter crescentus can effectively metabolize xylose, and many studies have heterologously expressed its related enzymes (including NAD þ -dependent xylose dehydrogenase and xylolactonase) [133] in other microorganisms to establish a new xylose utilization pathway.Related approaches include evolutionary engineering, which identified mutations conducive to the growth of B. subtilis on xylose.These mutations relieved the inhibition by AraR and produced recombinant B. subtilis that can grow rapidly using xylose.The strain design moved from random to rational/semi-rational to improve the efficiency of xylose transformation.Promoting the microbial xylose transport system by: strengthening the expression of the transporter AraE, engineering homologous or heterologous transporters, and screening xylose-specific transporters, can also greatly improve xylose utilization.Combined with the research on reverse metabolic engineering: the relevant endogenous metabolic nodes can be excavated, high-quality artificial cells can be established, and xylose metabolism in genetically engineered B. subtilis, can be improved.In short, synthetic biology will enable the development of new strategies to produce surfactin from xylose, thereby providing an alternative for inexpensive industrial production.

Prospects
The various activities of surfactin have stimulated great interest in applications including: microbial oil recovery, emulsifiers, and medical therapeutic agents.However, high costs and low yields have long represented challenges for the microbial synthesis of surfactin.This review summarized strategies for the production of surfactin at the molecular level, at the fermentation level, and for the future use of xylose, the main component of plant hydrolysates, as a feedstock.Strain engineering may be a valuable means to: reduce the substrate cost, increase the productivity, and finally achieve large-scale production of surfactin, that is economically feasible.For example, the synthesis of surfactin can be effectively improved by: optimizing precursor metabolism, increasing the expression of the gene clusters for surfactin synthesis, blocking competing pathways for its synthesis, and modifying various regulatory factors.Notably, with the development of omics, the modeling of the metabolic networks and the analysis of cellular metabolism combined with metabolomics will help us explore the interactions between metabolites and key metabolic modules, and more information that may be useful for surfactin production can be identified.Meanwhile, future research should consider the establishment of appropriate hosts for heterologous expression of surfactin biosynthesis genes, providing chassis strains to achieve high-efficiency surfactin production.Finally, synthetic biology can be used to improve the ability of strains to utilize inferior biomass feedstocks for surfactin production, which may provide an effective strategy for technically feasible surfactin production from inexpensive biomass hydrolysates.

Figure 2 .
Figure 2. The strategies for improving the biosynthesis of surfactin discussed in this review.(A) Medium optimization; (B) Genome engineering; (C) Heterologous expression; (D) Synthetic biology and metabolic engineering on xylose utilization.

Figure 3 .
Figure 3.The typical factors affecting the expression of srfA.Upper red box: ComA, a partial transcriptional regulator, is a key factor in regulating surfactin synthesis.ComA binds to the srfA promoter and regulates its transcription, thereby regulating srfA expression.The purple box below: Phr-Rap protein regulates surfactin synthesis by interacting with ComA.At low cell densities, in which the Phr concentration is relatively low, Rap proteins inhibit ComA and Spo0A activity.When the population density reaches high levels, the Phr signal peptide is re-imported into cells by the oligopeptide permease Opp and the oligopeptide permease Spo0K, and it represses Rap activity, thereby activating ComA-P and Spo0A-P to control gene transcription.

Figure 4 .
Figure 4.The pathway by which xylose is channeled into the surfactin pathway in Bacillus spp.

Table 1 .
Comparison of surfactin titers following medium optimization.

Table 2 .
Examples of enhanced surfactin biosynthesis via genome engineering.