Introduction

Fatty acids (FAs) are essential components of cell structure and play a crucial role in energy production. Common types of FAs include saturated/ unsaturated, cis/ trans, and straight-chain/ branched-chain FAs based on different configurations [1]; while according to the molecular chain lengths, they are usually divided into short-chain FAs (SCFAs), middle-chain, and long-chain FAs (MCFAs & LCFAs) [2]. Besides their functions regulating cell signals, FAs are essential substrates for energy and anabolic metabolisms [3,4,5].

Organisms can synthesize, decompose and utilize FAs through different oxidation pathways. Three known pathways of FAs degradation are β-, α-, and ω-oxidation. Among these, α- and ω-oxidation are minor ways primarily occurring in eukaryotes, whereas β-oxidation is the main pathway to degrade bacterial FAs [6]. Since the synthesis process of FAs is highly energy-consuming, organisms can also absorb and use exogenous FAs to participate in their own physiological activities [7].

FAs and their derivatives are necessary in extensive cellular activities, such as bacterial pathogenesis, membrane permeability, and phospholipid biosynthesis [8,9,10,11]. With the rapid enrichment of microbial genome sequence databases, most sequenced bacteria have been confirmed to possess β-oxidation-related genes. However, the mechanisms of bacterial β-oxidation need to be further studied. Here, we first introduce some new achievements in bacterial β-oxidation; then, using Xanthomonas as an example, we discuss the research progress on the degradation mechanism of the DSF-family quorum signal molecule; based on these findings, we predict the degradation mechanism of FAs with a specific structure. Finally, we outline some future follow-up researches directions.

Full Overview of FAs β-Oxidation in Escherichia coli

The metabolism of LCFAs (C12—C18) and MCFAs (C7—C11) in E. coli requires the fad regulon genes involved in the transportation (fadL), activation (fadD), and oxidation (fadA, B, E, H, and M) of FAs (Fig. 1). These fad genes are located at different positions on the E. coli chromosome and form a regulon consisting of a few operons [12]. In addition, fadR regulates the transcription of FA degradation-related genes by binding downstream of fad genes (fadL, fadD, fadE, and fadBA) and inhibiting their transcription [13, 14]. The transportation and utilization of SCFAs (C4—C6) are conducted by the atoDAEB operon, which is controlled by AtoC [12, 15].

Fig. 1
figure 1

The β-oxidation pathway of fatty acid in bacteria. Schematic representation of degradation of fatty acids. The common even-carbon saturated FAs are metabolized via β-oxidation (A). While to allow the degradation of FAs with various steric variants such as VLCFAs (B), branched chain fatty acids (C) or unsaturated fatty acids (D & E & F) to proceed, different strategies have been developed. These strategies include evolution of enzymes showing specificity with respect to either chain length or modified acyl-chain, or other special conformation, to catalyze a parallel set of reactions with different substrate specificities. The β-oxidation cycles are shown in grey. FadB and FadA play dual roles in β-oxidation pathway. FadB, the α-subunit of the multienzyme complex FadBA, has dodecenoyl-CoA delta-isomerase, enoyl-CoA hydratase, 3-hydroxybutyryl-CoA epimerase and 3-hydroxyacyl-CoA dehydrogenase activity (enzymatic step shown in orange). FadA has 3-ketoacyl-CoA thiolase activity, catalyzes the last step of a β-oxidation cycle, and this small β-subunit of FadBA is responsible for the release of two carbon atoms to form acetyl-CoA in thiolysis reactions in the fatty acid β-oxidation cycle (enzymatic step shown in blue)

Full size image

Fatty Acid Transmembrane Transporter (FadL)

FadL (48.5 kDa) is localized in the outer cell envelope and can bind exogenous LCFAs with high affinity [16]. Since FAs have low cell membrane permeability, FadL functions as a substrate-specific diffusion channel to achieve transmembrane [17]. The exogenous FAs transportation process is ATP- and CoA-required [12]. van den Berg firstly reported FadL’s crystal structures and found that FadL constitutes a presumed β-barrel with 14 transmembrane β-strands and 7 extracellular loops [18]. Thus, a lateral-diffusion transport mode of FadL-uptook LCFAs has been built, the FA transport activity is regulated by the C-terminal of FadL, and the FadL expression can largely control the whole-cell biotransformation rates of FAs [18,19,20]. FadL works together with fatty acyl-CoA ligase (FadD) to vectorial esterification-induced concomitant transport and activation of CoA thioester [21, 22], thereby rendering this process unidirectional.

Fatty Acyl-CoA Ligase (FadD)

FadD participates E. coli β-oxidation by catalyzing free FAs to acyl-CoA thioesters. Besides the catalyzation of endogenous FAs released from membrane lipids, FadD plays a pivotal role in the catalyzation of exogenous LCFAs [23]. The molecular weight of FadD is about 62 kDa [24]. FadD contains two structural binding domains for ATP and FA binding separately [25]. It can activate FAs with C6-C20 chain lengths, showing the best specificity for C14-C18 [21]. FadD catalyzes the conversion of LCFAs for their later reduction into intermediates, which are involved in serving as substrates for β-oxidation and phospholipid biosynthesis [26, 27]. Knocking out of the fadD gene leads to loss of transport function and degradation of LCFAs [21, 26], and resulting in significantly decreased E. coli viability in its stationary growth period [28], suggesting a critical function of FA oxidation for survival capacity. The FA transport system in microbes is partially shock-sensitive, suggesting the requirement of precise chemical composition (pH, periplasmic protein, etc.) [21]. It is interesting that loss-of-function of FadD decreased the expression of fadL via FadR [29], and the overexpression of fadD in E. coli strongly up-regulated the expression of fadBA and fadE [30]. A 3D model for the E. coli FadD has been proposed, and the crystal structure demonstrates that the N-terminal domain is the main part, composed by a twisted antiparallel β-barrel flanked by two β-sheets flanked by α-helices [31].

Fatty Acyl-CoA Ligase (FadK)

The molecular weight of FadK is about 62.77 kDa. Functional and expression identification of FadK indicate that it functions similarly to ligases FadD by producing an acyl-AMP intermediate before forming the final acyl-CoA, demonstrating that it is a second E. coli acyl-CoA ligase. With SCFAs as preferred substrates, FadK displays a stronger relative activity when compared with LCFAs for MCFAs but weaker absolute activity than FadD [32]. The fadK expression is suppressed during aerobic growth but promoted in the presence of terminal electron acceptor such as fumarate in an anaerobic environment [33].

Acyl-CoA Dehydrogenase (FadE)

As a flavinase, the flavoprotein of FadE (75 kDa) can transfer electrons from the cofactor of FAD/FMN to the electron-transport chain [12]. FadE catalyzes the first step of the activated FAs’ degradation, except for Δ2-unsaturated FAs. The fact that fadE encoding the sole acyl-CoA dehydrogenase is genetically confirmed in E. coli. FAs with various chain lengths cannot be metabolized in E. coli fadE mutants [25]. The substrates of FadE are fatty acyl-CoA esters, which limits the experiments because of their complicated chemical synthesis, especially for LCFAs [34].

Multienzyme Complex of β-Oxidation (FadA & FadB)

As a multienzyme complex of β-oxidation, FadA and FadB are critical for the degradation of FAs in E. coli. It consists of an α subunit (FadB, 79 kDa) and β-subunit (FadA, 41 kDa), catalyzing five reactions during the degradation of various FAs [35]. The β-oxidation complex FadBA comprises 2-enoyl-CoA hydratase, 3,2-enoyl-CoA isomerase, 3-hydroxyacyl-CoA dehydrogenase, 3-hydroxyacyl-CoA epimerase and 3-oxoacyl-CoA thiolase activities. Substrate specificity studies demonstrated that this β-oxidation complex is required for the substrate metabolism [12]. The multienzyme complex of FAs β-oxidation shows features that from non-mitochondrial systems (in bacteria and peroxisomes of eukaryotes).

FadA

The β-subunit of FadBA (product of fadA) is 40.9 kDa with 387 residues. FadA conducts 3-ketoacyl-CoA thiolase activity to catalyze the last step of β-oxidation. This small β-subunit of FadBA works on FAs with different chain lengths to release two carbon atoms, formimg acetyl-CoA in thiolysis reactions.

FadB

The α subunit (79.6 kDa) of FadBA (product of fadB) contains 729 residues. FadB can acts as 3-hydroxyacyl-CoA epimerase to catalyze 2-trans-enoyl-CoA to L-3-hydroxyacyl-CoA and as enoyl-CoA hydratase to convert from D-3-hydroxy of acyl-CoA to 3-trans-enoyl-CoA [12]. Taking this configuration isomerism into consideration, it is needed to mention that a kinked steric structure of cis-unsaturated FAs (cis-UFA) leads to highly mobile membranes; Conversely, the more extended trans-UFAs reduce membrane fluidity when compared to their cis-isomers. Since the adaptation mechanism of bacteria to environmental stress is that they can change the cis-unsaturated FAs into trans ones in membrane lipids, increasing the membrane stiffness to resist stress better [9], and LCFAs are difficult to be biodegraded [36], thus, FadB may be important for bacteria adaptation and for modulating membrane permeability and fluidity.

Typically, the enzymes that catalyze 2-enoyl-CoA into S- or R-3-hydroxyacyl-CoA are named enoyl-CoA hydratases (ECHs, a domain of FadB), in which the S- and R-specific type is named ECH-1 and ECH-2, respectively, and their substrates have different specificity for chains of different lengths [37], and their geometries active sites are similar, even if in a mirror image fashion [38]. It is remarkable that only ECH-2 (R-ECH) but not ECH-1 comprises epimerase activity and can specifically catalyze 3-hydroxyacyl-CoA from R- or D- type to generate S- or L- type. Without the epimerase activity of ECH-2, all cis-UFAs are difficult to undergo the following steps of β-oxidation. ECH-2 is speculated to re required in the lipidic intermediates metabolism in bacteria.

2,4-dienoyl Reductase (FadH)

FadH assists in the β-oxidation of some unsaturated FAs [39]. FadH (72.55 kDa) is a monomer, in which N-terminus contains a FAD-binding domain and C-terminus has a NADPH-binding domain [12]. FadH catalyzes the β-oxidation reactions of specific unsaturated FAs with double bonds at contiguous even-numbered positions, producing 3-trans-enoyl-CoA, which forms 2-trans-enoyl-CoA stimulated via FadB isomerase activity [12]. Feng and Cronan [39] reported that fadH is regulated by three independent regulators, oxygen response system ArcA-ArcB, the FA degradation repressor FadR, and the complex of cyclic AMP receptor protein-cyclic AMP (CRP-cAMP). The expression level of fadH increases threefold by the deletion of arcA, and 8- to tenfold after inactivation of fadR and arcA under anaerobic conditions, suggesting that anaerobic expression is inhibited by ArcA-ArcB and FadR [39]. The binding of FadR to the fadH promoter can be reverted by the thioesters of long-chain acyl-coenzyme A, consistent with that fadH is strongly induced by LCFAs, and the activation of fadH by CRP-cAMP complex and prototype fad is similar [39].

Acyl-CoA Thioesterase (FadM)

Researchers reported a novel long-chain acyl-CoA thioesterase FadM, which can cleave the thioester bonds of acyl-CoA during β-oxidation [40, 41]. FadM is 15 kDa, composed of 132 amino acids. The expression of fadM is confirmed to be negatively regulated by CRP-cAMP complex, which is different from other fad regulon genes that are positively regulated [40, 41]. FadM is required for β-oxidation of unsaturated FAs, such as oleic acid [42].

Transcriptional Regulator of fad Genes (FadR)

FadR is 26.97 kDa, composed of 239 amino acid residues. Coexpression of FadR led to enhanced production of free FAs in E. coli [43], it is a critical transcriptional regulator which are capable of down-regulate the expression of gene involved in β-oxidation (fadBA, fadD, fadE, fadH, fadL, fadM). Interestingly, FadR can also act as an activator of transcription [44]. For example, iclR is one of the genes activated by FadR [45]. IclR is a repressor of the aceBAK operon coding for the enzymes of the glyoxylate shunt. In addition, FadR activates almost all FA biosynthesis genes (such as unsaturated FA biosynthetic genes fabA and fabB) [13]. Hence, FadR is a transcriptional factor that positively regulates anabolism and negatively regulates the catabolism of the same family of molecules [46]. This dual role of FadR in FA degradation and synthesis seems specific to E. coli, as in other bacteria, two distinct regulators are commonly present to solve the two functions [44]. During β-oxidization, the activated C12-C18 LCFAs remove the repression of FadR by binding to it and causing the protein to be released from DNA. In contrast, MCFAs (C7–C11), do not exhibit this feature [47]. Thus, E. coli utilizes LCFAs (but not MCFAs), as the only source for generating carbon and energy. However, wild-type E. coli was once reported to utilize MCFAs as a unique carbon source in anaerobic environment, illustrating that FadR is not the key determinants regulating the FA metabolism [25]. Moreover, FadR is induced to express by LCFAs, but not by MCFAs [12]. On the FadR 3D structure, a dimer and a two-domain fold, a DNA-binding domain in N-terminal, and an acyl-CoA-binding domain in C-terminal were observed [48].

Fatty Acids β-Oxidation Research in Other Bacteria

Studies aimed at better characterizing FA catabolism in bacteria other than E. coli are increasingly intriguing. As a result of high-throughput technologies, a growing number of FA β-oxidation genes have been found in various microbes. Despite many FA β-oxidation genes present in bacterial genomes with essential functions in host survival, limited information is available regarding the metabolic details and the related regulatory mechanisms of these genes. Although the mechanism of β-oxidation in bacteria is relatively conserved, the biological functions mediated by β-oxidation pathways are diverse.

Bacillus subtilis: The FA β-oxidation process in B. subtilis is similar but not identical to that in E. coli, with a major difference being that B. subtilis can break down branched chain fatty acids while E. coli cannot [49,50,51]. Many β-oxidation-related genes, such as lcfA, lcfB (yhfL), fadB (ysiB), fadN (yusL), fadA (yusK), fadE (yusJ) and fadG, have been detected in B. subtilis. The disruption of fadA, fadE, fadG or fadN significantly affects FAs utilization [51]. Simultaneously, the global transcriptional regulator FadR, which can be inactivated by long-chain acyl-CoAs, inhibits the transcription of genes regulating FA degradation in B. subtilis [50].

Haemophilus parasuis, the pathogen of Glässer’s disease, can cause arthritis, fibrinous polyserositis, and meningitis [52]. The H. parasuis genes HAPS_1695 & HAPS_0217 (designated as fadD2 & fadD1) are identified as FadD and prefer LCFAs as substrate. Physiological function results show that either of these two genes is necessary for their survival, revealing that these genes are target candidates for antibiotics or other drugs to control Haemophilus pathogenicity [52]. In order to understand the virulence level among strains, Hill et al. detected the expression of virulence-related genes in H. parasuis by reverse transcription-polymerase chain reaction [53]. The expression of FadD was significantly up-regulated after acute infection, suggesting that FadD may be involved in virulence. A similar result about the relationship between FadD and virulence was reported in Pseudomonas aeruginosa [54], where FadD is implicated in bacterial pathogenicity.

Mycobacterium: Biofilms, a bacterial strategy to adapt to harsh environments, are synthesized using intermediates from the β-oxidation cycle [55, 56]. Unfortunately, the specific correlation between FA β-oxidation and mycobacterial lipid metabolism is still unclear. Xu and colleagues [57] identified a TetR-like transcription factor MmbR in M. smegmatis and found this regulator responsible for β-oxidation and biofilm formation. The genome of intracellular pathogen Mycobacterium tuberculosis (Mtb) contains approximately 100 redundant genes related to the β-oxidation process [58, 59]. To better understand the lipid metabolism in Mtb, Venkatesan and Wierenga [60] isolated and analyzed the complex of FA β-oxidation (trifunctional enzyme, TFE), which catalyzes the β-oxidation reactions. Structural analysis showed that TFE is a complex encoding for the fadA-encoded TFE-β and the fadB-encoded TFE-α. In addition, Cox et al. [58] reported the crystal structure of FadB2. They explained the functional difference between monofunctional FadB2 and the trifunctional FadA-FadB complex: FadB2 cannot dimerize or replace FadB in their complex because it doesn’t have the hydrates domain. However, this does not eliminate the effect of FadB2 on the FadB hydration product [58]. A recent investigation [61] provided evidence that the mycolic acid desaturase regulator (MadR) mutant of M. smegmatis showed a damaged cell wall and accumulation of desaturated α-mycolate, indicating that MadR controls the desaturation and biosynthesis of mycolic acid. Additionally, the transcriptomic profiling results also implicated that MadR indirectly regulates the β-oxidation pathways, as lipid metabolism genes were significantly down-regulated in madR mutant.

Myxococcus: Lipid bodies, such as triacylglycerides, are important for fruiting body formation, and lipid body catabolism by FA β-oxidation provides the major energy during Myxococcus xanthus development [62]. Bioinformatics and experimental analysis showed that M. xanthus has two gene sets, MXAN_5371 (designated as fadJ) and MXAN_5372 (designated as fadI), and their expression increased at the developmental stage when lipid body accumulation reaches its maximum, which involved in the degradation of lipid bodies and flavin adenine dinucleotide. Deletion of fadI (MXAN_5372 and MXAN_6988) and fadJ (MXAN_5371 and MXAN_6987) caused developmental defects, increased lipid body content, and slowed spore coat maturation [63, 64]. Therefore, this finding suggests that β-oxidation participates in fruiting body growth and spore coat maturation, as well as increases the ability of spores to resist UV light and/or heat [62].

Pseudomonas aeruginosa is a well-known pathogen causing lung infection in patients who have cystic fibrosis (CF) [65]. Kang et al. [65] found that the expression of P. aeruginosa fadD2 influences the production of phospholipases, rhamnolipids, lipases, proteases, and is also relevant to swimming and swarming motilities, whereas fadD1 gene only participates in swarming motility. Zhang et al. [66] illustrated that the expression levels of rhamnolipid biosynthetic genes, rhlA and rhlB, were increased by 200- to 600-fold, while fadA was only up-regulated 3- to 4- fold with the addition of octadecanoic acid. These findings demonstrated that there are indeed pathophysiological interconnections between virulence factors and FAs β-oxidation. Moreover, it was reported that psrA, a TetR-type transcriptional regulator, could control quinolone signal and β-oxidation-related genes in P. aeruginosa, such as fadE [67, 68]. Induction of the fadBA5 operon by PsrA produces a signaling response to LCFA, indicating that binding of LCFAs to PsrA relieves fadBA5 repression [67].

In Ralstonia eutropha, acetyl-CoA or FA metabolic intermediates can be used for high yield production of polyhydroxyalkanoates (PHAs) or different PHA precursors, respectively. PHA, as an energy and carbon storage compound produced by microorganisms, has the potential to replace the petroleum-based plastics because of their excellent biodegradability and biocompatibility. Besides PHAs and PHA-related polymers, R. eutropha can also produce many other polymers, making this species a model organism for PHA biosynthesis [69]. However, few studies have reported that these genes in this species are actually involved in FA degradation, even though the genomic sequence of R. eutropha H16 is known and some potential β-oxidation pathway gene homologs have been annotated [49, 70]. Microarray and mutation analyses demonstrated two operons candidates that encode the β-oxidation-related enzymes in R. eutropha, but at least one operon is needed for FA degradation [71, 72].

Salmonella enterica: Although the fatty acid metabolism pathways in S. enterica and E. coli are considered almost the same according to the sequence alignment, experiments have shown that S. enterica has greater efficiency of β-oxidation than E. coli. The difference is manifested in (1) S. enterica could grow normally with decanoic acid, while E. coli cannot; (2) FAs can be completely degraded to acetyl-CoA in S. enterica, whereas many intermediate products are produced in E. coli, which can be explained by the different functions of fadE and fadBA in different organisms [73]. Except for being involved in FA degradation, FadD also activates hilA gene expression, which is required for invasion. In S. enterica serovar Typhi HEp-2 cells, the hilA expression and invasiveness was reduced by fadD mutation [74].

Streptomyces coelicolor is commonly used to produce antibiotics for clinical use and can accumulate high-level triacylglycerol (TAG) to supply energy and carbon. TAG mobilization is correlated with high-energy production due to the high reductive status of acyl moieties [75]. S. coelicolor can efficiently take up exogenous FAs as specific carbon and energy sources for membrane phospholipids synthesis or store them in neutral lipid compounds, like TAG [75]. Menendez-Bravo et al. [76] discovered that three gene clusters (SCOfadAB_1, SCOfadAB_2 and SCOfadAB_3) are involved in the FA β-oxidation in S. coelicolor. They confirmed the correlation of SCOfadAB with TAG by flux balance and mutation analyses, suggesting that defective β-oxidation could inhibit the mobilization of TAG.

Vibrio: Vibrio cholerae is the leading cause of the diarrheal disease cholera. Bile unsaturated FAs drastically inhibit the transcript of virulence factors and enhance motility [77]. To verify the relationship between the TetR family transcriptional repressor PsrA and the pathogenicity of V. cholerae, Yang et al. [78] found a significant decreased in colonization ability, as well as reduced the expression of fadB, fadA, fadI and fadJ genes in the psrA mutant, demonstrating that PsrA is a key regulator that inhibits the expression of fad regulon genes. In addition, the fadR gene is essential for bacterial infection in many pathogens. Brown and Gulig [79] first demonstrated that fadR can cause infection in animal hosts during their study on V. vulnificus, with a mechanism closely related to FAs metabolism. The fadR mutants led to a low infection rate in localized skin damage and fatal systemic liver infections, insensitivity to envelope pressure, reduced mobility, and changes of membrane lipid structure. Furthermore, infection with oleate in vitro promoted infection in mice by fadR mutant. Simultaneously, it showed significantly decreased and increased content of unsaturated and saturated FAs, respectively, suggesting that these fadR mutants have defective ability to biosynthesize unsaturated FAs [79].

β-Oxidation of DSF-Family Quorum Sensing Signal Molecules and the Relevant Mechanisms Underlying Bacterial Pathogenicity

The Gram-negative plant pathogenic bacteria Xanthomonas comprise various species that can cause illness in more than 400 kinds of economically important crops [80]. All Xanthomonas can synthesize and secrete a type of DSF (diffusible signal factor)-family quorum sensing (QS) signal to sense population density and regulate various biological processes, such as virulence, biofilm diffusion, and ecological ability [81, 82]. The major quorum-sensing signal molecules from DSF-family identified from Xanthomonas campestris pv. campestris (hereafter referred to as Xcc) belong to a class of cis-unsaturated LCFAs, including the DSF (cis-11-methyl-2-decenoic acid) and BDSF (cis-2-decenoic acid) [83,84,85]. Barber [86] and Slater [87] demonstrated that in batch cultures of Xcc, DSF accumulated at the early stationary stage, but then decreased. Wang et al. [83] solved the structures of DSF by NMR, and further confirmed that the level of DSF remained relatively low at the early-growing stage, subsequently increased substantially, and then decreased rapidly at the post-logarithmic growth phase. For Xanthomonas oryzae pv. oryzae (Xoo hereafter) strain KACC10331 cultured in YEB medium, He et al. [88] found that BDSF production started to increase after 18 h inoculation and reached its maximum at 36 h, and then significant decreased at 42 h in the supernatants. These results indicated a naturally-occurring signal turnover phenomenon in Xanthomonas.

RpfB enzymatic activity is required for DSF and BDSF turnover. A recent study unveiled that RpfB functions as a fatty acyl-CoA ligase that plays a role in FA β-oxidation to give acyl-CoAs for membrane lipid synthesis in Xcc. It also plays a more important role in pathogenesis by counteracting the thioesterase activity of DSF synthase RpfF [89]. Deletion of rpfB (ΔrpfB hereafter) slightly increased Xcc virulence in Chinese radish, while overexpression of rpfB [ΔrpfB(rpfB) hereafter] significantly decreased virulence [85]. Zhou et al. [85] developed new methods to quantify the DSF family genes by the combined use of ultra-performance liquid chromatographic (UPLC) and time-of-flight mass spectrometry (TOF–MS). using these new methods, Zhou discovered that mutation of rpfB (ΔrpfB) increased the biological production of DSF and BDSF for ~ 10.0-fold compared to the parental strain Xcc or Xoo. Overexpression of rpfB or E. coli fadD in ΔrpfB [ΔrpfB(rpfB)] blocked the production of DSF and BDSF; ΔrpfB(rpfB) could rapidly degrade exogenously added DSF and BDSF, but ΔrpfB had no degradation ability [83]; RpfB with an E-365 mutation showed largely decreased DSF and BDSF degradation activities. Therefore, RpfB is needed for the degradation of DSF and BDSF signal molecules and represents to a naturally-occurring signal conversion system targeting at DSF-family QS signals in Xcc and Xoo (Fig. 2). RpfB-dependent signaling turnover is also widely observed in human opportunistic pathogens, like Burkholderia spp. and Stenotrophomonas maltophilia, as well as environmental bacteria including Leptospirillum, Frateuria, Lysobacter, Luteibacter, Methylobacillus flagellates, Rhodanobacter, and Thiobacillus denitrificans [90]. Thus, future studies on these microbes could also benefit from this review. Furthermore, the Xcc genome also contains homologues of other genes regulating E. coli FA β-oxidation, such as fadL(Xcc0017), fadE (Xcc2870), fadA (Xcc1978), fadB (Xcc1979, Xcc0810) and fadH (Xcc0933). Recently, preliminary results confirmed that Xcc1979 and Xcc1978 are involved in DSF degradation. Although the functions of other genes remain to be verified, one of the Xcc potential naturally occurring DSF signal turnover systems should be FAs β-oxidation pathway.

Fig. 2
figure 2

Schematic representation of DSF-mediated QS signaling network and QS signal turnover system in Xcc. Fatty acids β-oxidation pathway participates in the degradation of DSF and BDSF signal molecules and represents a naturally occurring signal turnover system that targets DSF-family QS signals in Xcc. Taking Xanthomonas as an example, the DSF-family quorum sensing signal molecular degradation mechanism is introduced. RpfF is one of the key enzymes required for DSF biosynthesis, DSF is sensed and transduced by the two-component signaling system RpfC–RpfG through a phosphorelay cascade, and the downstream second-messenger cyclic di-GMP, The global regulator Clp is the effector of cyclic di-GMP, which regulates directly and indirectly the expression of genes for virulence and adaptation [92]. Zur a transcriptional regulator involving in zinc uptake regulation, FhrR a transcriptional regulator required for flagella biosynthesis and hypersensitive response (Hrp)

Full size image

Compared with E. coli β-oxidation, the DSF signaling molecule degradation pathway in Xanthomonas has at least two unique phenomena. (1) An unique epimerase is needed, and the function of the epimerase that converts (R)-3-hydroxyacyl-CoAs to (S)-3-trans-hydroxyacyl-CoAs may come from FadB (Xcc1979); otherwise DSF cannot be degraded; (2) Xcc has some predicated homologues involved in FA metabolism but lacks a FadR homologue; rpfB is under the direct regulation of a global transcription factor Clp [85, 91].

DSF-producing microbes are widely distributed [92]. Bacteria other than Xanthomonas also synthesize DSF-family quorum sensing signals. P. aeruginosa can synthesize cis-2-decenoic acid, which can trigger dispersion during microbial biofilms [93]; Xylella fastidiosa can produce cis-2-tetradecenoic acid and cis-2-hexadecenoic acid (XfDSF1 & XfDSF2) [94]; and the Gram-positive bacterium Streptococcus mutans can produce signaling molecule trans-2-decenoic acid (SDSF) to inhibit Candida albicans hyphal formation [95].

The core rpf signaling-components, including genes such as RpfB, RpfG, RpfF, and RpfC, have been functionally verified in all sequenced Xanthomonas species. Homologous gene clusters also exist in Lysobacter dokdonensis DS-58, Methylobacillus flagellatus, Pseudoxanthomonas spadix BDA-59, Stenotrophomonas maltophilia, Thiobacillus denitrificans and X. fastidiosa. RpfF encodes a key enzyme required for DSF biosynthesis, whereas RpfC and RpfG constitute a two-component system involved in signal perception and transduction in Xcc [87, 96]. The activated HD-GYP domain of RpfG has phosphodiesterase activity and can degrade cyclic di-GMP (c-di-GMP), an inhibitory ligand of the global transcription factor Clp. Consequently, derepressed Clp drives the expression of several hundred genes, including those encoding virulence factor production [97]. BDSF (cis-2-decenoic acid) and other similar structure signal molecules have been found in Burkholderia cenocepacia and P. aeruginosa, and homologous proteins of RpfB are also present in the two strains [91]. Two RpfB proteins in Lysobacter enzymogenes were identified by Li et al. [98]. RpfB was shown to have substitutional function for FadD in Xcc [89]. In vitro experiments demonstrated that RpfB could activate various FAs to the corresponding CoA esters [89], and these RpfB-activated fatty acyl-CoAs would be degraded further during the β-oxidation process. Recent researche confirmed that the mutants of rpfG, rpfC, or clp, in Xcc and Xoo all showed increased transcriptional and translational levels of rpfB [85]. The above-mentioned results demonstrated that RpfB-dependent FAs β-oxidation may be a conserved quorum sensing turnover mechanism in bacteria (Fig. 2).

Steric-Problems-Solving Strategies for β-Oxidizing in Different Acyl-CoA Variants

The common even-carbon saturated FAs are metabolized via β-oxidation (Fig. 1). However, FAs with special structures also exist in living organisms. β-oxidation includes four reactions that occur in repeating cycles, producing intermediates such as trans-2-enoyl-CoA, 3-hydroxyacyl-CoA, and 3-ketoacyl-CoA. In this way, β-oxidation can be easily explained for straight-chain saturated FAs, but not for all modified general FAs. Several methods, such as evolutionary analysis of enzymes’ specificity regarding chain length, modified acyl-chain or any special conformation, have been explored to ensure the degradation of FAs with various steric variants such as very long chain fatty acids (VLCFAs), unsaturated fatty acids, and branched chain fatty acids [1]. In general, FAs can be degraded via different mechanisms, including α-, β- and ω-oxidation.

α- and β-oxidation for VLCFAs. The acyl-group with more than 20 carbons (C20) can undergo β oxidation (Fig. 1B). Currently, there are three strategies to oxidize substrates of different chain lengths, such as development of the adaptable active sites, acquisition of paralogues, and functional convergence of isoforms [1]. α-oxidation shortens FAs by one carbon from the carboxyl-end. It is a peroxisome reaction that activates the 3-methyl-branched FA to a 3-methyl-branched fatty acyl-CoA, which is aerobically oxidized by fatty acyl CoA 2-hydroxylase to produce a 2-hydroxy-3-methyl-branched fatty acyl-CoA.

β-oxidation of unsaturated FAs (Fig. 1). β-oxidation allows for the complete degradation of unsaturated FAs with an even number of carbon though FadE-mediated dehydrogenation to form double bonds. However, such cases are not suitable for all FAs. Extra steps are needed for the oxidation of double-bonds FAs based on the double bond position and the configuration of FAs. Each double bond reduces the generation of one FADH2, but one less step of FadE-mediated dehydrogenation. The double bond position and the configuration ofFAs also affect the β-oxidation process. (1) The prevented double bond between C2–C3 by that of C3–C4 can be solved by a reaction that converts the double bond of the cis-Δ3 into a trans-Δ2 type via cis-Δ3-enoyl CoA isomerase (ECI1). Thus, the double bond between C2 and C3 is finally generated, sharing the following reactions with those of saturated FAs, as trans-Δ2-enoyl CoA is a regular substrate. For example, in unsaturated palmitoleate with one double bond between C9 and C10, its cis-Δ3 -enoyl CoA is converted into trans-Δ2-enoyl CoA first, allowing the oxidation pathway continue. (2) The existence of a double bond between C2 and C3 improves the hydration reaction of both trans-Δ2-enoyl CoA and cis-Δ2-enoyl CoA into 3-hydroxyacyl-CoA, whereas the product of cis-Δ2-enoyl CoA is D-3-hydroxyacyl-CoA, not L-3-hydroxyacyl-CoA. Therefore, 3-hydroxyacyl-CoA epimerase is required to convert D-3-hydroxyacyl-CoA into L-3-hydroxyacyl-CoA, resulting in the continuation of the FAs β-oxidation. Typically, 3-hydroxyacyl-CoA epimerase participates in the metabolism of cis double bonds at even-numbered positions. The mechanism of epimerization indicates that the epimerization of 3-hydroxyacyl-CoA is accomplished by the combined functions of ECH-1 and ECH-2. (3) Polyunsaturated fatty acids (PUFAs) are precursors of signaling molecules. The β-oxidation of PUFAs is a little different because the 2,4-dienoyl intermediate produced by PUFAs β-oxidation cannot be used as a substrate for the next step’s enzymes. Fortunately, this issue is avoided by 2,4-dienoyl-CoA reductase and cis-Δ3-enoyl CoA isomerase, which transform the 2,4-dienoyl intermediate into a customary trans-Δ2.

β-oxidation of methyl-branched fatty acids (Fig. 1C) needs the joint collaboration of β- and α-oxidation pathways, as well as α-methylacyl-CoA racemases [1]. β-oxidation would be halted if the methyl groups cause steric problems and inhibit the second dehydrogenation step. Fortunately, this issue will be solved by the α-oxidation pathways. α-oxidation usually degrades FAs with special structure. For example, phytanic acids attached to CoA and form phytanoyl-CoA, which is then oxidized to 2-hydroxyphytanoyl-CoA. This is followed by the cleavage of 2-hydroxyphytanoyl-CoA to form pristanal and formyl-CoA, and then the oxidization of pristanal to form pristanic acid, which conducts β-oxidation.

Fatty acid ω-oxidation. In ω-oxidation, the methyl group is firstly hydroxylated at the ω-end under the canalization of cytochrome P450, followed by ω-end oxidation to produce 1, ω-dicarboxylic acid, which can be shortened by β-oxidation. It has been reported that the α-chlorinated FAs can be degraded by ω- and β-oxidation successively from the ω-end [99]. Some microbes can convert alkanes into FAs through ω-oxidation, which can be used for environmental protection, such as cleaning up oil spills. It has been also demonstrated that VLCFA can undergo ω-oxidation [100].

Concluding Remarks and Future Perspectives

Regarding the even-carbon saturated-FAs in E. coli, their primary metabolic pathways have been well-studied, and the metabolic mechanism has been basically clarified. Fatty acid oxidation pathways and mechanisms in other bacteria are also being studied. The specific roles of fatty acid β-oxidation in Xanthomonas, particularly its connection to DSF-mediated QS and bacterial pathogenicity, are being explored. However, several areas require further investigation to fully elucidate the complexities of fatty acid β-oxidation in bacteria, particularly in pathogenic contexts: (1) The selectivity of substrates, enzyme activity and molecular mechanism of the key enzymes within β-oxidation pathway, such as FadBA complex, need further elucidation, to shed light on how bacteria utilize β-oxidation to regulate virulence through QS signaling etc. (2) The degradation enzymes of many specialized structures of FAs (such as DSF-family QS signals, which comprises cis-2-unsaturated FAs of different chain lengths and branching) and their mechanisms require additional study. (3) The correlation and crossover between β-oxidation and other FAs metabolic pathways deserve further investigation. (4) An important future research direction should be the use of synthetic biology methods to modify the β-oxidation pathway and synthesize products with application value. (5) The function of FA degradation pathways in the symbiotic environment of animals, plants, humans and microorganisms needs further study. As a factor of the global regulatory network, QS controls many microbial traits, such as symbiosis, virulence, development and maintenance of population, and community structures. Addressing the above issues will be key to gaining a more detailed understanding of the bacterial utilization, modification or prevention strategies.