Characterization of the Flavin-Dependent Monooxygenase Involved in the Biosynthesis of the Nocardiosis-Associated Polyketide

Abstract

Some species of the Nocardia genus harbor a highly conserved biosynthetic gene cluster designated as the NOCardiosis-Associated Polyketide (NOCAP) synthase that produces a unique glycolipid natural product. The NOCAP glycolipid is composed of a fully substituted benzaldehyde headgroup linked to a polyfunctional alkyl tail and an O-linked disaccharide composed of 3-α-epimycarose and 2-O-methyl-α-rhamnose. Incorporation of the disaccharide unit is preceded by a critical step involving hydroxylation by NocapM, a flavin monooxygenase. In this study, we employed biochemical, spectroscopic, and kinetic analyses to explore the substrate scope of NocapM. Our findings indicate that NocapM catalyzes hydroxylation of diverse aromatic substrates, although the observed coupling between NADPH oxidation and substrate hydroxylation varies widely from substrate to substrate. Our in-depth biochemical characterization of NocapM provides a solid foundation for future mechanistic studies of this enzyme as well as its utilization as a practical biocatalyst.

Graphical Abstract

Introduction:

Assembly line polyketide synthases (PKSs) are multifunctional enzymes responsible for the biosynthesis of numerous bioactive natural products.^1,2 Much like assembly lines in manufacturing processes, PKSs mediate the stepwise assembly of complex small molecules through a series of enzymatic reactions that are carried out by discrete modules. Altogether, these modules are composed of multiple catalytic domains that mediate the elongation and modification of the nascent polyketide chain by a ketide unit.^3,4 Our laboratory is particularly interested in an orphan PKS exclusively found in strains of the actinomycete Nocardia that we designated as the NOCardiosis-Associated Polyketide (NOCAP) synthase.^5–7 Notably, most of these Nocardia strains have been classified as opportunistic human pathogens and causative agents of nocardiosis, a potentially life-threatening infectious disorder that affects immunosuppressive patients.^8,9 Although nocardiosis is commonly systemic to the pulmonary system, the disease can spread via the bloodstream to other parts of the human body such as the brain, skin, and spinal cord.^10,11

Our prior phylogenetic studies of the NOCAP cluster suggested a recent evolutionary origin of the biosynthetic pathway given that 40 strains harboring NOCAP were clustered in three distinct clades plausibly due to an added fitness provided by the natural product in human hosts.¹² Remarkably, 34 of the Nocardia strains were isolated from patients diagnosed with nocardiosis, while the remaining 6 strains are potentially also capable of infecting humans due to their clustering adjacent to NOCAP-positive pathogens in phylogenetic trees of Nocardia species with sequenced genomes.

The NOCAP cluster consists of sixteen genes, including four PKS-encoding genes and twelve non-PKS-encoding genes that putatively encode sugar biosynthesis and transfer enzymes.^5,6 Our recent efforts in refactoring the entire NOCAP cluster in E. coli revealed the terminal NOCAP glycolipid product (1C) as a fully substituted benzaldehyde headgroup linked to a polyfunctional alkyl chain and an O-linked disaccharide consisting of 3-α-epimycarose and 2-O-methyl-α-rhamnose (Figure 1).¹²

Figure 1.

Biosynthesis of the fully decorated NOCAP product (1C). Proposed biosynthetic pathway of 1C based on our previously reported pathway refactoring in E. coli.¹² Abbreviations: KS-ketosynthase, ACP-acyl carrier protein, AT-acyltransferase, KR-ketoreductase, DH-dehydratase, ER-enoylreductase, MT-methyltransferase, TR-thioester reductase, TEII-thioesterase II.

A key biosynthetic step that links the NOCAP PKS product (1) and incorporation of the disaccharide unit involves meta hydroxylation catalyzed by a flavin monooxygenase (FMO), NocapM (Figure 1). FMOs represent a large family of enzymes that are ubiquitous in natural product biosynthesis due to their roles in tailoring small molecules with functional groups that are critical for their bioactive properties.^13–15 FMOs are also important biocatalysts that perform highly specific oxidations on small molecules involved in industrial and biotechnological applications.^13,16 NocapM was shown to catalyze the hydroxylation of PKS products 1/2 to 1B/2B through in vivo heterologous expression (Figure 2).¹²

Figure 2.

In vitro reconstitution of MBP-NocapM with PKS products. a) Extracted ion chromatograms (EICs) showing production of 1B from a biochemical assay containing 1, FAD, NADPH, and NocapM subject to negative controls. Enzymatically synthesized 1B has the same retention time as 1B produced by E. coli BAP1 (pAD2/pCK-KPY285/pCK-KPY259). b) EICs showing production of 2B from a biochemical assay containing 2, FAD, NADPH, and NocapM subject to negative controls. Enzymatically synthesized 2B has the same retention time as 2B produced by E. coli BAP1 (pAD2/pCK-KPY285/pCK-KPY259). c) Steady-state kinetic analysis of NocapM-catalyzed conversion of 1 at room temperature. d) Steady-state kinetic analysis of NocapM-catalyzed conversion of 2 at room temperature.

In this present work, we expressed and purified soluble NocapM and reconstituted its native activity using in vitro biochemical analyses. Inspired by the versatility of flavoenzymes in catalyzing a myriad of biochemical reactions^17–22, we explored its substrate specificity to establish it as a promiscuous flavin monooxygenase capable of recognizing various substrates beyond substituted benzaldehydes, including acetophenones, methyl benzoates, and 1,3-benzodioxole derivatives. Subsequent kinetic analyses were conducted to establish the biocatalytic potential of NocapM in producing various meta-hydroxylated products. Notably, we observed markedly different degrees of reaction coupling between NADPH oxidation and substrate hydroxylation when different benzaldehyde derivatives were tested.

Materials and Methods:

Commercial Materials.

Q5 High-Fidelity PCR Master Mix (New England Biolabs) was used for PCR reactions. Oligonucleotides were obtained from ELIM BIOPHARM. All chemicals used in this work were from Sigma-Aldrich, Fisher Scientific, or Santa Cruz Biotechnology, unless otherwise noted.

Bacterial Strains and Growth Conditions.

Escherichia coli strains were cultivated on lysogeny broth (LB) agar plates and liquid terrific broth (TB). BL21 (DE3) cells were utilized for protein expression, while BAP1 cells were utilized for heterologous expression. Growth media was supplemented with 100 μg/mL of carbenicillin.

Construction of Plasmids for Expression in E. coli.

Individual genes were PCR amplified from genomic DNA (Nocardia araoensis DSM 44729; DSMZ Germany) and cloned into a pET-21 vector (CarbR, C-terminal 6X His Tag) through restriction enzyme digestion (ThermoFisher) and ligation with NEBuilder HiFi DNA Assembly (New England Biolabs). A maltose binding protein tag (MBP) and TEV protease cut site were fused to the N-terminus of NocapM to promote solubility. Stellar competent cells were utilized for routine cloning. The primers and vectors utilized in this study are reported in Table S1 and Table S2, respectively. Plasmids were extracted using a Zyppy Miniprep Kit (Zymo Research) and confirmed by DNA sequencing.

Expression and Purification of Recombinant Proteins.

A general procedure for immobilized metal affinity chromatography (IMAC) was conducted. E. coli BL21(DE3) competent cells were inoculated (1% inoculum ratio) to 1 L of terrific broth (TB) in a shake flask containing 100 μg/mL of carbenicillin. The cells were grown at 37°C at 220 RPM to an OD₆₀₀ of 0.6. The cells were then cooled on ice for 10 min and induced with 120 μM isopropyl-β-D-galactopyranoside (IPTG) for 20 h at 16°C and 220 RPM. Cells were harvested by centrifugation (5,000 x g, 4°C, 15 min), resuspended in 30 mL of lysis buffer (50 mM Tris pH 8.0, 500 mM NaCl, 10 mM imidazole), and lysed by sonication on ice. Cellular debris was removed by centrifugation (25,000 x g, 4°C, 1 h), and the resulting supernatant was filtered with a 0.45 μm filter before batch binding. Ni-NTA resin (Qiagen) was added to the filtrate at 3 mL/L of culture, followed by incubation for 1 h at 4°C. The protein-resin mixture was loaded onto a gravity flow column in which the flowthrough was discarded. The column was then washed with approximately 30 mL of wash buffer (50 mM Tris pH 8.0, 100 mM NaCl, 25 mM imidazole). Proteins were eluted in approximately 15 mL of elution buffer (50 mM Tris pH 8.0, 100 mM NaCl, 250 mM imidazole). The eluate from the Ni-NTA column was filtered and subjected to ion-exchange chromatography using a HiTrap Q HP column (5 mL, Cytiva). The protein was eluted using a linear gradient of 50–1000 mM NaCl in 50 mM Tris pH 8.0. The fractions containing the target protein were concentrated to 10 mg/mL with an Amicon Ultra-4 filter at 4°C. The concentrated protein was further purified using a HiLoad 16/600 Superdex 200 pg column with a mobile phase containing 20 mM Tris pH 8.0 and 150 mM NaCl. Fractions containing NocapM were pooled, concentrated, and glycerol was added to a final concentration of 10% (v/v). NocapM was flash-frozen in liquid nitrogen and stored in 30 μL beads at −80°C. The presence and purity of NocapM was assessed using SDS-PAGE and concentrations were determined using a NanoDrop Ultraviolet (UV)-Visible spectrophotometer (ThermoFisher).

The approximate molecular weight and protein yield is as follows: NocapM (90 kDa, 6 mg/L from TB, NCBI: WP_051020764.1).

Heterologous Production of 1, 2, 1B, and 2B.

The procedure for production of 1, 2, 1B, and 2B was as previously reported.^6,12 A single colony of an E. coli BAP1 strain was used to inoculate an overnight culture of LB broth (10 mL) grown at 37°C in a 50 mL Falcon tube. The BAP1 seed culture (10 mL) was used to inoculate 1 L of Terrific Broth in a shake flask containing 100 μg/mL carbenicillin, 50 μg/mL kanamycin, and 50 μg/mL streptomycin at 37°C at 220 RPM to an OD₆₀₀ of 0.4. The cells were then cooled on ice for 10 min and inoculated with 10 mL of 500 mM sodium malonate pH 7.4, 1 mL of 50 mM calcium pantothenate, and 0.1 mL of 1 M IPTG. The cultures were then agitated at 20°C (220 RPM) for 48 h.

The following combination of plasmids were utilized to make specific strains:

• 1 and 2 production: E. coli BAP1 [pCK-KPY178/pAD2/pCK-KPY259]

• 1B and 2B production: E. coli BAP1 [pCK-KPY285/pAD2/pCK-KPY259]

Detection, Isolation and Purification of 1, 2, 1B, and 2B.

Cells were harvested by centrifugation (5,000 x g, 4°C, 15 min) because these products accumulated in a biomass-associated manner. The analytical verification of 1, 2, 1B, and 2B was conducted by resuspending 250 mg of cells in 0.3 mL of n-hexane and 0.2 mL of isopropanol. The cell mixture was vortexed with glass beads for 30 min. After centrifugation at 15,000 x g for 10 min, the upper organic phases were collected, dried in vacuo, redissolved in methanol, and analyzed by LC-MS. An Agilent Ultivo QQQ LC-MS system or Agilent 6545 Q-TOF LC-MS LC-HRMS system equipped with an Agilent Eclipse Plus C18 column (2.1 x 50 mm, RRHD 1.8 μm) was utilized. A water/acetonitrile mobile phase with 0.1% (vol/vol) formic acid with a linear gradient of 2-98% acetonitrile at a flow rate of 0.5 mL/min was utilized.

For isolation and purification, the following HPLC protocols were used: HPLC A: 99.9% (v/v) water, 0.1% (v/v) formic, HPLC B: 99.9% (v/v) acetonitrile, 0.1% (v/v) formic acid. A previously established lipid extraction protocol²³ was adapted in which the cell pellet was resuspended in 65 mL/5 g cell pellet of water, 150 mL/5 g cell pellet of methanol, and 500 mL/5 g cell pellet of methyl tert-butyl ether. The cell mixture was sonicated in a Branson Ultrasonics M1800 ultrasonic cleaning bath for 1 h. Phase separation was induced by addition of a 65 mL/L solution of brine. The upper organic phases were collected, pooled, dried in vacuo, and redissolved in 20 mL of 25% (v/v) methanol in HPLC B. This mixture was diluted to a final volume of 90 mL with HPLC A and equally loaded onto six pre-equilibrated C18 solid-phase extraction cartridges (3M Empore 7 mm/3 mL). The cartridges were each washed with 1 mL of HPLC A, 1 mL of 25/75 HPLC B/HPLC A, 1 mL of 50/50 HPLC B/HPLC A, 1 mL of 75/25 HPLC B/HPLC A, and 1 mL of HPLC B. Fractions identified by LC-MS to contain 1, 2, 1B, or 2B (50/50 HPLC B/HPLC A, 75/25 HPLC B/HPLC A and HPLC B) were pooled, dried in vacuo, redissolved in 35/65 HPLC B/HPLC A, and filtered through a 0.45 μm PFTE membrane. This mixture was further purified with a gradient elution method (35/65 HPLC B/HPLC A to 80/20 HPLC B/HPLC A over 1 h at 2 mL/min) on an Agilent 1260 Infinity LC system equipped with an Agilent Eclipse XDB-C8 column (5 μm, 250 × 9.4 mm). The compounds were lyophilized and stored at −20°C for long-term storage.

NocapM Biochemical Reconstitution.

Immediately before each assay, purified NocapM was incubated with 0.1 mM FAD and run through a pre-equilibrated Bio-Rad Bio-Gel P-6 Gel column (20 mM Tris pH 8.0, 150 mM NaCl) to remove excess FAD. Reactions were performed at room temperature for 3 h (unless otherwise noted) in 50 μL of 100 mM Tricine pH 8.4 (buffered with KOH), 1 mM EDTA, 100 μM primary substrate (1, 2, 3, 5, 7, 9, 11, 13, 15, or 17), 0.5 mM NADPH, and 5 μM NocapM. Primary substrates were dissolved in DMSO to ensure full solubility and assays were maintained at a final concentration of 2% DMSO (vol/vol). After the incubation period, reactions were quenched with two volumes of chilled methanol. The precipitated protein was removed by centrifugation (16,000 x g, 10 min) and the supernatant was subjected to LC-HRMS analysis using an Agilent 6545 Q-TOF LC-MS equipped with an Agilent Eclipse Plus C18 column (2.1 x 50 mm, RRHD 1.8 μm). A water/acetonitrile mobile phase with 0.1% (vol/vol) formic acid with a linear gradient of 2-98% acetonitrile at a flow rate of 0.5 mL/min was utilized. At least three independent replicates were performed for each assay, and representative results are shown. Comparative metabolomics were conducted utilizing a similar workflow as previously reported.²⁴

Analysis of FAD Binding to NocapM.

Purified NocapM (yellow color) after IMAC purification was boiled for 10 min and precipitated protein was removed by centrifugation (15,000 x g, 10 min). The supernatant was analyzed using LC–UV–MS with an Agilent Technologies 6545 Q-TOF LC–MS equipped with an Agilent Eclipse Plus C18 column (2.1 x 50 mm, RRHD 1.8 μm). A water/acetonitrile mobile phase with 0.1% (vol/vol) formic acid with a linear gradient of 2-98% acetonitrile at a flow rate of 0.5 mL/min was utilized. At least three independent replicates were performed for each assay, and representative results are shown. The UV spectrum was monitored at 420 nm and the mass spectrum/retention time was compared to an authentic FAD standard. FAD occupancy was estimated to be 21% by utilizing ε_FAD= 11,300 M⁻¹cm⁻¹ at 450 nm.

Kinetic Analysis of NocapM with its Primary Substrates by Monitoring NADPH Consumption.

Reactions were performed at room temperature in a 96-well plate containing 50 μL of 100 mM Tricine pH 8.4, 1 mM EDTA, 0.5 mM NADPH/NADH, and 1 μM NocapM. The amount of primary substrate was varied, and all assays were performed at a final concentration of 2% DMSO. Oxidation of NADPH/NADH was monitored at 340 nm using a BioTek Synergy HT plate reader to determine reaction velocities. Reported values represent the average velocity and standard deviation, respectively, from triplicate experiments conducted at each substrate concentration. Absorbance values were converted to concentration utilizing ε_NADPH=6,200 M⁻¹cm⁻¹. Kinetic parameters were plotted using GraphPad Prism.

Determination of NocapM Kinetic Parameters Towards Primary Substrates using LC-MS.

Assays were performed in triplicate at room temperature in 50 μL of 100 mM Tricine pH 8.4, 1 mM EDTA, 0.5 mM NADPH/NADH, and 1 μM NocapM. The concentration of primary substrate was varied depending on the kinetic activity (see individual Michaelis-Menten graphs). The incubation times for the reactions were 1 min, 5 min, 10 min, 30 min, and 60 min, which were used to determine the initial velocity of the reaction. After each incubation period, the reactions were quenched with two volumes of chilled methanol. The precipitated protein was removed by centrifugation (15,000 x g, 5 min) and the supernatant was used for analysis. LC-HRMS analysis was performed using an Agilent Technologies 6545 Q-TOF LC-MS equipped with an Agilent Eclipse Plus C18 column (2.1 x 50 mm, RRHD 1.8 μm). A water/acetonitrile mobile phase with 0.1% (vol/vol) formic acid with a linear gradient of 2-98% acetonitrile at a flow rate of 0.5 mL/min was utilized. Product concentration was estimated by constructing a standard curve using authentic standards. Kinetic parameters were determined and plotted using GraphPad Prism 9.

Removal of MBP tag from MBP-NocapM Fusion Protein.

Following IMAC purification, 10,000 units of TEV protease (New England Biolabs) were added to 20 mg of SEC-purified MBP-NocapM and dialyzed overnight using a 10 kDa Slide-A-Lyzer cassette in 2 L of dialysis buffer (50 mM Tris pH 8.0, 1 mM EDTA, 100 mM NaCl) at 4°C. The dialysis buffer was changed twice during the overnight period every 2 h. The next morning, 3 mL of Ni-NTA resin (Qiagen) was added to the dialyzed product, followed by incubation for 1 h at 4°C. The protein-resin mixture was loaded onto a gravity flow column in which the flowthrough was discarded. The column was then washed with approximately 30 mL of wash buffer (50 mM Tris pH 8.0, 100 mM NaCl, 25 mM imidazole). Tagged proteins were eluted in approximately 15 mL of elution buffer (50 mM Tris pH 8.0, 100 mM NaCl, 250 mM imidazole). The eluate from the Ni-NTA column was filtered, concentrated using a 10 kDa Amicon Ultra-4 filter at 4°C, and subjected to size exclusion chromatography using a HiLoad 16/600 Superdex 200 pg column with a mobile phase containing 20 mM Tris pH 8.0 and 150 mM NaCl. Fractions containing NocapM were pooled, concentrated, and glycerol was added to a final concentration of 10% (v/v). NocapM was flash-frozen in liquid nitrogen and stored in 30 μL beads at −80°C. The presence and purity of NocapM was assessed using SDS-PAGE and concentrations were determined using a NanoDrop Ultraviolet (UV)-Visible spectrophotometer (ThermoFisher). The overall MBP cleavage efficiency was estimated to be around 25%.

Protein Used in this Study.

NocapM (Nocardia araoensis DSM 44729, NCBI Accession Number: WP_051020764.1).

Results and Discussion:

Purification of soluble MBP-NocapM.

We attempted to purify N- or C-terminal polyhistidine-tagged variants of NocapM (Table S3); however, poor solubility precluded protein recovery. To enhance expression of soluble protein, we designed a gene encoding a maltose binding protein tag (MBP) fused to the N-terminus of NocapM with a C-terminal polyhistidine tag. In between the MBP tag and nocapM gene, we inserted a TEV protease cleavage site (see pAD9, Table S2). Upon expression and IMAC purification, we obtained a soluble, yellow MBP fusion protein containing flavin adenine dinucleotide (FAD) at 21% occupancy (Figure S1–S2a–e). FMOs typically use FAD as a cofactor to catalyze the incorporation of an oxygen atom from molecular oxygen into substrates with NAD(P)H as the external reductant for the oxidized flavin cofactor.^25,26 Sequence alignment of NocapM with its well-characterized homolog ZvPNO (33%/45% sequence identity/similarity) from the locust Zonocerus variegatus revealed that NocapM possesses conserved FAD and NADPH binding motifs (GXGXXG/A) found in FMOs (Figure S2f).^27–29 After subjecting the fusion protein to size exclusion chromatography, we found that MBP-NocapM is a monomer in solution based on a standard curve with gel filtration standards (Figure S2g).

Biochemical reconstitution of MBP-NocapM.

To reconstitute the biochemical activity of MBP-NocapM, we sought to isolate the NOCAP PKS products (1 and 2) by using E. coli as a heterologous host for scalable polyketide biosynthesis.^6,30 We engineered three different plasmids containing compatible antibiotic resistance markers and origins of replication to refactor the PKS pathway (Table S3). The first plasmid, pAD2, encodes Modules 1 and 2 as a bimodular protein and Module 3 of NocapA with orthogonal docking domains (Figure 1, Figure S3a). The second plasmid, pCK-KPY-259, encodes Module 4-KS₅ and intact NocapB (DH₅-Module 8) (Figure S3a). Lastly, pCK-KPY-178 encodes tAT-TEII, MBP-Module L, MatB, and MatC from Rhizobium leguminosarum (Figure S3a).^6,31,32 Refactored E. coli BAP1 [pAD2/pCK-KPY-259/pCK-KPY-178] produced polyketides 1 and 2 as assessed by liquid chromatography-high resolution mass spectrometry (LC-HRMS), UV, and mass fragmentation (MS/MS fragmentation) (Figure S3b–e).⁶

Polyketides 1 and 2 were isolated from 4L of E. coli BAP1[pAD2/pCK-KPY-259/pCK-KPY-178] and purified as described in the Materials and Methods section, yielding 1 mg/L and 2 mg/L, respectively.⁶ Compounds 1 and 2 were individually utilized as substrates for NocapM assays containing FAD and NADPH. Two new products, putatively identified as 1B and 2B, respectively, were observed in these enzyme assays (Figure 2a–b). To test this hypothesis, we introduced pCK-KPY-285 (a derivative of pCK-KPY-178 containing the nocapM gene) into E. coli BAP1 along with pAD2 and pCK-KPY-259.¹² The resulting strain produced compounds with the same retention time, mass fragmentation, high-resolution mass, and UV as previously identified heterologous products (Figure 2a–b, Figure S4–S6). Steady-state kinetic analysis of the MBP-NocapM catalyzed reactions yielded kinetic parameters for 1 (k_cat= 94.2 ± 5.0 min⁻¹, K_M= 5.9 ± 1.3 μM) and 2 (k_cat= 84.5 ± 5.4 min⁻¹, K_M= 7.0 ± 1.9 μM) (Figure 2c–d).

NocapM is a promiscuous FMO.

FMOs are frequently involved in a wide array of biological processes as evidenced with their roles in catabolism of natural and anthropogenic compounds, while some aid in the biosynthesis of primary and secondary metabolites such as vitamins, hormones, antibiotics, and other small molecules.^33,34 A key challenge in synthetic chemistry involves regioselective functionalization of C–H bonds in complex molecules.^35–40 Nature cleverly employs strategies to catalyze C–H bond functionalization by using metalloenzymes, such as cytochrome P450 monooxygenases (CYPs) and non-heme iron(II) dioxygenases, or FMOs.^41–46 To probe the substrate scope of NocapM, we tested a panel of 12 additional substrates varying in substitutions in the benzene ring and aldehyde functionality (Figure 3). All biochemical reactions were monitored by LC-UV-HRMS and subjected to both targeted hydroxylated species detection, and an untargeted product search via comparative metabolomics to obtain insight into the substrate specificity of NocapM.

Figure 3.

Biochemical analysis of MBP-NocapM assays with multiple substrates. EICs are provided for reaction products from enzymatic assays with the substrate pool in the blue box. Each enzymatic assay was conducted in triplicate and representative chromatograms are shown.

Remarkably, NocapM appears to have broad tolerance for benzaldehyde, acetophenone, benzoate ester, and 1,3-benzodioxole substrates with diverse substituents on the aromatic ring (Figure S7–S14). Kinetic parameters were initially estimated for each substrate by monitoring NADPH oxidation (Figure 4, Figure S17). In general, compounds with a salicylaldehyde moiety were the most preferred substrates (Figure 4). The polyfunctional alkyl tails of 1 or 2 were not required for enzymatic activity. In contrast, replacing the aldehyde functionality with a ketone or ester reduced the catalytic efficiency of NocapM by ~10-20 fold (e.g., 5 vs 7, Figure 4). A few tested compounds, such as benzoates, were not recognized as substrates within detectable limits (Figure S16). The precise position of hydroxylation also appears to be variable from substrate to substrate. Based on comparison to authentic standards, we confirmed the production of 4, 14 and 16 from NocapM-catalyzed reactions with 3, 13 and 15, respectively, (Figure S7, S12–S13). Thus, NocapM appears to preferentially hydroxylate the C3 position of benzaldehyde substrates but can also hydroxylate the C5 position of certain substrates with blocked C3 positions.

Figure 4.

Kinetic analysis of MBP-NocapM. a) MBP-NocapM kinetic parameters were determined for each substrate by monitoring NADPH oxidation at 340 nm. b) Catalytic efficiency of MBP-NocapM for various 2,4-dihydroxybenzaldehyde substrates. c) Catalytic efficiency of MBP-NocapM for acetophenone, methyl benzoate, 1,3-benzodioxole, and singly hydroxylated benzaldehyde substrates. The parameters and uncertainty represent the average and standard deviation from three independently performed experiments, respectively.

An especially intriguing reaction product generated by NocapM was 14, given the importance and broad use of catechols in synthetic chemistry, pharmaceuticals, natural products, and fine chemical production.⁴⁷ Catecholamines are an example of hormones and neurotransmitters produced by mammals to modulate the body’s fight or flight response. Biosynthesis of catecholamines proceeds through production of a key catechol intermediate, L-dihydroxyphenylalanine (L-DOPA), that is synthesized from L-tyrosine by the enzyme L-tyrosine hydroxylase, an iron(II) and tetrahydrobiopterin-dependent monooxygenase.^48–50 The ability of NocapM to synthesize 14 highlights a new strategy to produce the catechol group through use of a flavin cofactor.

In terms of cofactor preference, NocapM had a modest preference for NADPH over NADH (Figure 4a, Figure S17).^27,28,34,51 Previous work explored the nicotinamide cofactor promiscuity in FMOs where the authors identified a modification of the phosphate binding region in the NADPH binding site that changes selectivity to NADH over NADPH for the enzymes CFMO and PSFMO.^52,53 However to date, only a handful of FMOs^28,29,51 have been shown to possess broad substrate specificity as exemplified by NocapM in this work. Interestingly, the closest structurally characterized homologs of NocapM are found in mammalian and insect genomes, including the domestic cat Felis catus⁵⁴ (29%/46% sequence identity/similarity), the locust Zonocerus variegatus²⁷ (33%/45% sequence identity/similarity), and human ancestral FMOs1-2⁵⁴ (29-31%/46% sequence identity/similarity).

Elucidating the degree of coupling in NocapM-catalyzed hydroxylation.

FMOs utilize a C4a-hydroperoxy flavin intermediate to promote oxygenation of substrates (Figure 6). Hydrogen peroxide is often released by the C4a-hydroperoxy flavin species, thus resulting in an unproductive catalytic cycle known as uncoupling.^55,16 To quantify the extent to which NADPH consumption was coupled to substrate hydroxylation, we determined the kinetic parameters of MBP-NocapM for substrates 3, 13, and 15 through direct quantification of hydroxylated products via LC-MS, while the parameters for 5 were estimated by quantifying substrate consumption (Figure 5a–b, Figure S17). A comparison between these kinetic parameters to those obtained from monitoring NADPH oxidation highlighted the critical role of the salicylaldehyde functionality in minimizing the uncoupled reaction (Figure 4a, Figure 5a–b, Figure S17). For example, in the presence of salicylaldehyde 13 as a candidate substrate, ~60% of NADPH consumption was coupled to hydroxylated product formation, whereas compound 15, which possesses a hydroxyl group at C3, yields only ~30% coupling. Similarly, tighter coupling was observed in the presence of substrate 3 than 5 (74% vs 30%). Coupling between hydroxylation and NADPH oxidation was independent of substrate concentration across a range of tested concentrations (Figure 5c).

Figure 6.

Proposed catalytic cycle of NocapM. R and R₂ refer to the nucleotide portion of FAD and the alkyl chain present in 1 and 2, respectively.

Figure 5.

Investigating the degree of coupling between NocapM and selected substrates. a) Steady-state kinetic analysis of MBP-NocapM-catalyzed conversion of 3, 5, 13, and 15 at room temperature using LC-MS. Authentic product standards were utilized to quantify production of 4, 14, and 16, while substrate consumption of 5 was monitored to estimate production of 6 (see Methods). b) Table of MBP-NocapM kinetic parameters determined from LC-MS biochemical assays. c) Degree of reaction coupling between NADPH oxidation and hydroxylation from MBP-NocapM biochemical assays with substrates 3, 5, 13, and 15 at room temperature. Substrate concentrations were chosen below and above the K_M, and at saturating conditions for each given substrate. d) Steady-state kinetic analysis of NocapM-catalyzed conversion of 3 at room temperature using LC-MS. The kinetic parameters, activity, and specificity of NocapM were similar to those for NocapM with an MBP tag.

Given the nature of the fusion protein analyzed in this work, we sought to investigate the potential role of the MBP tag in altering the activity, specificity, or oligomeric properties of NocapM. After extensive screening of enzymatic digestion protocols, soluble NocapM was obtained after TEV protease digestion and size exclusion chromatography to separate uncleaved protein from free NocapM (see Methods, Figure S18). NocapM thus eluted as a monomer on a gel filtration column, suggesting that the MBP tag did not influence the oligomerization state of NocapM (Figure S18b). The kinetic parameters of NocapM lacking the MBP tag also remained nearly unchanged (Figure 5a, Figure 5d).

Proposed molecular mechanism for aromatic hydroxylation by NocapM.

A typical reaction cycle for aromatic hydroxylation by FMOs (Figure 6) is composed of two parts: 1) reductive and 2) oxidative half reactions.^25,56 NocapM is proposed to be part of the Group A FMOs based on its conserved binding motifs and shared role in hydroxylation of aromatic substrates (Figure S2f).³³ In general, Group A members are encoded by a single gene, possess a glutathione reductase (GR-2) Rossman fold involved in FAD binding, and use NADP(H) as an electron donor. The prototypical FMO from Group A is p-hydroxybenzoate hydroxylase⁵⁷ that catalyzes the hydroxylation of 4-hydroxybenzoate to give 3,4-dihydroxybenzoate for which we propose that NocapM follows a similar mechanism for aromatic hydroxylation.³³ For the reductive half-reaction, the aromatic substrate enters the active site to form a reduced enzyme: substrate complex after reduction by NADPH. The reduced enzyme: substrate complex then interacts with molecular oxygen to afford a C4a-hydroperoxyflavin intermediate that is responsible for the hydroxylation step by acting as an electrophile (Figure 6). The last step involves product release from the enzyme and regeneration of the oxidized flavin cofactor. Future work will focus on solving the crystal structure of NocapM to identify key residues in the active site that play a role in substrate binding and catalysis to obtain insight into the molecular mechanism and promiscuous nature associated with NocapM.

Conclusions:

In summary, we reconstituted and investigated the biocatalytic potential of NocapM, an unusual FAD-dependent monooxygenase involved in the biosynthesis of the NOCAP glycolipid in many species of Nocardia. The results presented in this work expand our limited knowledge of FMOs involved in natural product biosynthetic pathways from human pathogens. To our knowledge, only two other FMOs have been biochemically characterized from Nocardia spp., including the N-hydroxylase, NbtG⁵⁸, and RIFMO⁵⁹ from Nocardia farcinica, which are involved in siderophore biosynthesis and rifampicin degradation, respectively.¹⁶ The biochemical and kinetic investigations in this work serve as a stepping stone for future studies of NocapM as a potential drug target given its presence in NOCAP-positive pathogens and precedent studies of FMOs as promising drug targets.¹⁶

Our preliminary substrate scope and kinetic analyses revealed that NocapM can catalyze the hydroxylation of various substituted benzaldehyde, acetophenone, methyl benzoate, and 1,3-benzodioxane substrates, although the coupling efficiency between NADPH oxidation and substrate hydroxylation varies significantly (25-75%). This work paves the way for future mechanistic interrogation, enzyme engineering, and biocatalytic applications of NocapM in generating hydroxylated benzylic species based on its highly promiscuous activity.

ABBREVIATIONS

LC-HRMS

liquid chromatography-high resolution mass spectrometry

EIC

extracted ion chromatogram

PK

polyketide

PKS

polyketide synthase

FMO

flavin monooxygenase

NOCAP

NOCardiosis Associated Polyketide

FAD

flavin adenine dinucleotide

NADPH

nicotinamide adenine dinucleotide phosphate