Quantum-chemical analysis for PK from Actinobacteria Bifidobacterium (BbPK)

The computational model was obtained using AlphaFold [28]. The complex of BbPK with ThDP and substrate was generated with PyMOL [29]. The model contains 218 atoms with a total charge of +1, including the side chains of His64, His553, Glu479, Tyr501, Gln321, Ser440, His142, Gly155, His320, Asn549, Gln546, His548, His97, and the cofactor ThDP. Generally, the glutamate (Glu479) is modeled in protonated state for forming a hydrogen bond with the N1’ atom of ThDP. His97 was considered as the most possible candidate of proton donor for dehydration process, and it was modeled in the doubly protonated state. According to the interaction mode of pocket residues, His64, His553, His320, His142, His548, were modeled in their singly protonated states. The structural models of short-chain ketoses were obtained from the PubChem.

All the calculations were performed using the Gaussian 09 package [30] and the B3LYP method. The 6-31G (d, p) basis set was used for the geometry optimizations, and the electronic energies of the stationary points were refined by single-point calculations with the 6–311++G (2d, 2p) basis set. Solvation energies were calculated with the SMD [31] implicit solvent method and a dielectric constant of ε = 4. Previous studies have shown that the effect of the solvation diminishes rapidly with the size of the active site model, rendering the particular value used for the dielectric constant less critical. The zero-point energy corrections were done at the same level of theory as the geometry optimizations.

As used in the cluster approach [32,33], a number of atoms were kept fixed to their crystallographic positions in the geometry optimizations (indicated by asterisks in the S6 Fig). This coordinate-fixing protocol is very important to avoid large unrealistic movements of the various groups at the active site. This approach deals only with the chemical steps of the enzymatic reactions, implying that the substrate binding and product release are usually not explicitly considered by the calculations. Therefore, an implicit assumption in the current model is that neither of these events are rate- or selectivity-determining.

Molecular docking of substrate to BbPK

The molecular docking tool of GNINA [34] was used to predict the potential interaction of different ligands (DHAP, Eu4P, Xu5P, F6P, D/L-EUS, and DHA) with PK from Actinobacteria Bifidobacterium. A structural model of BbPK was obtained using AlphaFold [28] and a 20-nanosecond molecular dynamics simulation was performed to optimize the side chain conformations. The refined PK structural model was treated as the receptor for docking. The structural models of the ligands and ThDP were obtained from the PubChem. The binding site was determined by the structural alignment to the crystal structure of a known phosphoketolase (PDB ID:3AHE) [15]. The docking models were ranked by GNINA’s CNN pose score. For each ligand, the top 1 docking model was selected and energy minimization was performed by OpenMM7 [35] with the Amber14 [36] force field to optimize the complex model.

Phosphoketolase selection

To investigate the function of PK genes in distant evolutionary branches, we screened and analyzed all potential PKs in the NCBI database. First, we predicted all potential PKs by search against the nonredundant database with the Pfam domain ID PF03894 (hmmscan—cpu 10—domtblout output.txt -E 1e-4 PF03894.hmm NR.fasta) [37]. Second, we retrieved all PKs from KEGG database (https://www.genome.jp/entry/pf:xfp). Third, we made a local blastp search using PKs from NCBI as the query sequences, and PKs from KEGG as the BLAST database. After blastp search, 12,185 potential PKs were screened with 3 standards: the best hit is a D-xylulose 5-phosphate/D-fructose 6-phosphate phosphoketolase (XFP, EC:4.1.2.9 4.1.2.22, KO: K01621), the identity is more than 40, and the align length is more than 600. Fourth, all PKs were classified into 4,101 groups by using OrthoMCL with the amino acid identity more than 90 in a group. For each group, we selected a PK gene, which is closest approximation to supposed optimal sequence, consisting of the highest frequency residues in multiple sequences alignment. Using the similar strategy, 23 PKs were screened based on the standard of the identity of 60.

Protein engineering of BbPK

To obtain full mutations, oligonucleotide primers were designed with the degenerate codon NNK, which cover almost all mutations with only 96 clones. Hence, a total of 3,264 clones were screened against 34 single-site saturation mutation libraries. Each single-site saturation mutant library was generated based on PCR. The PCR product was degraded the template with DpnI restriction endonuclease, and then transformed into E. coli BL21 (DE3) competent cells for library construction. Each colony was incubated in 200 μL of LB medium for 24 h to plateau at 37°C and then transferred to 1 mL of the same medium for protein expression. The cells, which induced by IPTG (isopropyl-β-D-thiogalactopyranoside) and cultured overnight at 16°C, were harvested by centrifugation. The bacterial pellet was washed and resuspended in reaction buffer (50 mM potassium phosphate buffer, 5 mM GALD or 20 mM DHA (pH 7.4)). After 3 h of reaction at 37°C, the supernatant was collected by centrifugation for detection of substrate or product.

For GALD, we determined the activity of the mutants by detecting the reduction of substrate. The detection method was as follows: added 120 μL chromogenic reagent to 60 μL sample and heated at 90°C for 15 min. Subsequently, the residual substrate concentration was measured spectrophotometrically at 650 nm. The chromogenic reagent: 1.5 g diphenylamine was dissolved in 100 mL acetic acid, and then 1.5 mL pure sulfuric acid was added.

For DHA, we screened for highly active mutants by detecting the product formaldehyde. The formaldehyde detection method was as follows: 40 μL sample was mixed with 160 μL chromogenic reagent, and then heated at 60°C for 10 min. Subsequently, formaldehyde production was measured spectrophotometrically at 440 nm, and 100 mL chromogenic reagent (pH 6.0) contains 25 g ammonium acetate, 3 mL acetic acid, and 0.25 mL acetylacetone solution.

Activity assay and kinetic properties of PK and mutants

The standard reaction mixture (100 μL) contained 50 mM potassium phosphate buffer (pH 7.5), 5 mM MgSO₄, 1 mM ThDP, 10 mM GALD (DHA or D-EUS), 1 mM ADP, 0.2 mg mL⁻¹ AckA, 5 U hexokinase, 2.5 U Glucose-6-phosphate dehydrogenase, 1 mM NADP⁺, and 10 mM glucose. Approximately 0.5 mg mL⁻¹ PK was added into the reaction system. The reactions conducted at 37°C. NADPH was detected spectrophotometrically at 340 nm. Enzyme kinetics with GALD (DHA or D-EUS) as substrate were determined in assays with GALD (DHA or D-EUS) concentrations of 0–110 mM. Kinetic parameters kcat and Km were determined by measuring the initial velocities of the enzymic reaction and curve-fitting according to the Michaelis–Menten equation, using GraphPad Prism 5 software.

Bacterial strains and growth condition

Escherichia coli DH5α strain (TransGenTM) was grown at 37°C in LB medium for gene cloning and other DNA manipulations. E. coli BL21 (DE3) (TransGenTM) was grown at 37°C or 16°C in 2YT medium for protein expression. Antibiotics for selection purposes were used with 100 μg mL⁻¹ spectinomycin, 100 μg mL⁻¹ ampicillin, 34 μg mL⁻¹ chloramphenicol, or 100 μg mL⁻¹ kanamycin.

Enzyme activity of BbPK for DHA, L-EUS, and D-EUS

The coding gene of phosphoketolase from Actinobacteria Bifidobacterium was ligated into the expression vector pET-28a via NdeІ and XhoI restriction sites. E. coli BL21(DE3) cells carrying recombinant plasmid were inoculated into 5 mL LB (Luria Broth) medium with Kanamycine (100 μg mL⁻¹) and cultured overnight at 37°C, and then scaled up to 800 mL 2YT medium (16 g L⁻¹ Tryptone, 10 g L⁻¹ yeast extract, 5 g L⁻¹ NaCl) containing Kanamycine (100 μg mL⁻¹). Gene expression was induced by adding IPTG to a final concentration of 0.5 mM when OD₆₀₀ reached 0.6. The cell cultures continued to grow overnight at 16°C before being harvested by centrifugation at 6,000×g and then was resuspended in 50 mL lysis buffer (50 mM potassium phosphate buffer (pH 7.4), 5 mM MgSO₄, 0.5 mM ThDP). The bacterial pellet was lysed by using a high-pressure homogenizer (JNBIO, China), and the cell debris was removed by centrifugation at 10,000×g for 60 min at 4°C. The soluble protein sample was loaded onto a nickel affinity column (GE Healthcare), which was rinsed with 50 mL wash buffer (50 mM potassium phosphate buffer (pH 7.4), 5 mM MgSO₄, 0.5 mM ThDP, and 50 mM imidazole) and then eluted with 20 mL elution buffer (50 mM potassium phosphate buffer (pH 7.4), 5 mM MgSO₄, 0.5 mM ThDP, and 200 mM imidazole). The eluted protein was concentrated and dialyzed against lysis buffer (50 mM potassium phosphate buffer (pH 7.4), 5 mM MgSO₄, and 0.5 mM ThDP) by ultrafiltration with an Amicon Ultra centrifugal filter device (Millipore, USA) with a 30 kDa molecular-weight cutoff. The protein concentration was determined using a BCA Protein Assay Reagent Kit (Pierce, USA) with BSA as the standard.

Activity of PKs on DHA was determined with 1 mg mL⁻¹ purified recombinant protein in 200 μL reaction mixtures. The reaction system comprised 10 mM DHA, 1 mM ThDP, 5 mM MgSO₄, and 50 mM phosphate buffer (pH 7.4). After incubation at 37°C for 0.5 h, the reaction was stopped by adding 200 μL of acetonitrile. Acetyl-phosphate in samples is determined by chromogenic and liquid chromatography, the by-product formaldehyde was confirmed by GC-MS.

Chromogenic detection of acetyl-phosphate (AcP): add 100 μL hydroxylamine hydrochloride solution (2 M, pH 6.5) to 100 μL sample, conduct at 30°C for 10 min. Then add 200 μL of chromogenic solution, which is prepared with 10 mL FeCl₃.6H₂O (FeCl₃.6H₂O powder dissolved in 0.1 M hydrochloric acid, 5% (m/v)), 10 mL trichloroacetic acid (15% (m/v)), and 10 mL HCl (4 M), conduct at 30°C for 5 min, then detect the absorbance at 505 nm.

HPLC detection for AcP: add equal volume of 5% sulfuric acid to the sample for completely decomposing AcP into acetic acid. Acetic acid was then detected by HPLC. HPLC conditions: column, Aminex HPX-87H (Bio-Rad); detection wavelength, 210 nm; mobile phase, 5 mM sulfuric acid; flow rate, 0.6 mL min⁻¹; sample volume, 20 μL; column temperature, 40°C.

Formaldehyde detection by GC-MS. Sample derivatization: 100 μL 2.4-dinitrophenylhydrazine solution was added to 100 μL samples, the mixture was performed at 60°C for 60 min in the dark. Then added 400 μL of n-hexane to the mixed solution, centrifuge at 5,000×g for 2 min. Separated the upper solution and then added appropriate amount of anhydrous sodium sulfate powder, centrifuged at 10,000×g for 10 min. Separated the upper solution and was detected by GC-MS. GC-MS conditions: Electron ionization (EI) GC-MS analyses were performed with a model 7890A GC (Agilent) with a DB-5 fused silica capillary column (30 cm length, 0.25 mm inner diameter, 0.25 μm film thickness) coupled to an Agilent 7200 Q-TOF mass selective detector. Injections were performed by a model 7683B autosampler. The GC oven was programmed from 160°C (held for 1 min) to 240°C at 10°C min⁻¹, and then held for 5 min; the injection port temperature was 250°C, and the transfer line temperature was 280°C. The carrier gas, ultra-high purity helium, flowed at a constant rate of 1.2 mL min⁻¹. For full-scan data acquisition, the MS scanned from 35 to 550 atomic mass units. Data analysis for GC-MS was performed with Mass Hunter software (Agilent, USA) and NIST Database.

The activity of PKs on L-EUS and D-EUS were determined with 1 mg mL⁻¹ purified recombinant protein in 200 μL reaction mixtures. The reaction system comprised 10 mM L-EUS or D-EUS, 1 mM ThDP, 5 mM MgSO₄, and 50 mM phosphate buffer (pH 7.4). After incubation at 37°C for 0.5 h, the reaction was stopped by adding 200 μL of acetonitrile. AcP in samples was detected by chromogenic and liquid chromatography. The by-product glycoaldehyde was confirmed by GC-MS. Chromogenic and liquid chromatography detection of AcP were same as DHA.

Glycoaldehyde was confirmed by GC-MS. Samples were freeze-dried, then 60 μL of 0.2 M PFBOA solution was added and mixed, the reaction was performed at 30°C for 90 min. A total of 300 μL of n-hexane was added and centrifuge at 5,000×g for 2 min. Separated the upper solution and then added appropriate amount of anhydrous sodium sulfate powder, centrifuged at 10,000×g for 10 min. Separated 100 μL upper solution, then 30 μL N-Methyl-N-(trimethylsilyl)-trifluoroacetamide with 1% trimethylchlorosilane was added and mixed, the reaction system conducted at 30°C for 30 min. The product was detected by GC-MS. GC-MS conditions: EI GC-MS analyses were performed with a model 7890A GC (Agilent) with a DB-5 fused silica capillary column (30 cm length, 0.25 mm inner diameter, 0.25 μm film thickness) coupled to an Agilent 7200 Q-TOF mass selective detector. Injections were performed by a model 7683B autosampler. The GC oven was programmed from 60°C (held for 1 min) to 100°C at 5°C min⁻¹, to 300°C at 25°C min⁻¹ and then held for 5 min; the injection port temperature was 250°C, and the transfer line temperature was 280°C. The carrier gas, ultra-high purity helium, flowed at a constant rate of 1.2 mL min⁻¹. For full-scan data acquisition, the MS scanned from 35 to 550 atomic mass units. Data analysis for GC-MS was performed with Mass Hunter software (Agilent, USA) and NIST Database.

Activity of PKs on E4P, D-ETS, D-GCD, and DHAP were determined with 1 mg mL⁻¹ of purified recombinant protein in 200 μL reaction mixtures. The reaction system included substrate (10 mM E4P, D-ETS, D-GCD, or DHAP), 1 mM ThDP, 5 mM MgSO₄, and 50 mM phosphate buffer (pH 7.4). After incubation at 37°C for 0.5 h, the reaction was stopped by adding 200 μL of acetonitrile. AcP in samples was detected by chromogenic and liquid chromatography. Chromogenic and liquid chromatography detection of AcP were same as DHA.

Activity of PKs on erythrulose-4-phosphate (Eu4P) were determined with 1 mg mL⁻¹ purified recombinant protein in 200 μL reaction mixtures. The reaction system included 10 mM Eu4P, 0.5 mg mL⁻¹ RpiB, 1 mM ThDP, 5 mM MgSO₄, and 50 mM phosphate buffer (pH 7.4). After incubation at 37°C for 0.5 h, the reaction was stopped by adding 200 μL of acetonitrile. AcP in samples was detected by chromogenic and liquid chromatography. Chromogenic and liquid chromatography detection of AcP were same as DHA.

Expression, purification, and enzyme kinetics of PKs from different species

The PKs coding genes from different species were ligated into the expression vector pET-28a via NdeІ and XhoI restriction sites. All genes were expressed in BL21 (DE3) and purified on the Ni-NTA column. Large-scale purification (800 mL) typically produced about 50 mg enzyme. The protein concentration was determined using the BCA Protein Assay Reagent Kit (Pierce, USA) with BSA as the standard.

Determination of kinetics of PKs on GALD, DHA, L-EUS, and D-EUS. The standard reaction mixture (200 μL) contained 50 mM potassium phosphate buffer (pH 7.5), 5 mM MgSO₄, 1 mM ThDP, 10 mM GALD (or, DHA, L-EUS, D-EUS), 1 mM ADP, 0.2 mg mL⁻¹ AckA, 1 U hexokinase, 0.5 U glucose-6-phosphate dehydrogenase, 1 mM NADP⁺, and 10 mM glucose. Various PKs (0.25 mg mL⁻¹) from different species were added into the reaction system. The reactions conducted at 37°C. The production of NADPH was detected at 340 nm. Enzyme kinetics with DHA, L-EUS, and D-EUS as substrates were determined in assays with concentrations of 0–110 mM. Kinetic parameters were determined from triplicate experiments using GraphPad Prism 5 (GraphPad Software, USA).

Determination of kinetics of PKs on Xu5P. The standard reaction mixture (200 μL) contained 50 mM potassium phosphate buffer (pH 7.5), 5 mM MgSO₄, 1 mM ThDP, 5 mM Xu5P, 0.05 mg mL⁻¹ TIM, 1 U glycerophosphate dehydrogenase, and 1 mM NAD⁺. Various PKs (0.1 mg mL⁻¹) from different species were added into the reaction system. The reactions conducted at 37°C. The production of NADH was detected at 340 nm.

Determination of kinetics of PKs on F6P. The standard reaction mixture (200 μL) contained 50 mM potassium phosphate buffer (pH 7.5), 5 mM MgSO₄, 1 mM ThDP, 5 mM F6P, 0.3 mg mL⁻¹ erythrose-4-phosphate dehydrogenase, and 1 mM NAD⁺. Various PKs (0.1 mg mL⁻¹) from different species were added into the reaction system. The reactions conducted at 37°C. The production of NADPH was detected at 340 nm.

Expression, purification, and enzyme kinetics of phosphatases from different species

The phosphatases coding genes of different species were ligated into the expression vector pET-28a via NdeІ and XhoI restriction sites. All genes were expressed in BL21 (DE3) and purified on the Ni-NTA column. Large-scale purification (800 mL) typically produced about 5 to 50 mg enzyme. The protein concentration was determined using a BCA Protein Assay Reagent Kit (Pierce, USA) with BSA as the standard.

The kinetics of phosphatases on F6P, Xu5P, E4P, G3P, and DHAP were determined by monitoring the production of fructose, xylulose, ETS, GCD, and DHA by HPLC. The standard reaction mixture (100 μL) contained 50 mM potassium phosphate buffer (pH 7.5), 5 mM F6P (or Xu5P, E4P, G3P, DHAP). Various phosphatases (0.1 mg mL⁻¹) from different species were added into the reaction system. The reactions conducted at 37°C for 0.5 to 2 h, and 100 μL acetonitrile was added to the samples to terminate the reactions, and then analyzed by HPLC.

HPLC condition for GCD or DHA: column, Aminex HPX-87H (Bio-Rad); detection wavelength, 210 nm; mobile phase, 5 mM sulfuric acid; flow rate, 0.6 mL min⁻¹; sample volume, 20 μL; column temperature, 40°C.

HPLC detection for fructose, xylulose, and erythrose. Sample derivatization: Reaction samples (50 μL) were mixed with a solution of O-benzylhydroxylamine hydrochloride (50 μL, 0.14 mmol mL⁻¹) (pyridine: methanol: water = 33:15:2). After incubation at 50°C for 60 min, samples were diluted in methanol (100 μL) and directly analyzed by HPLC chromatography. HPLC condition: column, X-BridgeTM C18, 5 μm, 4.6 × 250 mm column from Waters (Milford, USA); sample volume, 30 μL; solvent system, (A) aqueous trifluoroacetic acid (TFA) (0.1% (v/v) and (B) TFA (0.095% (v/v)) in CH₃CN/H₂O (4:1), gradient elution from 20% to 60% B in 16 min; flow rate, 1 mL min⁻¹; detection wavelength, 215 nm; column temperature, 35°C.

Formaldehyde circulation system verification

Comparison of PK-GALS formaldehyde recycling system and PK-FLS formaldehyde recycling system: The 100 μL reaction system contained 50 mM potassium phosphate buffer (pH 7.5), 5 mM MgSO₄, 2 mg mL⁻¹ PK, 2 mg mL⁻¹ GALS or FLS, 10 mM DHA, and 1 mM ThDP. The reactions conducted at 37°C for 2 h, and 100 μL sulfuric acid (5%) was added to the samples to terminate the reactions. Acetic acid was then detected by HPLC. HPLC conditions: column, Aminex HPX-87H (Bio-Rad); detection wavelength, 210 nm; mobile phase, 5 mM sulfuric acid; flow rate, 0.6 mL min⁻¹; sample volume, 20 μL; column temperature, 40°C.

Acetic acid and DHA detection method is the same as above. Formaldehyde was detected by chromogenic method. Chromogenic detection of formaldehyde is as follows: dilute the sample to the appropriate concentration (0.1 to 1 mM), add 80 μL chromogenic solution to 120 μL of the diluted sample. The reaction conducted at 60°C for 10 min and centrifugal at 12,000×g for 5 min. Take 150 μL to measure the absorbance at 414 nm. Calculate the concentration of formaldehyde in the sample according to the standard curve. Chromogenic solution is prepared as follows: Dissolve 250 g of ammonium acetate in 900 mL of ddH₂O, then add 30 mL of acetic acid and 2.5 mL of acetylacetone, adjust the pH of the solution to 6 with acetic acid, and finally add ddH₂O to a volume of 1 L.

Process analysis of the APK pathway for F6P

The 1 mL reaction system contained 50 mM potassium phosphate buffer (pH 7.5) 5 mM MgSO₄, 2 mg mL⁻¹ PK, 0.5 mg mL⁻¹ EcHAD, 1 mg mL⁻¹ Ps-LRHI, 1 mM ThDP, 10 mM F6P. The reactions conducted at 37°C for 10 h. Samples were taken out every 2 h for analysis. The reactions were terminated by heating at 95°C for 5 min, and then cooled down to 37°C. Added the 5 U alkaline phosphatase, and conducted at 37°C for 4 h, equal volume acetonitrile was used to terminate the reaction. Fructose, D-ETS, D-EUS, GALD, and acetic acid were detected by HPLC.

Sample derivatization: Reaction samples (50 μL) were mixed with a solution of O-benzylhydroxylamine hydrochloride (50 μL, 0.14 mmol mL⁻¹) (pyridine: methanol: water = 33:15:2). After incubation at 50°C for 60 min, samples were diluted in methanol (100 μL) and directly analyzed by HPLC. HPLC conditions: column, X-Bridge TM C18, 5 μm, 4.6 × 250 mm column from Waters (Milford, USA); sample volume, 30 μL; mobile phase, solvent system (A) TFA (0.1% (v/v) and (B) TFA (0.095% (v/v)) in CH₃CN/H₂O (4:1), gradient elution from 20% to 60% B in 16 min; flow rate, 1 mL min⁻¹; detection wavelength, 215 nm; column temperature, 35°C.

Process analysis of the APK pathway for Xu5P

The 1 mL reaction system contained 50 mM potassium phosphate buffer (pH 7.5) 5 mM MgSO₄, 2 mg mL⁻¹ PK, 1 mg mL⁻¹ CpHAD, 0.1 mg mL⁻¹ TIM, 2 mg mL⁻¹ FLS, 1 mM ThDP, and 10 mM Xu5P. The reactions conducted at 37°C for 10 h, and samples were taken out every 2 h for analysis. The reactions were terminated by heating at 95°C for 5 min, then cooled down to 37°C. Added the 5 U alkaline phosphatase, and conducted at 37°C for 4 h, equal volume acetonitrile was used to terminate the reaction. D-xylulose, GCD, DHA, and acetic acid were detected by HPLC. Formaldehyde was detected by chromogenic method as before.

HPLC detection for D-xylulose, GCD and DHA. Sample derivatization: Reaction samples (50 μL) were mixed with a solution of O-benzylhydroxylamine hydrochloride (50 μL, 0.14 mmol mL⁻¹) (pyridine: methanol: water = 33:15:2). After incubation at 50°C for 60 min, samples were diluted in methanol (100 μL) and directly analyzed by HPLC chromatography. HPLC conditions: column, X-BridgeTM C18, 5 μm, 4.6 × 250 mm column from Waters (Milford, USA); sample volume, 30 μL; mobile phase, solvent system (A) TFA (0.1% (v/v) and (B) TFA (0.095% (v/v)) in CH₃CN/H₂O (4:1), gradient elution from 20% to 60% B in 16 min; flow rate, 1 mL min⁻¹; detection wavelength, 215 nm; column temperature, 35°C. HPLC conditions of acetic acid were the same as before.

The APK pathway for the utilization of C1-C4 carbon sources

The APK pathway for D-EUS and GALD. The 200 μL reaction system contained 50 mM potassium phosphate buffer (pH 7.5), 5 mM MgSO₄, 2 mg mL⁻¹ PK, 1 mM ThDP, 10 mM GALD, or 5 mM D-EUS. The reactions conducted at 37°C overnight.

The APK pathway for FALD and DHA. The 200 μL reaction system contained 50 mM potassium phosphate buffer (pH 7.5), 5 mM MgSO₄, 2 mg mL⁻¹ PK, 2 mg mL⁻¹ FLS, 1 mM ThDP, 20 mM FALD, or 10 mM DHA. The reactions conducted at 37°C overnight.