Isotope Probing of the UDP‐Apiose/UDP‐Xylose Synthase Reaction: Evidence of a Mechanism via a Coupled Oxidation and Aldol Cleavage

Abstract The C‐branched sugar d‐apiose (Api) is essential for plant cell‐wall development. An enzyme‐catalyzed decarboxylation/pyranoside ring‐contraction reaction leads from UDP‐α‐d‐glucuronic acid (UDP‐GlcA) to the Api precursor UDP‐α‐d‐apiose (UDP‐Api). We examined the mechanism of UDP‐Api/UDP‐α‐d‐xylose synthase (UAXS) with site‐selectively 2H‐labeled and deoxygenated substrates. The analogue UDP‐2‐deoxy‐GlcA, which prevents C‐2/C‐3 aldol cleavage as the plausible initiating step of pyranoside‐to‐furanoside conversion, did not give the corresponding Api product. Kinetic isotope effects (KIEs) support an UAXS mechanism in which substrate oxidation by enzyme‐NAD+ and retro‐aldol sugar ring‐opening occur coupled in a single rate‐limiting step leading to decarboxylation. Rearrangement and ring‐contracting aldol addition in an open‐chain intermediate then give the UDP‐Api aldehyde, which is intercepted via reduction by enzyme‐NADH.

Cells were harvested by centrifugation at 5000 rpm and 4 °C using a Sorvall Evolution RC centrifuge (Thermo Fisher Scientific, Waltham, Massachusetts, USA) equipped with a F10-S rotor and the pellet was stored at -20 °C. For purification, the cells were disrupted by a French Pressure Cell Press (American Instrument Company, Silver Spring, Maryland, USA) and solid parts were removed by centrifugation (16000 g, 4 °C, 60 min) on an Eppendorf 5415 R centrifuge (Eppendorf). The supernatant was filtered through a 1.2 µm Minisart filter (Sartorius Stedim) and loaded on a self-packed Cu 2+ -loaded IMAC sepharose column (GE Healthcare) using a BioLogic DuoFlow liquid chromatograph (Bio-Rad, Vienna, Austria). Elution was done using 50 mM Tris/HCl pH 7.5 buffers containing 5% (v/v) glycerol (buffer A and B) and 400 mM imidazole (buffer B). UAXS started eluting at 120 mM imidazole. All enzymecontaining fractions were pooled and the buffer was exchanged to 50 mM Tris/HCl pH 7.5 containing 5% (v/v) glycerol and 1 mM dithiothreitol using Amicon Ultra-15 centrifugal concentrators (Millipore, Vienna, Austria). Enzyme preparations were stored at -70 °C. The purity of the enzyme was checked with SDS-PAGE (NuPAGE 4-12% Bis-Tris-Gel, Life Technologies, Vienna, Austria) and Silver Staining ( Figure S1). UAXS concentrations were determined by UV spectroscopy (λ = 280 nm) on a DeNovix DS-11+ microvolume spectrophotometer (DeNovix) using a molar extinction coefficient ϵ of 48360 M -1 cm -1 and a molecular mass of 44703 Da (calculated by the ExPASy ProtParam web service).
NaCl was removed from nucleotide sugar preparations using ÄKTA FPLC with a 2 mL sample loop and a Superdex Peptide 10/300 GL size exclusion column (GE Healthcare). Elution was performed with deionized water at a flow rate of 1 mL min -1 . The target compound was detected by UV absorption (λ = 254 nm, Figure S5). Product-containing fractions were collected, pooled and concentrated on the Laborota 4000 at 45 °C and 20 mbar to a final volume of approximately 5-10 mL. Residual H 2 O was removed by lyophilization on a Christ Alpha 1-4 lyophilizer (B. Braun Biotech International, Melsungen, Germany), after which the sugar nucleotide product was obtained as white powder. NMR data are given in Figures S6 -S11.
Purification was done according to substrate 3, however, a 1 mL SuperQ 650M column (Tosoh Bioscience GmbH, Stuttgart, Germany), a flow rate of 3 mL min -1 and the following steps were used: 16 mL NaCl 0 mM, 92 mL NaCl 20 mM, 20 mL NaCl 150 mM, 17 mL NaCl 500 mM, 16 mL NaCl mM ( Figure   S13). NaCl was separated from compound 2-deoxy-3 as described for substrate 3 ( Figure S14). Water was always removed under N 2 flow due to the instability of 2-deoxy-3. NMR and HPLC data are shown in Figures S15 -S18.

UAXS enzymatic reactions and assays
Reactions were performed at 30 °C on a Thermomixer Comfort (Eppendorf, Hamburg, Germany) without agitation. The reaction volume was typically 500 µL. If not stated otherwise, 2.0 mM of substrate were used. The buffer was 50 mM potassium phosphate, pH 8.5. In the case that D 2 O was used, the pD was set as pH meter reading plus 0.4. Purified UAXS (20 µM) was added to start the reaction and a sample was drawn immediately. Further samples were taken in regular intervals. For stopping the reaction, samples were heated to 99 °C for 5 min or mixed with acetonitrile (1:1 ratio). The samples were analyzed by HPLC as described below. Concentrations of substrate stock solutions were always confirmed by UV spectroscopy (λ = 262 nm) on a DeNovix DS-11+ microvolume spectrophotometer (DeNovix, Wilmington, Delaware, USA) using a molar extinction coefficient ε of 10 mM -1 cm -1 , as described in literature. 8 For measuring the pH profile of the activity of UAXS, the phosphate buffer was used in the pH range 6.0 -8.5, 50 mM sodium citrate was used between pH 5.0 and 6.0, and 50 mM Tris/HCl was used between pH 8.5 and 9.5. Reactions were performed identically as described above.
In the case of using UDP-2-deoxy-GlcA as the substrate of UAXS, alkaline phosphatase was added to the reaction (10 U mL -1 ) to convert any UDP present into uridine and phosphate. The UDP originates from spontaneous degradation of the unstable substrate, and it inhibits the enzyme.

High performance liquid chromatography
Samples were analyzed on a Shimadzu Prominence HPLC system (Shimadzu, Korneuburg, Austria) equipped with a 5 µm Kinetex C18 analytical HPLC column (4.6 × 50 mm; Phenomenex, Aschaffenburg, Germany) and a UV detector (λ = 262 nm). Precipitated protein was removed from samples by centrifugation (16000 g, 4 °C, 5 min) on an Eppendorf 5415 R centrifuge (Eppendorf), and after proper dilution, samples were measured using an injection volume of 5 µL, a temperature of 35 °C and a flow rate of 2 mL min -1 . Isocratic elution was performed using 87.5% 40 mM tetra-n-butylammonium bromide in 20 mM potassium phosphate buffer (pH 5.9) and 12.5% acetonitrile. Analysis time was 3.5 min (without UTP) or 4.5 min (with UTP). Authentic standards were used for calibration.

KIEs from intermolecular competition experiments: reactions and sample preparation
Reaction conditions were the same as for standard enzymatic assays, except that approximately 1 mM unlabeled substrate 3 and approximately 1 mM 3-2 H or 4-2 H substrate 3 were mixed to yield the final substrate concentration. After the reaction had proceeded to the desired level (10 -70% conversion), enzymes were removed by ultrafiltration, as described for synthesis of substrate 3. Purification of the remaining substrate or the reaction products was done using an ÄKTA FPLC liquid chromatograph (GE Healthcare) equipped with a 5 mL HiTrap Q HP anion exchange column (GE Healthcare) and a 2 mL sample loop ( Figure S19). A step gradient of 7.5 mL min -1 ammonium formate buffer (buffer A: 20 mM and buffer B: 500 mM; pH 4.2) was used for elution of bound compounds ( Figure S4). Collected 5-mLfractions were lyophilized and subjected to desalting on a Superdex Peptide 10/300 GL column (GE Healthcare), as described above. After lyophilization and dissolution in D 2 O, all fractions were analyzed by NMR-spectroscopy. The isolation procedure was also applied to unreacted substrate and shown not introduce an isotope effect. 9 A Varian (Agilent) INOVA 500-MHz NMR spectrometer (Agilent Technologies, Santa Clara, California, USA) and the VNMRJ 2.2D software were used for all measurements. 1 H NMR spectra (499.98 MHz) were measured on a 5 mm indirect detection PFG-probe, while a 5 mm dual direct detection probe with z-gradients was used for 13 C NMR spectra (125.71 MHz). NMR data were recorded from purified substances or reaction mixtures. In addition, they were also recorded from enzymatic in situ reactions performed at 30 °C in a total volume of 500 µL potassium phosphate buffer (50 mM; pD 8.5) in D 2 O containing 2 mM of the respective substrate and 20 -100 µM UAXS. 1 H NMR spectra were recorded with pre-saturation of the water signal by a shaped pulse in case of in situ experiments. Standard pre-saturation sequence was used: relaxation delay 2 s; 90° proton pulse; 2.048 s acquisition time; spectral width 8 kHz; number of points 32 k. 13 C NMR spectra during in situ experiments were recorded with the following pulse sequence: standard 13 C pulse sequence with 45° carbon pulse, relaxation delay 2 s, Waltz decoupling during acquisition, 2 s acquisition time. Up to 256 scans were accumulated in one measurement. Arrayed spectra were acquired with an array of pre-acquisition delay of 30 min or 60 min.

NMR spectroscopy (including in-situ analysis of enzymatic reactions)
HSQC spectra were measured with 128 scans per increment and adiabatic carbon 180° pulses. HETCOR spectra were recorded with 4 scans per increment and 256 increments. For KIE analysis, a standard proton experiment with relaxation delay of 25 s was used to record 1 H NMR spectra. ACD/NMR Processor Academic Edition 12.0 (Advanced Chemistry Development Inc.) was used for evaluation of spectra.

KIEs from intermolecular competition experiments: NMR analysis and KIE determination
Evaluation of the spectra for determination of the 1 H/ 2 H ratio was done using ACD/Processor Academic Edition NMR Processor Academic Edition 12.0 (Advanced Chemistry Development Inc., Toronto, Canada). Peaks were fitted with a Gauss+Lorentz function and optimized using a Levenberg-Marquadt algorithm, yielding peak areas. By comparison of the area of the isotopically replaced atom with the unaf-fected atoms, the 1 H/ 2 H ratio was determined. The total concentration of substrate (UDP-GlcA) or product (UDP-Xyl; UDP-Api measured as UMP) was measured by HPLC and converted to the individual concentrations of 1 H and 2 H substrate using the ratio determined by NMR spectroscopy. KIEs were then calculated for each point applying the formula KIE = ln(1 -F 1 )/ln(1 -F 2 ) (measuring the residual substrate). F 1 and F 2 correspond to the fractional conversions of the 1 H and 2 H substrate, respectively. The KIE determined thus is on the catalytic efficiency (V max /K m ).

KIEs from direct comparison of reaction rates
Enzymatic reactions were performed under the conditions described above (50 mM potassium phosphate buffer, pH 8.5) using unlabeled 3 or [4-2 H]-3 as the substrate. The substrate concentration was saturating (≥ 1 mM) and varied at 4 concentrations in the range 1 -3 mM. The conversion of substrate and the formation of products were measured by analyzing samples with HPLC at different times between 15 and 200 min. Initial reaction rates (V) were calculated from the linear dependencies of substrate consumed and time used. The KIE was calculated as the ratio of the V with unlabeled and deuterium-labeled substrate. The effect of substrate deuteration on the ratio of products formed from 3 was also analyzed.

Transient formation of enzyme-bound NADH
The UAXS as isolated contains tightly bound NAD + . Judging from absorbance at 340 nm in the purified protein, the enzyme does not contain NADH. External NAD + was not necessary for activity and addition of NAD + (up to 500 µM) did not affect the enzymatic reaction rate. The formation of enzyme-bound NADH was monitored spectrophotometrically at 340 nm (Beckmann Coulter DU 800, Brea, CA, USA) during conversion of 3 under the conditions of the enzymatic assay described above. Using 20 µM UAXS and assuming a molar extinction coefficient of 6.22 mM -1 cm -1 for NADH, reduction of just 5 -   The average value of the KIE is 2.47 ± 0.43.  The average value of the KIE is 1.20 ± 0.03. The measurement at full conversion of 3 serves as a control. The initial isotope Ratio 3-2 H/3-1 H in substrate 3 was 0.736 (see Table S4).