A Kinetic Photometric Assay for the Quantification of the Open‐Chain Content of Aldoses

Abstract Aldoses exist predominantly in the cyclic hemiacetal form, which is in equilibrium with the open‐chain aldehyde form. The small aldehyde content hampers reactivity when chemistry addresses the carbonyl moiety. This low concentration of the available aldehyde is generally difficult to ascertain. Herein, we demonstrate a new kinetic determination of the (minute) open‐chain content (OCC) of aldoses. This kinetic approach exploits the aldehyde‐selectivity of 2‐aminobenzamidoxime (ABAO), which furnishes a strongly UV‐active adduct. Simple formation curves can be measured in a photometer or plate reader for high‐throughput screening. Under pseudo‐first order kinetics, these curves correlate with a prediction model yielding the relative OCC. The OCCs of all parent aldoses (pentoses and hexoses) were determined referencing against the two tetroses with exceptionally high OCCs and were in very good agreement with literature data. Additionally, the assay was extended towards higher‐carbon sugars with unknown OCC and also applied to rationalise a lack of reactivity observed in a recent synthetic investigation.


General information
All chemicals were used directly from commercial sources and used without further purification. NMR spectra were recorded at 297 K in the solvent indicated with an Avance UltraShield 400 and an Avance III HD 600 spectrometer. All spectra were calibrated to the solvent residual peak. 1 Chemical shifts (δ) and coupling constants (J) were expressed in ppm and Hz, respectively. Assignments are based on 2D-NMR (COSY, HSQC, HMBC). Optical rotation was measured on an Anton Paar MCP 500 at the specified conditions, [α]D values are given in 10 −1 deg cm 2 g −1 . UV/Vis measurements were performed on a platereader Zentyth 3100 from Anthos or on a Shimadzu UV1800 spectrometer equipped with a thermostat at 20 °C. For data analysis the software Graphpad Prism 6 was used.
The mixture was neutralized with HCl (2 N), monitored using pH paper, and concentrated under reduced pressure. The residue was extracted with EtOAc (5× 75ml), until no further product was detected in the organic phase by TLC (LP:EtOAc 1:1). The combined organic layer was washed with brine, dried over Na2SO4 and concentrated. Purification of the crude material was achieved via trituration with DCM/LP (1:1, 300 ml, 60 min) yielding 21.8 g (85%) of product 2, pure according to 1  In analogy to the case of a classic pre-equilibrium approximation, consequently the following rate law for the formation of the product can be formulated for the ABAO reaction: With the initial product concentration being zero (P(t = 0) = 0) the specific term for the integration constant c can be determined: And after substituting Equation S13 into Equation S10: Equation S15 : And finally, as at complete conversion [C]0 = [P]max: With the product being substantially more UV-active at 405 nM absorbtion compared to the starting materials, this equation can be adapted to reflect the measured absorption.

Determination of OCC shown on the example of ribose
In the following the determination of the OCC will be shown on one representative example: ribose. All other sugars were processed in accordance.

Extraction of the term K·k2 from the measured absorption spectra
Below, the measured UV-curve for the ABAO-adduct formation of ribose is depicted, which represents the average of the triplicate measurements after deduction of the equally averaged blank value.

Figure S1: Measured absorption curves of ribose
The datapoints from the measured curve were then fitted using the following general model equation using the program Graphpad Prism 6 (non-linear regression, one-phase association): The obtained values for the Y0, Plateau and α (and Tau = 1/α, half-time = ln(2)/α and Span = Plateau -Y0 as derived values thereof) with their respective standard errors and the coefficient of determination R² are given in the following

Calculation of OCC values from the term K·k2 based on the suitable k2 values
Finally, with the two k2-values at hand, the term k2•K can be separated. For ribose the k2, erythro-value of 8.89 L·mol -1 ·s -1 is applicable, as it has a cis-2,3-configuration and hence belongs to the erythro-family. This gives a K-value of 0.000922.
Transforming the following equation allows to calculate the OCC from the K derived from the model fitting. The numbers for the deduced figures and terms, (K·k2), k2; K and OCC are given in the lower right section next to the plotted and modeled absorption for each respective sugar.

Determination of k2, erythro and k2, threo-values from the measurements of erythrose and threose.
In analogy to the deduction discussed for ribose (section 5), the term k2•K (as described in

Determination of standard deviations for the case of idose
The accuracy of the assay was determined from the triplicates of idose, as this series experiences the greatest inaccuracy. This is due to handling time impacting this measurement the most, as idose is the sugar with the fasted conversion with ABAO investigated herein. For this purpose, each blank was deducted from each of the three curves obtained for idose. The resulting curves are depicted in Figure S2. Following the procedure described in 5.1 and 5.2, an OCC was determined for each of these 9 resulting curves. These OCC values, along with their average, standard deviation and coefficient of variation are depicted in Table S1. The low coefficient of variation, being less than 3%, indicates the high accuracy of the employed method, even for this most effected sugar.  One assumption of the used model described in section 0 was that the adduct formation is significantly slower than the preceding equilibrium step. We chose to support this assumption to be true, even for the slowest case of D-glucose (in the ABAO assay), by measuring the mutarotation and to compare the time it takes for it to complete with the ABAO assay.

Time (h)
Model Glucose which strongly corroborates our assumptions. In addition, any superimposed effects based on mutarotation can also be ignored.  Figure S3 depicts the absorption spectra of the lyxo-series in regular form, whereas for the publication a representation with normalized data was chosen (Figure 3 and Figure S4; obtained by scaling each curve to obtain a maximal absorption of 1). Through this visual aid, the relation between the different reaction rates can be more easily graphically deduced, which is why that representation was chosen for the publication. This way it becomes apparent that the k2 values for 15, 17 and 18 are very similar but distinctively different form 9. Figure S3: Effect of formal terminal chain elongation onto the OCC of sugars as determined by the ABAO assay without normalization.

Figure S4: Effect of formal terminal chain elongation onto the OCC of sugars as determined by the ABAO assay after normalization
Due to the high reactivity of erythrose and threose, those samples had to be measured at a lower concentration to obtain high quality data. We set out to confirm that the k2 values determined at this 0.4 mM concentration can be used for the analysis of the curves of the more regular sugars at 4 mM solution a screening. In this light, a screening varying both the concentration of ribose (as a mediocrely fast sugar) and ABAO was conducted. As depicted in the following Secondly, the spectra of a representative selection of ABAO-adducts are given. Adducts were prepared from a solution of the respective sugar (300 mM) and ABAO (360 mM) in NH4OAc buffer (pH 4.5, 100 mM). After 8-72 h, depending on the sugar, excess ABAO was extracted using Et2O and the aqueous solution was lyophilized. The residue was taken up in D2O. Further purification was omitted due to the instability of the formed product. Signals of the NMR spectra were labeled in accordance with the following scheme.

Threose 8
In order to on the one hand confirm the reported values for threose but also to deduce a most relevant own value for this study, an 1 H-NMR spectrum (600 MHz) was recorded at ~4m M concentration. From it, an OCC of 11.7% was determined from the integrals of the respective H1-proton signals (10.7% of the hydrate and 1.0% from the aldehyde form; 38.2% β-furanoside and 50.1% α-furanoside). Due to the vicinity of the dominant water peak and its broad signal, the OCC was also determined and thus confirmed comparing diagnostic peaks more distant to the residual water peak: The H2-protons of both cyclic forms combined give an intensity of 93.0, while the hydrate's H2-proton has an intensity of 11.2 (=10.7%).

Erythrose 7
Again, in order to, on the one hand confirm the reported values for erythrose but also to deduce an as most relevant as possible own value for this study, NMR spectra of erythrose in D2O were measured. As described in the literature, erythrose forms various dimeric and oligomeric structures in more concentrated solution, 9 which is why both the assay and the NMR were measured at low enough concentrations (0.4 mM for the assay, 4 mM for the NMR), where these forms are practically non-existent. Nonetheless, an impurity not ascribed to any monomeric erythrose-form was found in two commercial and one self-prepared sample of erythrose, which had to be considered due to its overlapping signals in the calculation. Doing so, a viable value for the OCC can be determined from the comparison of H1 intensities: 0.23% aldehyde, 12.6% hydrate 25.5% β-furanoside and 61.6% α-furanoside yield an OCC of 12.8%.
A second hydrate's signal (H2) with a net-integral of ~12.3 confirms, that the former H1-value Figure S5: Standard sugars divided into the erythro-and threo-family