Rare sugars are monosaccharides and their derivatives that are rare in nature, such as D-allose, D-erythritol, D-ribulose, and D-psicose. Owing to their beneficial effects on human health, they have good development potential in the food and pharmaceutical industries[1]. D-allose, a diastereomer of D-glucose at the C-3 position, is the aldose form of D-psicose. It is approximately 80 % as sweet as sucrose while low in calories [2]. It is primarily found in plants such as shrubs [3], seaweed [4], and human umbilical cord blood [5]. Besides being low in calories, D-allose has the ability to inhibit the proliferation of various cancer cells [6] and enhance the effects of radiotherapy [7] and chemotherapy [8] in cancer treatment. Other physiological functions of D-allose include lowering hypertension [9], serving as an antioxidant [10], and providing cryoprotection [11].

Currently, D-allose is mainly synthesized via chemical synthesis [12] and biosynthesis [13]. It can be synthesized from D-psicose using enzymes such as L-rhamnose isomerase [14], ribose 5-phosphate isomerase (RPI, EC 5.3.1.6) [15], and glucose-6-phosphate isomerase [6]. Previous studies have engineered to enhance the activity of L-rhamnose isomerase from Bacillus subtilis, improving the conversion rate by 55.73 % [16]. However, L-rhamnose isomerase is a metal-dependent enzyme, and the addition of metal ions in carbohydrate production can have detrimental effects. Therefore, we focused on the directed evolution of non-metal-dependent enzymes for more efficient and stable production.

RPI catalyzes the interconversion of ribulose-5-phosphate and ribose 5-phosphate [17] and is part of a large class of isomerases responsible for catalyzing the interconversion of chemical isomers. Two distinct types of RPI exist: RpiA and RpiB. Studies on the aldose-ketose isomerization catalyzed by RpiB from various microbial sources are shown in Table 1.

Due to the catalytic mechanism of RPI, which follows the “ene glycol mechanism”, metal ions do not enhance catalytic activity for most RPIs [30]. RPIBs in Table 2 have been applied to the production of D-allose. Park et al. [31] successfully heterologously expressed CtRpiB in Escherichia coli. The enzyme is active on D-ribose 5-phosphate, D-psicose, and D-allose but shows no catalytic activity toward other sugars. Yeom et al. [15] mutated the predicted catalytic center of CtRpiB and obtained the R132E mutant, which exhibited a catalytic efficiency 1.5 times higher than CtRpiB for D-psicose.

AtRpiB from A. thermocellus can convert D-psicose to D-allose. However, its stability, optimal temperature and conversion efficiency require improvement. This study aimed to investigate the catalytic properties, stability, and substrate-binding affinity of AtRpiB, with a focus on the AtRpiB-D-psicose complex. We employed a multidimensional mutation design and analysis to gain new insights into the directed evolution of AtRpiB variants.