Characterization of Sorbitol Dehydrogenase SmoS from Sinorhizobium meliloti 1021

Sinorhizobium meliloti 1021 is a Gram-negative alphaproteobacterium with a robust capacity for carbohydrate metabolism. The enzymes that facilitate these reactions assist in the survival of the bacterium across a range of environmental niches, and they may also be suitable for use in industrial processes. SmoS is a dehydrogenase that catalyzes the oxidation of the commonly occurring sugar alcohols sorbitol and galactitol into fructose and tagatose respectively using NAD+ as a cofactor. The main objective of this study is to evaluate SmoS using biochemical techniques. The nucleotide sequence was codon optimized for heterologous expression in E. coli BL21 (DE3) GOLD cells, the protein was subsequently overexpressed and purified. Size exclusion chromatography and X-ray diffraction experiments suggest that SmoS is a tetrameric peptide. SmoS was crystallized to 2.1 Å in the absence of substrate and 2.0 Å in complex with sorbitol. SmoS was characterized kinetically and shown to have a preference for sorbitol despite a higher affinity for galactitol. Computational ligand docking experiments suggest that galactitol oxidation proceeds slowly because tagatose binds the protein in a more energetically favorable complex than fructose, and is retained in the active site for a longer time frame following oxidation which reduces the rate of the reaction. These results supplement the inventory of biomolecules with the potential for industrial applications and enhance our understanding of metabolism in the model organism S. meliloti.


144
In an attempt to uncover the residues involved in substrate binding, SmoS was 145 also crystallized in the presence of sorbitol and a structure determined to 2.0 Å ( Fig.3; 146 Table 1). Consistent with other described Zn-independent SDR enzymes, conserved 147 active site residues Tyr153, Lys157, Ser140 and Asn111 form the active site (Fig. 4A).
148 Residue Asn111 resides on a π-bulge motif formed by an atypical backbone hydrogen 149 bond disrupting helix α4. This deformation allows the backbone carbonyl group of 150 Asn111 to form a hydrogen bond with a water molecule likely to be involved in the formation of a proton relay system similar to what has been described for the  appropriate pH ranges. Activity at the optimal pH was defined as 100%.
196 217 Kinetic analysis revealed that galactitol turnover is much less efficient than 224 sorbitol oxidation (Fig. 6, Table 2). This observation was particularly interesting due to 225 the K M value of galactitol oxidation, which suggested that the enzyme's affinity for 226 galactitol was higher than for sorbitol (Table 2). This led to the hypothesis that tagatose is  leguminosarum 3841 has also been reported to be tetrameric in solution [31]. The data presented clearly shows that SmoS from S. meliloti is present as a tetramer in solution but 284 with a small subset seemingly present as a hexamer or an octamer made up of a dimer of 285 tetramers (Fig. 2). Of note, it appears that both the tetrameric as well as the higher 286 oligomeric forms show sorbitol dehydrogenase activity (Fig. 2). Tetrameric

287
configurations are reported most often and likely represent the majority of SDR protein 288 structures in solution [43].

289
The SmoS-sbt structure shows that the hydroxyl group bonded to C1 of sorbitol 290 associating with catalytic residue Tyr153, and that the structure has a subtle difference 291 from the apo structure in that residues His190 and Trp191 in alpha helix 7 are contorted 292 slightly to accommodate the presence of the substrate (Fig. 4C). As well, residues 293 Asn111, Ser140, Tyr153, and Lys157, which have been proposed to be involved in 294 electron transfer, are too distant from the substrate for catalysis (Fig. 4A).

295
If the positioning of Tyr153 were correct, it would imply that sorbitol should be forming a planar carbonyl carbon [18,34]. We also note that enzymes catalyzing the 300 oxidation of sorbitol into glucose are known as sorbitol oxidase (SOX) proteins [44,45].

301
These enzymes are dissimilar to SDH enzymes of the SDR family [46,47].

302
This anomaly could be due to the absence of NAD + in the binding pocket. NAD + was 303 left out of the crystallization solution because its presence would result in an enzymatic 304 reaction, which would prevent the capture of a substrate-bound complex. However, SDR 305 reactions proceed with the coenzyme binding first and leaving last [48], which may help to explain not only why sorbitol is found in an atypical position, but also why fructose 307 was not found in the active site of the fructose grown crystal structures despite its 308 presence at high concentrations. In addition, modeling of NAD + and sorbitol into the R.

309
sphaeroides predicted direct contact and a sandwiching of the C2 carbon of sorbitol 310 between the active site tyrosine, and the nicotinamide ring. Taken together these may 311 explain the observed structure.

312
Thermal stability of an enzyme can affect its ability to be exploited in industrial 313 processes [8]. It has been proposed that the increased thermal stability of SDH is due to 314 the abundance of proline residues and the proline to glycine ratio in its primary amino 315 acid sequence [25]. Proline is a rigid residue with low configurational entropy due to its 316 pyrrolidine ring hindrance, there are several studies that suggest protein thermostability

320
meliloti has 5 proline residues and the Pro/Gly ratio is 0.2, additionally the position of the 321 residues appears to be conserved, indicating that it's thermostability is likely more similar 322 to RsSDH (Fig. 8).