Archives

  • 2018-07
  • 2018-10
  • 2018-11
  • 2019-04
  • 2019-05
  • 2019-06
  • 2019-07
  • 2019-08
  • 2019-09
  • 2019-10
  • 2019-11
  • 2019-12
  • 2020-01
  • 2020-02
  • 2020-03
  • 2020-04
  • 2020-05
  • 2020-06
  • 2020-07
  • 2020-08
  • 2020-09
  • 2020-10
  • 2020-11
  • The GH family includes also xylanases which are not MeGlcA

    2020-11-20

    The GH30 family includes also xylanases which are not MeGlcA-dependent. Most of them are grouped to GH30_7 subfamily that includes XynIV from Trichoderma reesei showing exo- and endo-xylanase activity [11] and XYLD from Bispora sp. MEY-1 [12] having activity towards glucuronoxylan and arabinoxylan, releasing Xyl, Xyl2 and Xyl3 as main hydrolysis products. However, classification of xylanases to GH30_7 and GH30_8 subfamilies does not precisely determine their catalytic properties. XynVI of T. reesei belonging to GH30_7 subfamily is MeGlcA-dependent xylanase [13] while CpXyn30A from Clostridium papyrosolvens belonging to GH30_8 subfamily is not appendage-dependent [14]. There is no doubt that the differences in catalytic properties arise from the topology of the substrate binding site which is not conserved among all GH30 members. The most studied member of GH30_8 subfamily is xylanase A of a phytopathogenic bacterium Erwinia chrysanthemi (EcXyn30A). EcXyn30A was first isolated by Braun and Rodrigues [15]. The corresponding gene cloning [16] was later used for its high-level expression, crystallization and determination of the 3-D structure [17]. The catalytic properties have also been reported [5,18]. Solution of the crystal structure of EcXyn30A complex with aldotetraouronic imidazoline (MeGlcA2Xyl3) as a ligand has revealed the structural features responsible for its unique mode of action [19]. It was found that the MeGlcA residue is tightly coordinated in -2b subsite through several hydrogen bonds involving Trp289, Tyr295, Tyr255 and Ser258 residues. Moreover, Arg293 was shown to form a strong bidentate ionic interaction with the carboxyl group of the ligand. It was calculated that this ionic interaction corresponds to about 36% of the total binding energy of MeGlcA2Xyl3 and can be responsible for a proper positioning of a substrate for its effective binding [19]. The topology of glycone subsites was also confirmed by ligand-bound crystal structure of another GH30_8 glucuronoxylan xylanohydrolase, XynC from Bacillus subtilis [20]. Based on the reduced number of protein–ligand contacts in the −1 and −2a subsites of XynC and the precise coordination of the MeGlcA residue in -2b subsite provided by Arg272, Trp268, Tyr274, Tyr231 and Ser235, the authors suggested the MeGlcA side residue binding mechanism which forces xylan main chain into a proximity of catalytic glutamates for hydrolytic cleavage. The bidentate ionic bond between the carboxyl group of the ligand and Arg293 of the enzyme was found to be the strongest interaction in the EcXyn30A–ligand complex. Therefore, in this study we decided to construct a variant wherein the Arg293 was substituted for Ala. Kinetic constants of the R293A variant on glucuronoxylan and its derivatives missing the free carboxyl group (4-O-methylglucuronoxylan methyl ester and 4-O-methylglucoxylan) were determined and the product formation from polymeric and oligomeric substrates was studied and compared to the wild-type enzyme.
    Materials and methods
    Results
    Discussion The xylanase A from Erwinia chrysanthemi, EcXyn30A, is a well-studied member of subfamily GH30_8. Several papers describing its physico-chemical and catalytic properties, as well as crystal structure have been published [5,[15], [16], [17], [18]]. 3-D structure of the EcXyn30A complex with aldotetraouronic acid (MeGlcA2Xyl3) showed that Arg293 forms a strong bidentate interaction with the carboxyl group of the ligand and the interactions with MeGlcA residue are stronger than a sum of the interactions with the xylopyranosyl residues in the negatively numbered subsites [19]. It was interesting to find out how elimination of the Arg293–MeGlcA carboxylate interaction would affect the enzyme activity. There are two approaches to investigate this issue – either to eliminate carboxyl group of a substrate, or to replace the Arg293. The first approach was tested by Biely et al. [21] and showed that esterification of glucuronoxylan MeGlcA carboxyl group or its reduction resulted in a drop in EcXyn30A specific activity. We have determined kinetic parameters of EcXyn30A on GX, GXE and GXR and confirmed that esterification or reduction of the carboxyl group of the polymeric substrate led to a dramatic decrease in catalytic efficiency of EcXyn30A. It was caused by a combination of an increase in Km, i.e. a lower affinity of the enzyme to the modified substrates, and high decrease in kcat (126-times and 227-times for GXE and GXR, respectively). Urbániková et al. [19] proposed that in order to be efficiently cleaved the substrate must undergo a strong bending mediated by ionic and stacking interactions. Missing ionic interaction can cause the cleavage slowdown as a consequence of inappropriate bending of the modified substrates. To test if the replacement of the Arg293 has the same effect, the R293A variant was constructed and applied on the same substrates. The R293A replacement led to 18-times decrease in catalytic efficiency on GX caused mostly by increased Km. It is only a moderate reduction compared to the drop observed for wild type enzyme caused by esterification or reduction of MeGlcA carboxyl group of GX. The effective binding of the substrate is thus not mediated solely by the Arg293-substrate carboxylate ionic interaction since other interactions in the substrate binding site of the R293A variant are apparently strong enough to ensure GX hydrolysis. Catalytic efficiency of the R293A variant on GXE and GXR was reduced only slightly in comparison to the catalytic efficiency on GX. This would indicate that the recognition of carboxyl group of the substrate, in contrast to native EcXyn30A, does not play a significant role in catalysis of the R293A variant. Analysis of the hydrolysis products formed from GX, GXE and GXR showed, that the R293A variant, similarly to EcXyn30A, produced a series of xylooligosaccharides having a branched residue on the xylose penultimate to the reducing terminus (Fig. 1). This serves as an evidence that all polymeric substrates are bound to both enzymes in the same way. The absence of Arg293 did not cause a change in the substrate specificity which means that there are other non-ionic interactions navigating the substrate side residue into -2b subsite of the enzyme regardless it is MeGlcA or any of its modified forms. Activity of EcXyn30A and the R293A variant on Xyl5 is very similar and markedly lower than on aldopentaouronic acid, showing that a xylooligosaccharide lacking a branch is a poor substrate for both proteins. While the low activity on linear xylooligosaccharides is not influenced by the Arg293 modification, the replacement of Arg293 for Ala caused 15-fold decrease in specific activity on MeGlcA3Xyl4, confirming that elimination of Arg293 interaction with MeGlcA reduces the ability of the enzyme to cleave MeGlcA-substituted xylooligosaccharide.