xga is an important constituent of the hairy regions of certain pectins. Although it has been suggested that an array of enzymes which are able to hydrolyze xga exist in nature, only an exogalacturonase which is able to remove a Xyl-GalA disaccharide from xga has been identified (1
). Because of the availability of xga (19
) as a very specific substrate and the availability of an A. tubingensis
expression library in K. lactis
, we performed the work described in this paper, which was aimed at finding enzymes that degrade xga in an endo fashion.
Expression libraries are most commonly screened by performing plate assays. In this paper we describe a BCA assay for detecting carbohydrolase activity in the supernatants of K. lactis recombinants in which microwell plates were used. The advantage of this method is that less substrate is required than the amount required for plate assays. Furthermore, it is very sensitive and fast.
Two identical K. lactis recombinants from the approximately 3,400 colonies screened exhibited hydrolyzing activity with xga. An analysis of the cDNA insert demonstrated that it codes for a 406-residue polypeptide with a signal sequence consisting of 18 amino acids. A single copy of the gene is present in the A. tubingensis genome. HPLC analysis and MALDI-TOF mass spectrometry of xga degradation by XghA-containing supernatant of the K. lactis recombinant clearly showed that the mechanism of the enzyme was an endo type of mechanism. To our knowledge, this is the first report of a true endo-xylogalacturonase. Since XghA exhibits no activity with polygalacturonic acid, linear araban, soy galactan, or xylan from oat spelt, we concluded that XghA requires a galacturonic acid backbone substituted with xylose.
When mhr-s was incubated with XghA, only two major products were produced, whereas several peaks were formed when xga was incubated with XghA (Fig. ). The difference could have been the result of the different Xyl/GalA ratios; the Xyl/GalA ratio was two to three times higher in mhr-s than in xga obtained from gum tragacanth. Steric hindrance by other subunits of mhr-s could also have been the reason for the difference. Also, HPSEC analysis of mhr-s incubated with XghA revealed a shift to the lower-molecular-weight fraction, which was less significant than it was in the elution profile of xga with XghA. A possible explanation for this is the presence of xga at the ends of the mhr-s chains. The mhr-s degradation data obtained in the HPAEC and HPSEC analyses show that XghA is valuable for determining the structures of complex polysaccharide fractions, like mhr-s.
A search of the EMBL data library for sequences homologous to the XghA sequence revealed that the levels of similarity of XghA to PGs and RHGs were low. The percentages of similarity were approximately the same for both groups of enzymes. However, an analysis of a multiple alignment of conserved domains in these proteins showed that XghA resembled PGs more than RHGs. The putative active site residues aspartate and histidine (present in domains I to III [Fig. ]) are conserved in the PGs and XghA. In RHGs, however, the histidine is replaced by a glycine, and only two aspartic acid residues are conserved. This suggests that the catalytic mechanism of RHGs may be different from the catalytic mechanisms of the PGs and XghA and that PGs and XghA might have similar catalytic mechanisms. This could also correspond to the cleavage sites in the backbones of the respective substrates. RHGs cleave a rhamnosyl-galacturonic acid linkage, whereas the PGs and XghA cleave a galacturonic acid-galacturonic acid linkage.
The positively charged sequence Arg-Ile-Lys present in domain IV of PGs, postulated to play a role in substrate binding, is not fully conserved in XghA and RHGs. This arginine residue is not conserved in either XghA or RHGs, but the substrate for these enzymes is different from the substrate for PGs, which may account for the discrepancy (14
). Although the substrate backbone for both PGs and XghA consists of GalA, XghA requires a substrate with a xylose-substituted backbone.
In conclusion, XghA is the first enzyme that has been reported to hydrolyze xga in an endo fashion. Based on sequence similarities of the active site residues, we propose that the catalytic mechanism of XghA is similar to that of PGs. This possibility must be investigated further, however. XghA should be a useful tool that allows workers to study plant cell wall structures. In practice, XghA can be a valuable component of tailor-made enzyme preparations that are used in fruit juice manufacturing.