Page 60 - Read Online
P. 60

Page 12 of 15                Sasaki et al. Microbiome Res Rep 2023;2:12  https://dx.doi.org/10.20517/mrr.2023.08

               DISCUSSION
               This study characterized MCC10289_0425, a GH39 AAfase from B. pseudocatenulatum exhibiting β-L-Arap
               -(1→3)-L-Ara-releasing activity against gum arabic, larch AGP, and sugar beet arabinan. An amino acid
               sequence comparison between AAfase and GAfase revealed that almost all amino acid residues involved in
                                                                       119
               catalysis and sugar-binding were conserved. However, only Asn , which is located at the -2 subsite, was
                                     116
               replaced by tyrosine (Tyr ) in AAfase. Furthermore, we converted GAfase-like activity to AAfase-like
                                                                    119
               activity by introducing a single amino acid substitution of Asn  to Tyr. In a study by Ichinose et al., GH27
               β-L-arabinopyranosidase from Streptomyces avermitilis (SaArap27A) was converted into α-galactosidase by
               substituting Asn  with Glu . Structural analysis of the SaArap27A-saccharide complex revealed that the
                                       [25]
                             99
               targeted site, Glu , is located near the O  atom of 5-(hydroxymethyl) group of galactose-a group lacking the
                             99
                                                 6
               L-arabinopyranose molecule. α-D-Gal forms a strong hydrogen bond with the Glu -O  atom and reduces
                                                                                         ϵ1
                                                                                      99
               catalytic turnover, suggesting that Glu  is suitable for β-L-arabinopyranosidase activity. In this study, Asn
                                               99
                                                                                                        119
               in GAfase and Tyr  in AAfase recognized the nonreducing ends of  α-galactopyranosyl and  β-L-
                                 116
               arabinopyranosyl disaccharide, respectively. The homologous genes with > 57% amino acid sequence
               identity with AAfase were categorized as Asn- and Tyr-type, exhibiting GAfase-like and AAfase-like
               activities, respectively. Depending on the species, these distinct substrate specificities would provide
               different fitness abilities to respond to diverse plant polysaccharides in the human gut environment. Tyr-
               type AAfase is a strain-specific enzyme with a prevalence of 36% in B. pseudocatenulatum (n = 168; identity
               > 97%, coverage > 99%), 39% in B. dentium (n = 47; identity > 73%, coverage > 99%), and 7.5% in B.
               asteroides (n = 40; identity > 63%, coverage > 95%) strains, according to the National Center for
               Biotechnology Information database. In contrast, Asn-type GAfase has a low prevalence in B. longum (6.8%,
               n = 307) and B. adolescentis (7.0%, n = 57) strains. Although the distribution of α-D-Gal-(1→3)-α-L-Araf-
               and β-L-Arap-(1→3)-α-L-Araf- structures is not common to all AGPs, it may be present in other plant
               polysaccharides and AGPs in addition to gum arabic and larch AGP. Wheat AGP was reported to possess β-
               L-Arap-(1→3)-α-L-Araf- structure at the end of the side chain . However, AAfase exhibited no such
                                                                       [26]
               activity in this study, suggesting that AAfase requires other enzymes for trimming the sugars attached to the
               side chain of wheat AGP. Using mild hydrolysis, β-L-Arap-(1→3)-L-Ara was detected in the soluble
               fractions of green seaweed (Codium fragile) , and the terminal β-L-Arap- structure was found in pectic
                                                     [27]
               arabinan in the roots of marshmallow (Althaea officinalis)  and pigeon pea (Cajanus cajan) . Thus, there
                                                                                             [29]
                                                                [28]
               may be other candidate substrates for AAfase.
               B. pseudocatenulatum is found in human intestines from infants to adults. It was reported to possess the
               transport system for using human milk oligosaccharides  and some GHs that act on plant polysaccharides,
                                                              [30]
                          [10]
               such as xylan . We revealed that the GH39 AAfase acts on larch and gum arabic AGP as a strain-specific
               enzyme in B. pseudocatenulatum. This finding supports the hypothesis that B. pseudocatenulatum can
               survive in the human intestine owing to the presence of various plant polysaccharide-degrading enzymes. In
               vitro assimilation test revealed that B. pseudocatenulatum MCC10289 utilized β-L-Arap-(1→3)-L-Ara as a
               sugar source. These results support the hypothesis that B. pseudocatenulatum MCC10289 can transport β-L-
               Arap-(1→3)-L-Ara released by AAfase into the bacterial cell via an ABC transporter (MCC10289_0427-
               0429)  and  that  β-L-Arap-(1→3)-L-Ara  is  degraded  into  L-arabinose  by  the  intracellular  β-L-
               arabinopyranosidase (MCC10289_0426) and is metabolized as an energy source. However, other AAfase-
               carrier Bifidobacterium strains could not grow even though they possessed this gene cluster for assimilating
               β-L-Arap-(1→3)-L-Ara. The amino acid sequences of the AAfase gene clusters within the AAfase-carrier
               strains are almost identical but have slight differences in the ABC transporter and intracellular GH36
               [Figure 5A]. Although it is possible that the difference may affect the protein conformations, the
               relationship for the assimilation remains unclear. Interestingly, B. kashiwanohense MCC10250 showed good
               growth on larch AGP even though it does not encode the genes for type II AG degradative enzymes, such as
   55   56   57   58   59   60   61   62   63   64   65