The whole ABO system really hints at just how tricky prediction of activity from amino acid sequence is.
The transferase that puts the A (GalNAc) or B (Galactose) sugar on switches specificity at the slightest hint of a sequence variation.
Look at the two enzymes found in this paper - one was classified as a CBM (i.e. only that it bound sugars) and the other in a family that primarily removes galactose.
As an aside, you have to appreciate this genomic arrangement where the bacteria are helpful enough to put all the genes needed to do one job (in this case break down blood antigen) in one place. That way, if you know what one thing does, you can jump to conclusions about what the others do (e.g. see a recent paper by Crouch et al).
Final point - if you are wondering where the hell we pick up antibodies to the different blood type antigens so early, if I remember correctly, it's because we develop antibodies to bacteria that ALSO have the blood type antigens.
The transferase that puts the A (GalNAc) or B (Galactose) sugar on switches specificity at the slightest hint of a sequence variation.
Look at the two enzymes found in this paper - one was classified as a CBM (i.e. only that it bound sugars) and the other in a family that primarily removes galactose.
As an aside, you have to appreciate this genomic arrangement where the bacteria are helpful enough to put all the genes needed to do one job (in this case break down blood antigen) in one place. That way, if you know what one thing does, you can jump to conclusions about what the others do (e.g. see a recent paper by Crouch et al).
Final point - if you are wondering where the hell we pick up antibodies to the different blood type antigens so early, if I remember correctly, it's because we develop antibodies to bacteria that ALSO have the blood type antigens.
Glycobiology is insanely cool.