Glycosylation inactivation is one of the important macrolide resistance mechanisms. The accumulated evidences attributed glycosylation inactivation to a glucosylation modification at the inactivation sites of macrolides. Whether other glycosylation modifications lead to macrolides inactivation is unclear. Herein, we demonstrated that varied glycosylation modifications could cause inactivation of midecamycin, a 16-membered macrolide antibiotic used clinically and agriculturally. Specifically, an actinomycetic glycosyltransferase (GT) OleD was selected for its glycodiversification capacity towards midecamycin. OleD was demonstrated to recognize UDP-D-glucose, UDP-D-xylose, UDP-galactose, UDP-rhamnose and UDP-N-acetylglucosamine to yield corresponding midecamycin 2′-O-glycosides, most of which displayed low yields. Protein engineering of OleD was thus performed to improve its conversions towards sugar donors. Q327F was the most favorable variant with seven times the conversion enhancement towards UDP-N-acetylglucosamine. Likewise, Q327A exhibited 30% conversion enhancement towards UDP-D-xylose. Potent biocatalysts for midecamycin glycosylation were thus obtained through protein engineering. Wild OleD, Q327F and Q327A were used as biocatalysts for scale-up preparation of midecamycin 2′-O-glucopyranoside, midecamycin 2′-O-GlcNAc and midecamycin 2′-O-xylopyranoside. In contrast to midecamycin, these midecamycin 2′-O-glycosides displayed no antimicrobial activities. These evidences suggested that besides glucosylation, other glycosylation patterns also could inactivate midecamycin, providing a new inactivation mechanism for midecamycin resistance. Cumulatively, glycosylation inactivation of midecamycin was independent of the type of attached sugar moieties at its inactivation site.