The Hajnia alvei strain 121 1 0-specific polysaccharide is composed of 3-amino-N-(~-3'-hydroxybutyryl)-3,6-dideoxy-D-galactose, N-acetyl-D-galactosamine, N-acetyl-D-glucosamine and D-glucose (1 : 1 : 2 : 2). On the basis of sugar and methylation analyses, Smith degradation, and one-and two-dimensional 'H-and I3C-NMR spectroscopy, the polysaccharide was shown to be an 0-acetylated polymer of the repeating hexasaccharide unit, + 2~(4-OAc)Fucp3NAcylj?I + 6~GlcpNAccll --+ (~Glcppl 4 3)4~GalpNAccll 4 3~GlcpNAcBl + 2~Glcppl +, where nFucp3NAcyl= 3-amino-N-(~-3'-hydroxybutyryl)-3,6-dideoxy-~-galactopyranose. The 0-specific polysaccharide showed some microheterogeneity due to incomplete substitution by terminal glucose.Hafnia alvei, a member of Enterobacteriaceae, is an opportunistic pathogen found in some incidences of nosocomial infections [l]. Preliminary chemical characterization of lipopolysaccharides isolated from the H . alvei group has been reported [2]. Recently, we described the structure of 0-specific polysaccharides of four representatives of H. alvei: standard strain ATCC 13337 and strains 1187, 39 and 2 [3-51. As a part of a systematic work, we now report the structural studies on the 0-antigen of H . alvei strain 121 1.The assignment of 'H-NMR signals for the particular sugar residues was performed using two-dimensional (2D) correlation spectroscopy (COSY), 2D relayed coherence transfer (RCT) [6] and one-dimensional (1 D) totally correlated (TOCSY) homonuclear Hartmann-Hahn (HOHAHA) [7, 81 spectroscopy. For sequence and linkage analysis, 2D NOE (NOESY) and rotating frame NOE (ROESY) [9, 101 spectra were recorded. Due to the different substitution pattern of each of the constituent sugar units, these units could be unambiguously identified by comparison of the linkage information obtained from NMR spectroscopy with the data provided by chemical methods and methylation analysis. In this way, we could dispense with the de novo identification of sugar residues by NMR, usually applied in our laboratory [3, 11 -131, which would be difficult in this case in view of seriousCorrespondence to E. Romanowska,