Different interactions between particles of Escherichia coli capsule bacteriophage 29 and its receptor, the E. coli serotype 29 capsular polysaccharide have been studied. The inactivation of phage 29 (8 x 108 PFU/ml) by isolated host capsular glycan was found to be physiologically insignificant (50% inactivation dose = 100 ug after 1 h at 37 C). No adsorption (<2 x 104 PFU/ug) of the viruses to K29 polysaccharide-coated erythrocytes (at 0 or 37 C) was observed either. The phage particles were, however, found to catalyze the hydrolysis of fl-Dglucosido-(1-3)-Dglucuronic acid bonds (arrow) in the receptor polymer, leading, ultimately, to the formation of a mixture of K29 hexasaccharide (one repeating unit), dodecasaccharide, and octadecasaccharide: pyruvate (ketal) 4 /\ 6 D-Glcp #I l 2 D-Manp al l 4 2)-D-Manp-(1 b3)-D-Glcp-(1 c c 3)-D-GlciJAp-(1 $3)-D-Galp-(l _a t Testing derivatives of K29 polysaccharide, as well as 82 heterologous bacterial (mainly Enteriobacteriaceae) capsular glycans, the viral glycanase was found to be highly specific; in accordance with the host range of phage 29, only one enzymatic cross-reaction (with the Klebsiella K31 polysaccharide) was observed. These and previous results, as well as the electron optical findings of M. E. Bayer and H. Thurow (submitted for publication), are discussed in terms of a unifying mechanism of phage 29-host capsule interaction. We propose that the viruses penetrate the capsules by means of their spike-associated glycanase activity, which leads them along capsular polysaccharide strands to membrane-cell wall adhesions where ejection of the viral genomes occurs. Bacteriophages generally infect a restricted range of related host bacteria only. This specificity is largely, but by no means exclusively,