Heterogeneity and distribution of lipopolysaccharide in the cell wall of a gram-negative marine bacterium.

Lipopolysaccharide (LPS) extracted from Alteromonas haloplanktis 214, variants 1 and 3, separated into three fractions when subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The fractions appeared in the gels as bands which stained for carbohydrate with the periodate-Schiff reagent. Variant 1, a smooth variant of the organism, and variant 3, a rough colonial variant, produced identical banding patterns. Under similar conditions, LPS from Neisseria meningitidis SDIC, Escherichia coli O111:B4, and Salmonella typhimurium LT2 gave rise to one, two, and three bands, respectively. LPS from Pseudomonas aeruginosa (ATCC 9027) failed to stain clearly with the reagent used. The banding pattern obtained with A. haloplanktis LPS was found not to be due to artifacts produced by the extraction or solubilization procedures employed or to the amount of protein associated with the LPS. When Triton X-100 replaced sodium dodecyl sulfate in the electrophoresis system, LPS failed to migrate into the gel. The lipid A but not the degraded polysaccharide fraction obtained by mild acid hydrolysis of the LPS migrated into the gel on electrophoresis. The three carbohydrate-staining bands obtained with A. haloplanktis LPS and referred to as LPS I, II, and III, in order of increasing electrophoretic mobility, were detected in each of the three outer layers of the cell wall of the organism. Estimations from densitometer scans indicated that 17% of the total LPS in the cell was present in the outer membrane, with the remainder divided almost equally between the loosely bound outer layer and the periplasmic space. Of the three fractions, LPS II was present in each of the layers in greatest amounts. Less LPS I and more LPS III were present in the outer membrane than in the periplasmic space. Pulse-labeling studies indicated that LPS I and II may be synthesized independently, whereas LPS III, which appeared only in cells in the stationary phase of growth, may be a degradation product of LPS I.