Date of Award

Spring 5-2009

Document Type

Dissertation - Restricted

Degree Name

Doctor of Philosophy (PhD)


Biological Sciences


Canonical Gram-negative bacteria have outer membranes abundant in lipopolysaccharide (LPS). The structure of most LPSs can be discussed in terms of three general regions: the lipid A membrane anchor, the core oligosaccharide, and the distal polysaccharide O antigen. Chemical analysis shows the Rhizobium etli CE3 O antigen to be a fixed-length heteropolymer, with two O-methyl residues that vary according to growth conditions. Prior genetic analysis has identified regions within the genome necessary for O-antigen synthesis in this bacterial strain, and the predicted functions of the open reading frames (ORFs) in these regions can account for nearly all the enzymes thought to be required for the synthesis of the known O-antigen structure. Which genes are required for which predicted steps in the synthesis of the R. etli CE3 O antigen were investigated in this study.

The LPS of mutants defective in these ORFs was analyzed. On high-resolution sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gels, most mutants synthesized truncated LPSs of varying sizes and abundance. The presence or absence of specific O-antigen sugars in these truncated LPSs was investigated by sugar composition analysis. From these analyses, a model for the synthesis of the R. etli CE3 O antigen is proposed. This model defines N-acetyl-quinovosamine as the first sugar of the O antigen and, beginning with this sugar, correlates almost every gene within the Oantigen genetic clusters to a specific function in the synthesis of the O antigen. These functions include: eight predicted glycosyltransferases for the addition of each sugar in the known O-antigen structure; enzymes for synthesis of most of the O-antigen-specific sugar nucleotides, including those for N-acetyl-quinovosamine, fucose, 3-O-methyl-6- deoxytalose, and the variable multiply-O-methylated terminal fucose; and the machinery for transport of the O antigen across the inner membrane.

Investigations were initiated to identify, to isolate, and to biosynthesize CE3 Oantigen intermediates. Particular combinations of genetic defects were constructed to test specific predictions of the O-antigen synthesis model. These mutants coupled with several approaches involving radiolabeling in vivo and in extracts in vitro were used to identify candidates for O-antigen intermediates in R. etli CE3.