the shell. During the time course of organelle formation conducted here, no distinct stages of polyhedral organelle formation were noted. In order to examine the packaging of diol dehydratase into the S. enterica organelles, complementation studies were conducted on strain BE87 (lpduCDE, diol dehydratase minus) using a plasmid carrying the diol dehydratase under control of an IPTG inducible promoter. Diol dehydratase was observed to localize to polyhedral organelles in a succinate/PD minimal medium culture that had been induced immediately after transfer from rich medium (prior to polyhedral organelle formation) and also in a mid-log succinate/PD minimal medium culture (subsequent to polyhedral organelle formation). This suggested that diol dehydratase could be packaged into polyhedral organelles after formation. However the possibility that diol dehydratase was incorporated into polyhedral organelles formed during the midlog phase could not be excluded. Clearly, further study will be needed to clarify this issue. Nevertheless, the fact that plasmid-encoded diol dehydratase could be packaged into the polyhedral organelles could be industrially important providing that the presently unknown function of the polyhedral organelles could provide some advantage to the enzymes that it encased. Future studies to elucidate the mechanism by which diol dehydratase is marked for transport to the polyhedral organelles will be a crucial area to pursue.

      The formation of aberrant structures in strain RT818 containing a Mud element insertion somewhere downstream of the pduG gene demonstrates that genes other than pduA and pduB are involved polyhedral organelle formation. Possible candidates are the pduJKNTU genes, which encode putative homologs to proteins required for the formation of carboxysomes. Although the studies conducted here suggest that the pduA, pduB, and