Open Access Research article

Genetic analysis of the roles of agaA, agaI, and agaS genes in the N-acetyl-D-galactosamine and D-galactosamine catabolic pathways in Escherichia coli strains O157:H7 and C

Zonglin Hu, Isha R Patel and Amit Mukherjee*

Author Affiliations

Division of Molecular Biology, Office of Applied Research and Safety Assessment, Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, Laurel, MD 20708, USA

For all author emails, please log on.

BMC Microbiology 2013, 13:94  doi:10.1186/1471-2180-13-94

Published: 1 May 2013



The catabolic pathways of N-acetyl-D-galactosamine (Aga) and D-galactosamine (Gam) in E. coli were proposed from bioinformatic analysis of the aga/gam regulon in E. coli K-12 and later from studies using E. coli C. Of the thirteen genes in this cluster, the roles of agaA, agaI, and agaS predicted to code for Aga-6-P-deacetylase, Gam-6-P deaminase/isomerase, and ketose-aldolase isomerase, respectively, have not been experimentally tested. Here we study their roles in Aga and Gam utilization in E. coli O157:H7 and in E. coli C.


Knockout mutants in agaA, agaI, and agaS were constructed to test their roles in Aga and Gam utilization. Knockout mutants in the N-acetylglucosamine (GlcNAc) pathway genes nagA and nagB coding for GlcNAc-6-P deacetylase and glucosamine-6-P deaminase/isomerase, respectively, and double knockout mutants ΔagaA ΔnagA and ∆agaInagB were also constructed to investigate if there is any interplay of these enzymes between the Aga/Gam and the GlcNAc pathways. It is shown that Aga utilization was unaffected in ΔagaA mutants but ΔagaA ΔnagA mutants were blocked in Aga and GlcNAc utilization. E. coli C ΔnagA could not grow on GlcNAc but could grow when the aga/gam regulon was constitutively expressed. Complementation of ΔagaA ΔnagA mutants with either agaA or nagA resulted in growth on both Aga and GlcNAc. It was also found that ΔagaI, ΔnagB, and ∆agaI ΔnagB mutants were unaffected in utilization of Aga and Gam. Importantly, ΔagaS mutants were blocked in Aga and Gam utilization. Expression analysis of relevant genes in these strains with different genetic backgrounds by real time RT-PCR supported these observations.


Aga utilization was not affected in ΔagaA mutants because nagA was expressed and substituted for agaA. Complementation of ΔagaA ΔnagA mutants with either agaA or nagA also showed that both agaA and nagA can substitute for each other. The ∆agaI, ∆nagB, and ∆agaInagB mutants were not affected in Aga and Gam utilization indicating that neither agaI nor nagB is involved in the deamination and isomerization of Gam-6-P. We propose that agaS codes for Gam-6-P deaminase/isomerase in the Aga/Gam pathway.