81116 is a commonly used laboratory strain that was originally isolated from an outbreak of enteritis in a UK boarding school in 1983 
. Since that time the strain has been widely disseminated and maintained in laboratories around the world. Despite the fact that this strain of C. jejuni
is laboratory adapted and has been subjected to repeated laboratory passage, evidence suggest that the isolate has maintained an amazing level of genetic stability over the last 20 years and that the AFLP pattern determined in 2001 was identical to the co-isolates that had not been passaged since 1983 
. Despite this considerable genome stability, the data presented herein clearly demonstrates the presence of a single nucleotide substitution in the open reading frame of the luxS
gene of 81116AI2- that results in a single amino acid substitution and the loss of AI-2 phenotype. Identification of such a mutation is not precluded by the genetic stability described above since the sensitivity of the AFLP technique for the detection of specific single amino acid changes is extremely low for the portions of the genome not included in the restriction enzyme recognition sites.
Interestingly, two previous studies regarding AI-2 competence of C. jejuni
have utilized isolates of 81116 and have demonstrated it to be AI-2 positive 
. To confirm our results we were able to obtain genomic DNA of the 81116 isolate used by one of those laboratories as well as a second 81116 isolate (81116AI2+) originating from a third laboratory. Our results indicated that 81116AI2- is a variant of strain 81116 and harbors a single nucleotide substitution in luxS.
This mutation results in the substitution of an invariant glycine residue for a much larger charged aspartate residue in LuxS. Based on crystallographic structures of several LuxS enzymes this residue is located at the amino end of β-strand 4 
. In this position it is located within a fraction of an angstrom from both the Cis-84 residue located 2 residues upstream and the Arg-39 residue located on the adjacent β2-strand. Both of these amino acids play important roles in the kinetics of the enzymatic reaction and the stabilization of the enzymatic intermediates 
. We speculate that the substitution with a larger charged amino acid induces steric hindrance and electrostatic repulsion, which induces conformational changes that result in decreased enzymatic kinetics. Specifically, we speculate that the presence of the larger aspartate residue results in either the β4 strand being displaced away from the active site and pulling the Cis-84 residue out of its typical 3 dimensional position or the displacement of the β2 strand towards the active site, resulting in the Arg-39 residue displacement. In either case the relative locations of these two residues would be disrupted, negatively impacting enzymatic function. Given the lack of significant change in circular dichroism spectra and the retention of some enzymatic activity, this conformation change is assumed to be minor and may only result in slight repositioning of some key active site residues.
To our knowledge, while at least one other naturally occurring E. coli luxS
mutant has been described 
, this isolate represents the first description of a naturally occurring luxS
mutant phenotype in C. jejuni
. Previous studies have utilized site directed mutagenesis to explore the amino acids involved in the active site and have demonstrated significant reductions in the enzyme activity associated with mutagenesis of these active site residues 
. Data herein demonstrated that the single amino acid substitution described in this isolate results in significant reductions in enzymatic activity and ultimately the loss of AI-2 phenotype. The circular dichroism spectra of the wildtype and mutant LuxS were nearly identical (), suggesting that the secondary structures of the proteins are largely conserved. While the ability of circular dichroism to accurately predict secondary structure in silico
is limited, its ability to compare two similar proteins secondary structure is much more robust 
Results of both the V. harveyi
bioluminescence and FRET method demonstrate the phenotypic loss of AI-2 in the 81116AI2- variant when compared to 81116AI2+ and other C. jejuni
strains ( and ). In contrast, the in vitro s
-ribosylhomocysteinase enzymatic kinetics demonstrated that relatively low levels of enzymatic activity were still retained in the mutant LuxS. This retained but limited enzymatic ability explains the detection of AI-2 production following in vitro
incubation of concentrated recombinant mutant enzyme with excess of SRH ( and ). Based on the enzyme kinetics of the wildtype protein kcat/Km of 332/M/s and that of the mutant protein at roughly 3.5/M/s we can estimate that the mutant protein has a roughly 100-fold decrease in enzymatic activity. As a result, we believe that the physiological concentrations of AI-2 produced during the in vitro
growth of 81116AI2-, using the media described, fall below the limits of detection of our AI-2 assays (LOD for FRET ~1 uM) resulting in a phenotypic loss of AI-2 activity while maintaining a minimally-functional LuxS enzyme. Such levels would also fall well below the 40 uM concentrations observed in the wildtype strain during growth in the same media, however data regarding the concentration of AI-2 necessary to result in signaling of C. jejuni
is currently unavailable and thus interpretation of the biological meaning of these levels in hampered. The biological role of AI-2 competent strains of C. jejuni
is still unclear. Evidence suggest that luxS
mutants have both altered phenotypes and altered gene expression, however some of these changes are clearly associated with altered metabolism and the role of AI-2 mediated quorum sensing in this species remains uncertain 
. Furthermore, luxS
mutants do demonstrate an altered ability to colonize animal models 
. In the present study the mutant exhibited a decreased motility consistent with previously published findings of luxS
mutants in this species 
The use of the FRET AI-2 assay in this manuscript provides, to our knowledge, the first truly quantifiable data concerning the physiological concentrations of AI-2 produced by C. jejuni grown in in vitro environments. While the ability of the organism to truly “sense” the compound is still unproven, this quantification data does provide much needed information concerning what range of potential concentrations of chemically synthesized AI-2 should be used to mimic natural production potentials. Based on our results we believe that concentrations between 0 and 40 uM are physiologically possible in logarithmic growth in MH media routinely used for C. jejuni growth.
When comparing the results of the in vitro recombinant proteins enzymatic reactions we observe that while no homocysteine can be detected following incubation of SRH with the 1198AI2- enzyme we do see measurable synthesis of AI-2 (). This provides additional evidence of retention of low-level enzymatic activity of the 81116AI2- strain but also provides clues into the biological sensitivity of two of the assays utilized in detection of LuxS enzymatic activity, homocysteine formation using the Ellman reaction and AI-2 activity using the Vibrio harveyi assay. Given that homocysteine will be produced in the in vitro reaction in a stoichiometrically identical concentration to DPD, that then becomes AI-2, the ability to detect low levels of AI-2 in the absence of homocysteine suggest that the lower limit of detection of the Ellman reaction, as performed in this manuscript, is greater then that of the Vibrio bioluminescence assay.
As has been described for other organisms, the absence of a fully functional luxS gene in C. jejuni does result in increased extracellular concentrations of SRH (). The presence of increased SRH concentrations in the naturally occurring mutant strain suggest that despite measurable residual enzymatic activity of the mutant enzyme, the metabolic ability of the organism to recycle SRH into the SAM recycling pathway is hindered. Thus, in addition to the phenotypic loss of AI-2 activity this strain has a phenotypic change in SAM recycling which precludes the ability to attribute any of the phenotypic characteristics observed directly to the loss of quorum sensing.
In conclusion, we have fully described the molecular basis for the loss of AI-2 phenotype associated with a naturally occur luxS mutant of C. jejuni. Given that the amino acid responsible for this loss of function is highly conserved over a broad range of bacterial species, we speculate that homologous mutations in other bacterial species would also result in a loss of AI-2 phenotype. Additionally, the fact that this amino acid mutation allows for the maintenance of relatively normal secondary structure while diminishing enzymatic activity to levels similar to that observed with luxS deletion mutants suggest that this amino acid is critical for normal enzyme function. Use of this isolate may be helpful in further defining the roles of AI-2 in quorum sensing in a wide variety of bacterial species since it does provide a potential enzymatic intermediate between fully functional and enzymatically incompetent.