Progress updates on the toxin assays for the “Human-derived gnotobiotic murine models of CDI”. The paper title is a work in progress! Original Post 6/18/15

Initial Round

Background

The initial round of toxin assays was performed last week Friday by Jhansi Leslie in Vince Young’s lab. She used the vero cell rounding assay which has been previously used by their lab and others as a way of measuring toxin activity levels. She used the supernatant from the stool or cecal samples plated to quantify C. difficile and sequenced to characterize the microbiota. The samples tested were from Day 2 post challenge, which was a day when many of the severely infected mice ended up dying. I expected high values of toxin activity on these days for these mice.

Results

The initial round resulted in overall low or undetectable levels of toxin activity. Looking at the level of toxin production measured in vitro in the paper by Paul E. Carlson “The relationship between phenotype, ribotype, and clinical disease in human Clostridium difficile isolates”, Table S1, this clinical strain of C. difficile, strain “431”, has lower toxin production (3.053 ng/ml) compared with the commonly used lab VPI strain (7.140 ng/ml). So to some degree we should have expected a lower overall activity level compared with VPI. I think, however, that the overall low levels we are seeing could be due to sample handling/storage issues (freeze-thaw with plating and subsequently DNA extractions, and storage at both -80 and -20 degrees Celsius). These factors may adversely affect toxin activity levels, based on observations by other members of the Young Lab (Casey Theriot). However, these factors would not affect either the plating of C. difficile, which happened immediately after sample collection, or the DNA extractions and subsequent sequencing.

Future Directions

Jhansi is testing biological replicates from Day 2 as well as new samples from Day 3, which also included samples from mice with severe infection that needed to be euthanized. This is happening 6/19/15. Instead of 1:10 dilutions of sample and/or pure toxin positive controls, she is using 1:5 dilutions, which hopefully will yield more refined results since our overall levels seem low. If it works well we will try it out with more samples. If not, then doing these assays for the remaining samples might be a waste of time/energy/resources.

Subsequent Rounds

Results

Since the initial round, Jhansi has been using the 1:5 dilutions. They’ve yielded a better range for these samples. Jhansi also tried to repeat it for some samples from the initial round. From the first 80ish samples, Jhansi plotted the weight loss as a function of amount of toxin in sample and the plot roughly corresponded with what we would expect based on weight loss from baseline. Jhansi has since finished testing all available fecal (mostly, but some cecal) supernatant samples for toxin activity. We also had three other groups of mice which not only were challenged with 431 but also various other strains of spores with different physiological properties (see Paul Carlson’s in vitro work). In total Jhansi tested 328 samples. These included days 1, 2, 3, 4, 7, and 10 for testing. Of the 328 samples, 4 were not sequenced: 2 cecal content samples, 1 mislabeled sample (needs to be checked), 1 C. difficile contaminated sample that accidentally was included.

Future Directions

I’m getting the final results of the remaining samples 7/6/15, completed/counted in last few days. It’s interesting that the spores I used for the germfree experiments, 431, were isolated from a patient with severe CDI, and yet in vitro it’s been shown to have much lower toxin activity compared with VPI or even 630. I wonder what sort of complications this particular patient had that could have contributed to disease severity.