Introduction C. difficileinfection (CDI) is a global health care problem. Recurring infections and increasing antibiotic resistance have complicated treatment of CDI. There is now, therefore, an urgent need for the development of novel non-antibiotic-based therapeutic and preventative strategies. Recent evidence suggests that the availability of iron may be important for C. difficilegrowth. Lactoferrin, a multifunctional iron-binding glycoprotein, exhibits bacteriostatic activity due to its ability to sequester iron, thereby depriving potential pathogens of this essential nutrient. The aim of this research is to evaluate the comparative effects of iron-depleted (1% saturated) bovine apolactoferrin (apo-bLf) and iron-saturated (84% saturated) bovine hololactoferrin (holo-bLf) in a human in vitro gut model predictive of CDI.

Method Two parallel triple-stage chemostat gut models designed to facilitate the formation of intestinal biofilms were primed with human faecal emulsions and spiked with C. difficilespores (∼10⁷ cfu/ml, PCR ribotype 027). Bacterial populations were allowed to equilibrate before simulated CDI was induced by instillation of Clindamycin (33.9 mg/L, four times daily for 7 days). Samples were collected for determination of microflora populations, C. difficilevegetative cells and spores, cytotoxin titres, short chain fatty acids, bile acids, lactoferrin, iron and% iron saturation of Lf at multiple time points.

Results Preliminary analyses indicate that there were no differences in biofilm formation in the two models. Both apo- and holo-bLf appeared to have little effect on microflora populations or C. difficileduring control periods pre-antimicrobial instillation. Two major differences were observed in C. difficilepopulations post-antibiotic: germination was seen 6 days post-Clindamycin instillation in the apo-bLF model, but not until 15 days post clindamycin in the holo-bLf model; no toxin was detected in the holo-bLf model for the duration of the experiment whereas toxin was rapidly detected in the apo-bLf model (1 day post germination).

Conclusion In conclusion, we have shown for the first time that disrupting iron homoeostasis has major effects on C. difficile growth and development in a human gut model. As the clinical effects of C. difficileinfection are mediated through toxin production, it may be that iron toxicity (rather than iron deprivation) has the greater potential as a novel preventative or treatment strategy.

Disclosure of interest T. Monaghan: None Declared, C. Chilton: None Declared, G. Crowther: None Declared, K. Spiewak: None Declared, M. Brindell: None Declared, G. Singh: None Declared, M. Wilcox Grant/ Research Support from: Multiple therapeutic and diagnostic companies, Consultant for: Multiple therapeutic and diagnostic companies.