Development and in vitro evaluation of an Escherichia coli probiotic able to inhibit the growth of pathogenic Escherichia coli K88

Weaned piglets commonly suffer from gastroenteritis caused by enterotoxigenic Escherichia coli K88. Our aim was to produce E. coli strains that inhibited the growth of E. coli K88 and could be used as a probiotic against postweaning diarrhea. The inclusion criteria for the probiotics were that in addition to being able to inhibit E. coli K88, they also needed to be negative for virulence genes commonly associated with E. coli. A total of 463 E. coli isolates from the cattle rumen, cattle feces, swine feces, and soil were screened against 18 E. coli K88 clinical isolates using an agar diffusion technique. Growth inhibition of the most sensitive K88 indicator strain 2-12 occurred for 121/463 isolates: 96/358 from cattle feces, 0/33 from rumen fluid, 9/35 from swine feces, and 16/37 from soil. Of the 121 positive strains, 71/121 were negative for toxin genes (LT, STa, STb, VT1, and VT2). The 14 most inhibitory strains were screened against a range of substrates to assess the ability to utilize carbohydrates that could be included in the diet to enhance their ability to compete in the gut. Two strains, UM-2 and UM-7, were weak utilizers of starch and inulin. In vitro competition assays between the probiotic strains and E. coli K88 strain 2-12 were conducted with glucose as the only carbon source (minimal medium; MM), MM + 2% starch, or MM + 2% inulin. The UM-2 and UM-7 strains were able to outcompete strain 2-12 when glucose was the only carbon source, indicating that inhibitory activity was produced against 2-12 independent of carbon source. The UM-2 strain outcompeted strain 2-12 in assays in which potato starch or inulin was the only carbon source; the ability of 2-12 to maintain its concentrations in the culture were probably the result of cross feeding of breakdown sugars of starch and inulin that could be utilized by 2-12. In contrast, UM-7 did not grow as well as UM-2 on starch and inulin and 2-12 declined rapidly in successive cultures likely because of the lack of breakdown products of starch and inulin produced by UM-7. We conclude that probiotic E. coli without known toxins and that produce inhibitory activity against E. coli that cause postweaning diarrhea can be produced. In addition, the ability to utilize starch or inulin is an important phenotype because it likely gives the probiotic a competitive advantage in the gut.