**Background:** Probiotics are live nonpathogenic microorganisms that confer health benefits, but their mechanisms of action remain incompletely understood. Inducible human β-defensins (hBD) are antimicrobial peptides produced by epithelial cells that play an important role in intestinal barrier function. Previous in vitro studies demonstrated that clinically effective probiotics such as E. coli Nissle 1917 induce hBD-2 expression in intestinal epithelial cells. However, whether this occurs in vivo at standard therapeutic doses was unknown. This study aimed to assess whether oral administration of the probiotic E. coli preparation Symbioflor 2 induces hBD-2 secretion in healthy volunteers and to characterize the bacterial strains responsible.
**Methods:** Thirty-one healthy individuals were enrolled; 5 received placebo and 26 received Symbioflor 2 (containing viable commensal E. coli bacteria). Three verum participants were excluded (1 for irregular drug intake, 2 for noncompliance), leaving 23 in the final verum group. Symbioflor 2 was administered for 21 days in escalating doses starting at 2×5 drops daily, increasing to 15 drops daily (1 ml containing 1.5–4.5×10^7 viable bacteria). Stool samples were collected at baseline, after 3 weeks (end of treatment), and in a subset (n=10) at week 12 (9 weeks post-treatment). Fecal hBD-2 protein was measured by ELISA. In vitro, Caco-2 colonic epithelial cells were incubated with heat-killed Symbioflor 2 genotypes (G1, G2, G3), E. coli Nissle 1917 (positive control), and E. coli K12 (negative control). hBD-2 mRNA expression was quantified by real-time RT-PCR. Antimicrobial susceptibility was assessed using radial diffusion assays with hBD-2, hBD-3, and lysozyme.
**Key Results:** After 3 weeks of Symbioflor 2 treatment, fecal hBD-2 peptide increased in 78% of participants, with a mean 3.7-fold increase compared to baseline (P<0.0001). The placebo group showed no significant changes. Remarkably, at week 12 (9 weeks after treatment cessation), fecal hBD-2 remained significantly elevated compared to baseline (P=0.008). In vitro, only one of the three Symbioflor 2 genotypes (G2) induced hBD-2 in Caco-2 cells, with induction comparable to E. coli Nissle 1917 (10- to 15-fold). hBD-2 induction by G2 was time-dependent (peak at 12 hours) and dose-dependent (maximum at optical density 0.3). The other genotypes (G1 and G3) showed no induction at any time point or concentration. Antimicrobial assays revealed that all tested E. coli strains—including probiotic strains that induce hBD-2 (E. coli Nissle, DSM 17252 G2) and those that do not (G1, G3, K12)—were equally susceptible to killing by hBD-2, hBD-3, and lysozyme, with no significant differences between strains.
**Clinical Implications:** This study provides the first in vivo evidence that standard-dose oral probiotic E. coli administration induces antimicrobial peptide secretion in the human intestine. The sustained elevation of hBD-2 for 9 weeks after treatment cessation suggests a durable effect on host defense mechanisms. The finding that defensin-inducing probiotic strains are not resistant to antimicrobial killing raises the possibility of a "suicidal" mechanism, where probiotics stimulate host defenses that also eliminate the probiotic bacteria themselves—potentially explaining why continuous administration is required. The study also reveals strain-specific differences in defensin induction capacity, with only one of three Symbioflor 2 genotypes being active, highlighting the importance of selecting appropriate probiotic strains. Notably, the active G2 strain lacks flagella, indicating that the hBD-2 induction mechanism differs from that of E. coli Nissle 1917 (which uses flagellin), suggesting multiple pathways for probiotic-mediated immune stimulation. These findings support the therapeutic potential of probiotics in conditions characterized by impaired antimicrobial defense, such as Crohn's disease, and provide a mechanistic rationale for probiotic use in inflammatory bowel disease. However, the study was limited to healthy individuals, and further research is needed to confirm these effects in patient populations and to elucidate the specific bacterial factors responsible for defensin induction.