Rapid-dissolving electrospun nanofibers for intra-vaginal antibiotic or probiotic delivery

Abstract

The emergence of probiotics as an alternative and adjunct to antibiotic treatment for microbiological disturbances of the female genitourinary system requires innovative delivery platforms for vaginal applications. This study developed a new, rapid-dissolving form using electrospun polyethylene oxide (PEO) fibers for delivery of antibiotic metronidazole or probiotic Lactobacillus acidophilus, and performed evaluation in vitro and in vivo. Fibers did not generate overt pathophysiology or encourage Gardnerella growth in a mouse vaginal colonization model, inducing no alterations in vaginal mucosa at 24 hr post-administration. PEO-fibers incorporating metronidazole (100 µg MET/mg polymer) effectively prevented and treated Gardnerella infections (∼3- and 2.5-log reduction, respectively, 24 hr post treatment) when administered vaginally. Incorporation of live Lactobacillus acidophilus (10⁷ CFU/mL) demonstrated viable probiotic delivery in vitro by PEO and polyvinyl alcohol (PVA) fibers to inhibit Gardnerella (10⁸ CFU/mL) in bacterial co-cultures (9.9- and 7.0-log reduction, respectively, 24 hr post-inoculation), and in the presence of vaginal epithelial cells (6.9- and 8.0-log reduction, respectively, 16 hr post-inoculation). Administration of Lactobacillus acidophilus in PEO-fibers achieved vaginal colonization in mice similar to colonization observed with free Lactobacillus. acidophilus. These experiments provide proof-of-concept for rapid-dissolving electrospun fibers as a successful platform for intra-vaginal antibiotic or probiotic delivery.

Figure 1.. PEO-fibers (EFs) do not cause overt pathophysiology in the vaginas of estrogenized mice.

(A) Blank PEO-fibers were inserted vaginally into estrogenized C57Bl/6J mice. Immediately after fiber insertion, mice were infected with PBS or Gardnerella “G.v.”. One day (~24 h) later, vaginal washes were collected, and mice were sacrificed to collect vaginas for histology. X-axis denotes days. (B) Vaginas from PBS-infected (top row) or Gardnerella-infected (bottom row) mice were fixed, processed and sections stained with hematoxylin and eosin (H&E). The vaginal epithelium remained protected by a layer of keratinized epithelial cells (typical of estrogenized mice) and the squamous epithelium appeared intact. Further, no robust polymorphonuclear (PMN) cell infiltration into the epithelium or the stroma was observed. In line with previous observations, no PMN recruitment into the epithelium was observed in Gardnerella-infected animals +/− fibers (n=5 per group; Figure 1B). Scale bars = 200 μm. Images are representative of n=10 mice for “No EFs” group, n=10 EFs, n=5 No EFs + Gardnerella and n=5 EFs + Gardnerella. See Figure S3 for additional images. (C) Gardnerella titers in vaginal washes. Each dot represents an individual mouse. Line denotes the geometric mean.

Figure 2.. Metronidazole-containing PEO-fibers (EFs) treat Gardnerella in a mouse vaginal inoculation model.

(A) For the prevention arm of the experiment, at time point zero, estrogenized mice were administered PEO-fibers, with and without MET, and were immediately intravaginally infected with Gardnerella Vaginal washes and tissues were collected 1 day later. (B) In the treatment arm, estrogenized mice were inoculated with Gardnerella at time zero, and 1 day later vaginal washes were collected immediately prior to insertion of fibers with or without MET. Vaginal washes were collected again on day 2. Vaginal wash titers are shown for both (C) prevention and (D) treatment experiments. (E) Gardnerella titers in uterine horn tissue combined from both experimental arms. In panels C-E, each data point represents one mouse. Box blots denote the 25th to 75th percentile, with line at the median, and whiskers at the min and max. For samples from which Gardnerella was not detectable, the detection limit was plotted (dotted lines). * P < 0.05, ** P < 0.01, ns = not significant by Mann-Whitney test.

Figure 3.. PEO- and PVA-fibers L. acidophilus viability, and lactic acid and hydrogen peroxide production.

The concentration of L. acidophilus aliquoted to the polymer solution prior to electrospinning (5 × 10⁷ CFU/mg polymer), was determined via CFU counts, and was considered as 100% viability. High probiotic viability on the order of 1.70 x 10⁷ and 2.26 x 10⁷ CFU/mg fiber was achieved with the incorporation of fresh L. acidophilus in (A) PEO- and (B) PVA-fibers, respectively. In comparison, incorporation of lyophilized bacteria decreased overall viability to 4.40 to 7.20 x 10² CFU/mg and 1.30 to 2.60 x 10⁴ CFU/mg, for PEO- and PVA-fibers, respectively. For the majority of formulations, fibers electrospun with fresh L. acidophilus had significantly higher L. acidophilus viability relative to fibers electrospun with lyophilized L. acidophilus (P ≤ 0.05), demonstrating as much as a 4- to 5-log increase in L. acidophilus viability in fibers that contained fresh L. acidophilus. The only exceptions were for PEO- and PVA-fibers electrospun in MRS and MRS with glycerol, respectively. In comparison, no statistical significance in L. acidophilus viability was observed between otherwise similar PEO- and PVA-fiber formulations, as a function of polymer choice (PEO or PVA) or electrospinning solution (P > 0.05). Production of lactic acid and hydrogen peroxide of PEO- and PVA-fibers is presented in panels (C) and (D), respectively. All values are shown as the mean ± standard deviation from three independent batches of fibers. Statistical significance between experimental groups was calculated by Tukey’s Honest Significant Difference test, is represented by *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 and ****p ≤ 0.0001.

Figure 4.. L. acidophilus incorporated in PEO- and PVA-fibers inhibits Gardnerella viability in bacterial co-culture assays.

(A) L. acidophilus PEO-fibers, (B) L. acidophilus PVA-fibers, or (C) free L. acidophilus were co-cultured at one log less (left) or equal (right) concentration relative to Gardnerella The viability of L. acidophilus or Gardnerella alone, or as a result of co-culture is noted in the legend. Gardnerella viability significantly decreased after administration of both tested concentrations of L. acidophilus fibers or free L. acidophilus. Columns with zero viability (or viability below the limit of plate-based detection) are represented with a star, and viability values are shown as the mean ± standard deviation from three independent batches of fibers or free bacteria. Statistical significance between the viability of free Gardnerella and Gardnerella in competition, as calculated by Tukey’s Honest Significant Difference test, is represented by *P ≤ 0.05 and ****P ≤ 0.0001.

Figure 5.. L. acidophilus incorporated in PEO- and PVA-fibers inhibits Gardnerella growth in the presence of cervical epithelial cells.

(A) L. acidophilus PEO-fibers, (B) L. acidophilus PVA-fibers, or (C) free, planktonically grown L. acidophilus were co-administered with Gardnerella to HeLa cell monolayers. Columns with zero viability (or viability below the limit of plate-based detection) are represented with a star, and viability values are shown as the mean ± standard deviation from three independent batches of fibers or free bacteria. Statistical significance between the viability of free Gardnerella and Gardnerella in competition, as calculated by Tukey’s Honest Significant Difference test, is represented by *P ≤ 0.05, **P ≤ 0.01, and ****P ≤ 0.0001.

Figure 6.. L. acidophilus incorporated in PEO-fibers colonize the mouse vagina as well as free L. acidophilus bacteria.

L. acidophilus (L.a.) titers in vaginal washes (A-B) and vaginal tissue homogenates (C-D) from mice inoculated vaginally either with free L. acidophilus (open circles) or L. acidophilus in PEO-fibers (closed circles). Lines represent the median values.