The Proton Pump Inhibitor Omeprazole Does Not Promote Clostridioides difficile Colonization in a Murine Model

Murine fecal microbiomes were minimally affected by omeprazole treatment.To test whether omeprazole treatment alters the microbiome and promotes susceptibility to CDIs, we gavaged mice with 40 mg of omeprazole per kg of body weight for 7 days before C. difficile challenge (Fig. 1A). A principal coordinate analysis (PCoA) of the Bray-Curtis distances over the initial 7 days of treatment revealed that the bacterial communities of omeprazole-treated mice remained relatively unchanged (Fig. 1B). We observed no significant changes in the relative abundance of those taxa previously shown to respond to PPI treatment throughout the course of the 16-day experiment (Fig. 1C and D; see also Fig. S1 in the supplemental material). We also observed no significant changes in relative abundances at the family and genus level over the course of the experiment for the omeprazole-treated mice (all corrected P values  >  0.36). These results demonstrated that the omeprazole treatment alone had minimal impact on the murine fecal bacterial community after 7 days of pretreatment.

• Open in new tab
• Download powerpoint

FIG 1

Omeprazole treatment had minimal impact on the murine fecal microbiota. (A) Mouse experiment timeline and logistics. The PPI omeprazole was administered throughout the duration of the experiment. Clindamycin was administered intraperitoneally (IP) 1 day before C. difficile challenge on day 0. Stool samples for 16S rRNA sequencing analysis were collected on the days that are labeled (days −7, −5, −3, −1, 0, 1, 2, 3, 4, 5, 7, and 9). C. difficile CFU in the stool was quantified daily through 6 days postinfection by anaerobic culture. (B) Principal coordinate analysis (PCoA) of Bray-Curtis distances from stool samples of mice in the omeprazole treatment group during the initial 7 days of the experiment. Each color represents stool samples from the same mouse, and lines connect sequentially collected samples. (C and D) Relative abundances of families previously associated with PPI use in humans at the start of the experiment (C) and after 7 days of omeprazole treatment (D). Each circle represents the value for an individual mouse. There were no significant differences across treatment groups for any of the identified families in the sequence data at day −7 (all P values  >  0.448) and day 0 (all P values  >  0.137), analyzed by Kruskal-Wallis test with a Benjamini-Hochberg correction for multiple comparisons. In panels C and D, the gray vertical line indicates the limit of detection.

Omeprazole treatment did not promote susceptibility to C. difficile infection in mice.Next, we examined whether omeprazole treatment altered susceptibility to C. difficile infection in mice. After omeprazole or clindamycin treatment, mice were challenged with 10³ C. difficile strain 630 spores. Although C. difficile colonized the clindamycin-treated mice, it did not colonize the omeprazole-treated mice (Fig. 2A). Interestingly, only one cage of mice that received both omeprazole and clindamycin were colonized, while the other cage of mice were resistant (Fig. 2A). The greatest shifts in bacterial communities occurred in the clindamycin-treated mice (Fig. 2B and Fig. S2). Regardless of whether the mice were colonized by C. difficile, all of the mice had cleared C. difficile within 5 days (Fig. 2A), suggesting that omeprazole did not affect the rate of clearance. Our results suggest that omeprazole treatment had no effect on bacterial community resistance to C. difficile colonization in mice. Instead, most of the differences between the three treatment groups appeared to be driven by clindamycin administration (Fig. 2C and Fig. S2). These findings demonstrated that high-dose omeprazole treatment did not promote susceptibility to C. difficile colonization.

• Open in new tab
• Download powerpoint

FIG 2

Omeprazole treatment alone does not promote CDIs in mice. (A) C. difficile CFU/gram of stool measured each day after C. difficile challenge for mice treated with clindamycin, clindamycin plus omeprazole, and omeprazole. The lines represent the mean CFU/gram for each treatment group, while points represent CFU/gram for individual mice within each group. The black dashed line indicates the limit of detection. (B) PCoA of Bray-Curtis distances from stool samples collected after antibiotic treatment (last 9 days of the experiment). The hue of the symbol indicates the treatment day. Symbols represent the C. difficile colonization status of the mice measured 2 days postinfection. Circles represent resistant mice (C. difficile was undetectable in stool samples), while × symbols represent mice that were colonized with C. difficile, although all mice cleared C. difficile within 5 days of infection. Omeprazole-treated fecal samples primarily cluster together throughout the experiment. (C) Genera that vary the most across treatment groups for stool samples collected from mice 2 days postinfection. Data were analyzed by Kruskal-Wallis test, and no P values were significant after Benjamini-Hochberg correction for multiple comparisons (all P values  >  0.092). The gray vertical line indicates the limit of detection.

Experimental procedures. (i) Animals.All mouse experiments were performed with 7- to 12-week-old C57BL/6 male and female mice. Each experimental group of mice was split between two cages with two or three mice housed per cage and male and female mice housed separately. All animal experiments were approved by the University of Michigan Animal Care and Use Committee (IACUC) under protocol number PRO00006983.

(ii) Drug treatments.Omeprazole (Sigma-Aldrich) was prepared in a vehicle solution of 40% polyethylene glycol 400 (Sigma-Aldrich) in phosphate-buffered saline (PBS). Omeprazole was prepared from 20 mg/ml frozen aliquots and diluted to 8 mg/ml prior to gavage. All mice received 40 mg of omeprazole per kg of body weight (a dose previously used in mouse experiments [17]) or vehicle solution once per day through the duration of the experiment with treatment starting 7 days before C. difficile challenge (Fig. 1A). Although the omeprazole dose administered to mice is higher than the recommended dose for humans, omeprazole has a shorter half-life in mice compared to humans (21) and lacks an enteric coating (22). One day prior to C. difficile challenge, two groups of mice received an intraperitoneal injection of 10 mg/kg clindamycin or sterile saline vehicle (11). All drugs were filter sterilized through a 0.22-μm syringe filter before administration to animals.

(iii) C. difficile infection model.Mice were challenged with C. difficile strain 630 7 days after the start of omeprazole treatment and 1 day after clindamycin treatment. Mice were challenged with 10³ spores in ultrapure distilled water as described previously (11). Stool samples were collected for 16S rRNA sequencing or C. difficile CFU quantification throughout the duration of the experiments at the indicated time points (Fig. 1A). Samples for 16S rRNA sequencing were flash frozen in liquid nitrogen and stored at –80°C until DNA extraction, while samples for CFU quantification were transferred into an anaerobic chamber and serially diluted in PBS. Diluted samples were plated on TCCFA (taurocholate- cycloserine-cefoxitin-fructose agar) plates and incubated at 37°C for 24 h under anaerobic conditions to quantify C. difficile CFU.

(iv) 16S rRNA gene sequencing.DNA for 16S rRNA gene sequencing was extracted from 10 to 50 mg fecal pellet from each mouse using the DNeasy Powersoil HTP 96 kit (Qiagen) and an EpMotion 5075 automated pipetting system (Eppendorf). The 16S rRNA sequencing library was prepared as described previously (23). In brief, the ZymoBIOMICS microbial community DNA standard (Zymo, CA, USA) was used as a mock community (24), and water was used as a negative control. The V4 hypervariable region of the 16S rRNA gene was amplified with Accuprime Pfx DNA polymerase (Thermo Fisher Scientific) using previously described custom barcoded primers (23). The 16S rRNA amplicon library was sequenced with the MiSeq system (Illumina). Amplicons were cleaned up and normalized with the SequalPrep normalization plate kit (Thermo Fisher Scientific), and pooled amplicons were quantified with the KAPA library quantification kit (KAPA Biosystems).

(v) 16S rRNA gene sequence analysis.mothur (v1.40.5) was used for all sequence processing steps (25) using a previously published protocol (23). In brief, forward and reverse reads for each sample were combined, and low-quality sequences and chimeras were removed. Duplicate sequences were merged, before taxonomy assignment using a modified version (v16) of the Ribosomal Database Project reference database (v11.5) with an 80% cutoff. Operational taxonomic units (OTUs) were assigned with the opticlust clustering algorithm using a 97% similarity threshold. To adjust for uneven sequencing across samples, all samples were rarefied to 3,000 sequences 1,000 times. PCoAs were generated based on Bray-Curtis distance. R (v.3.5.1) was used to generate figures and perform statistical analysis.

(vi) Statistical analysis.To test for differences in relative abundances in families and genera across our three different treatment groups at different time points (clindamycin, clindamycin plus omeprazole, and omeprazole on days −7, 0, 2, and 9) or within the omeprazole treatment group across three time points (days −7, 0, and 9), we used a Kruskal-Wallis test with a Benjamini-Hochberg correction for multiple comparisons.

Code availability.The code for all sequence processing and analysis steps as well as a Rmarkdown version of this manuscript is available at https://github.com/SchlossLab/Tomkovich_PPI_mSphere_2019.

Data availability.The 16S rRNA sequencing data have been deposited in the NCBI Sequence Read Archive (accession no. PRJNA554866).