Freeze-dried fecal samples are biologically active after long-lasting storage and suited to fecal microbiota transplantation in a preclinical murine model of Clostridioides difficile infection

Abstract

Fecal microbiota transplantation is now recommended for treating recurrent forms of Clostridioides difficile infection. Recent studies have reported protocols using capsules of either frozen or freeze-dried stool allowing oral administration in in- and out-patient settings. However, a central question remains the viability, engraftment, and efficacy of the microbiome over time during storage life. This study shows that both the freeze-drying and freezing procedures for fecal samples allowed preserving viability, short-chain fatty acids concentration, and anti-Clostridioides difficile properties of microbiota without significant alteration after storage for 12 months. Fecal transplantation with freeze-dried microbiota allowed engraftment of microbiota leading to clearance of Clostridioides difficile infection in a preclinical murine model with a survival rate of 70% versus 53-60% in mice treated with frozen inocula, and 20% in the untreated group. Moreover, the freeze-dried powder can be used to fill oral hard capsules using a very low amount (0.5%) of glidant excipient, allowing oral formulation. Altogether, this study showed that freeze-dried inocula can be used for the treatment of Clostridioides difficile infection with long-lasting stability of the fecal microbiota. This formulation facilitates biobanking and allows the use of hard capsules, an essential step to simplify patient access to treatment.

Keywords: Clostridioides difficile; Fecal microbiota transplant; freeze-dried microbiota; frozen microbiota; pre-clinical model.

Figures

Figure 1.

Microbiota viability of freeze-dried and frozen human fecal samples over 12 months of storage. (a) Dominant microbiota: EOS bacteria (n = 6), Bacteroides sp. (n = 6), and Bifidobacterium sp. (n = 6) (b) Subdominant microbiota: Enterobacteriaceae (n = 6), Enterococcus sp. (n = 5), and Lactobacillus sp. (n = 3). Time study: Week 1, W1; Month 3,6 and 12: M3, M6, M12. Results are expressed as the mean ± SEM. Statistical significance: *<0.05, **<0.01, **<0.001.

Figure 2.

Fecal metabolic activity of frozen and freeze-dried human fecal microbiota after 12 months of storage. (a) Amount of Short-Chain Fatty Acids at M12. Results are expressed as the mean ± SEM. (b) Anti-CD activity of stool samples at M12. Left: results are given as the ratio between inhibition diameters of pure bacteria or stool samples and chemical control, and expressed as the mean ± SEM. Right: representative image of an anti-CD agar spot test: A: Bifidobacterium infantis longum CUETM 89–215, B: Lactobacillus spp., C: stool frozen in G10, D: stool frozen in G80, E: chemical control (lactic acid, 500 mM), F: stool frozen in G10, G: stool lyophilized FD. White lines represent inhibition diameter. Statistical significance: *<0.05, **<0.01, ***<0.001.

Figure 3.

Ability of freeze-dried human fecal microbiota to colonize the digestive tract of germ-free mice. (a) Scheme of the experiment. (b) Follow-up of main bacterial groups per gram of feces after colonization with rehydrated FD microbiota [: #S4-FD (n = 4),: #S5-FD (n = 4), feces sampled per cage]. Results are expressed as the mean for each of the 2 FD samples. (c) Representative image of a colon section from germ-free mice colonized with rehydrated FD microbiota (D30). Scale bar, 50 µm; crypt depth in the colon of germ free and mice colonized with rehydrated FD microbiota (30 crypts per mouse were measured) and Ki67-positive cells expressed as the percentage of total cell number in colon crypts from germ-free mice colonized with rehydrated FD microbiota (60 crypts per mouse were analyzed). Statistical significance: *

Figure 4.

Evolution of clinical parameters after frozen and freeze-dried FMT in Clostridioides difficile-infected mice. (a) Experimental design of the CD-infected-FMT mouse model. Two stools samples were evaluated #S1 and #S6. (b) Monitoring of body weight and a clinical score of all survival mice (up to the day of death). Daily measurements were made starting from Day 0 (day of infection). Grey squares show the total number of mice before randomization into the various FMT groups. Results are expressed as the mean (c). Survival curves of CD-infected mice receiving FMT with FD or frozen microbiota. CD-: control group receiving PBS (n = 4), CD+: CD-infected mice (n = 10); FMT-G10: CD-infected mice receiving FMT with frozen microbiota diluted in G10; FMT-G80: CD-infected mice receiving FMT with frozen microbiota diluted in G80 and FMT-FD: CD-infected mice receiving FMT with freeze-dried microbiota. For each CD group receiving FMT, n = 20, i.e. 10 mice per stool sample.

Figure 5.

Evolution of the biological parameters of Clostridioides difficile-infected mice following frozen or freeze-dried FMT (a) Level of CD in the cecal content and level of MPO in colonic tissue at the time of necropsy. (b) Representative images of colonic and cecal sections from control (CD- and CD+) mice and mice receiving FMT with frozen or FD microbiota (day 4 post-infection). Scale bar, 100 μm. (c) Histopathological score measured from CD-infected mice receiving FMT with frozen or FD microbiota. Six mice randomly selected for each point. Statistical significance: *<0.05, **<0.01.

Figure 6.

Comparison of the microbiota at day 4 between control mice and Clostridioides difficile-infected mice that received, or not, frozen or freeze-dried FMT. (a) α-diversity. (b) β-diversity using Bray–Curtis distance. (c) Composition of the microbiota at the family level. The two stool samples, #S1 and #S6, used for FMT are represented. Statistical significance: *<0.05.

Figure 7.

Development of hard capsules for oral administration of freeze-dried fecal microbiota. Upper panel: (a) examples of macroscopic aspects (n = 4). (b) microscopic aspects Optical Microscopy (x10) and (c) SEM. Lowe panel: flowability and bulk density of FD samples without excipient (n = 5), with 25% to 75% filler (n = 2) or with glidant (G1 n = 1, G2 n = 1, G3 n = 6), ND: not determined.