key: cord-0258521-zd88e8j9
authors: Pensinger, Daniel A.; Fisher, Andrea T.; Dobrila, Horia A.; Van Treuren, William; Gardner, Jackson O.; Higginbottom, Steven K.; Carter, Matthew M.; Schumann, Benjamin; Bertozzi, Carolyn R.; Anikst, Victoria; Martin, Cody; Robilotti, Elizabeth V.; Chow, JoMay; Buck, Rachael H.; Tompkins, Lucy S.; Sonnenburg, Justin L.; Hryckowian, Andrew J.
title: Butyrate differentiates permissiveness to Clostridioides difficile infection and influences growth of diverse C. difficile isolates
date: 2022-05-21
journal: bioRxiv
DOI: 10.1101/2022.05.20.492898
sha: e711181bc5ff225ef79a761b8039a7c3276d915b
doc_id: 258521
cord_uid: zd88e8j9

A disrupted “dysbiotic” gut microbiome engenders susceptibility to the diarrheal pathogen Clostridioides difficile by impacting the metabolic milieu of the gut. Diet, in particular the microbiota accessible carbohydrates (MACs) found in dietary fiber, is one of the most powerful ways to affect the composition and metabolic output of the gut microbiome. As such, diet is a powerful tool for understanding the biology of C. difficile and for developing alternative approaches for coping with this pathogen. One prominent class of metabolites produced by the gut microbiome are short chain fatty acids (SCFAs), the major metabolic end products of MAC metabolism. SCFAs are known decrease the fitness of C. difficile in vitro and that high intestinal SCFA concentrations are associated with reduced fitness of C. difficile in animal models of C. difficile infection (CDI). Here, we use controlled dietary conditions (8 diets that differ only by MAC composition) to show that C. difficile fitness is most consistently impacted by butyrate, rather than the other two prominent SCFAs (acetate and propionate), during murine model CDI. We similarly show that butyrate concentrations are lower in fecal samples from humans with CDI relative to healthy controls. Finally, we demonstrate that butyrate impacts growth in diverse C. difficile isolates. These findings provide a foundation for future work which will dissect how butyrate directly impacts C. difficile fitness and will lead to the development of diverse approaches distinct from antibiotics or fecal transplant, such as dietary interventions, for mitigating CDI in at-risk human populations. IMPORTANCE Clostridioides difficile is a leading cause of infectious diarrhea in humans and it imposes a tremendous burden on the healthcare system. Current treatments for C. difficile infection (CDI) include antibiotics and fecal microbiota transplant, which contribute to recurrent CDIs and face major regulatory hurdles, respectively. Therefore, there is an ongoing need to develop new ways to cope with CDI. Notably, a disrupted “dysbiotic” gut microbiota is the primary risk factor for CDI but we incompletely understand how a healthy microbiota resists CDI. Here, we show that a specific molecule produced by the gut microbiota, butyrate, is negatively associated with C. difficile burdens in humans and in a mouse model of CDI and that butyrate impedes the growth of diverse C. difficile strains in pure culture. These findings help to build a foundation for designing alternative, possibly diet-based, strategies for mitigating CDI in humans.

the short chain fatty acids (SCFAs), which are the metabolic end products of 116 MAC metabolism by the microbiome [17] , impact C. difficile fitness in pure culture 117 and in animal models of infection [8, 18, 19 their metabolic end-products on CDI and the promise for rapid translation to 125

In previous work, we demonstrated that inulin, a β-2,1-linked fructan, 140 suppresses C. difficile burdens in a murine model of CDI [8] . To begin to test the 141 generalizability of these findings to other purified MAC sources, we focused on 142 FOS, which is structurally identical to inulin except for its degree of 143 polymerization (DP) (FOS DP = 2-8 and inulin DP = 2-60) [30] . In contrast to 144 mice fed inulin, mice fed FOS retain high burdens of C. difficile 630 during CDI 145 (Figure 1 ). These results generated two possible hypotheses, that the effect of 146 with pulsed amperometric detection (HPAEC-PAD) to determine the extent of 162 FOS utilization by C. difficile grown in FOS-supplemented minimal medium. We 163 determined that C. difficile does not utilize FOS but instead consumes the trace 164 amounts of glucose and fructose in the FOS preparation ( Figure 2B , peaks 165 within gray bars correspond to glucose and fructose based on reference 166 chromatograms in Figure S1 ). Therefore, this work supports previous findings 167 that C. difficile does not readily consume MACs [31] and that it is likely that 168 factors unrelated to FOS metabolism by C. difficile contribute to the inability of 169 FOS to clear murine CDI. 170

The major metabolic end products of MAC metabolism by the gut FOS-fed mice relative to mice fed a MAC deficient diet (Figure 2C) C. difficile)) and patients without CDI (negative for CDI (via Cepheid Xpert C. 231 difficile)). In stool from the symptomatic C. difficile patients, we observed 232 significantly lower concentrations of butyrate (but not acetate or propionate) 233 relative to patients without CDI (Figure 4) , which demonstrates that our findings 234 in mice ( Figure 3B) 1-3) . 243

Though similar butyrate-dependent effects were observed in 4 unsequenced C. Future diet-based strategies to mitigate CDI will similarly be informed by the growing literature surrounding the impact of other dietary inputs on CDI (see 299

Because butyrate levels differentiate mice and humans that have CDI from 301 those that do not (Figure 2C, 3B, 4) , continued focus on this SCFA in the context 302 of CDI will yield important insights into the biology of C. difficile, the ecology of 303 CDI, and future therapeutic approaches. We and others previously showed that 304 butyrate negatively impacts growth in 5 distinct C. difficile strains [8, 18] and in the 305 current study we extend these findings to 12 additional C. difficile strains (Figure 306 and presumably helps to re-establish facets of microbiome community function 318 that allow C. difficile to thrive. 319

Future work based on the above conceptual model and the data 320 presented in the current study will seek to understand the variety of host-by-microbiome-by-diet interactions that influence C. difficile fitness in the gut. 322

Specific foci on the molecular mechanisms and genetic circuitry underlying the 323 responses of C. difficile to butyrate will facilitate a better basic understanding of 324 C. difficile and how it interacts with the host and the gut microbiome. In addition, 325 continued research on these and other diet-driven effects on CDI are likely to 326 yield insights that will aid in the development of specific and targeted 327 1 minute and the OD 600 of the cultures was recorded using Gen5 software 365 (version 1.11.5). 366

All animal studies were conducted in strict accordance with Stanford 369 University Institutional Animal Care and Use Committee (IACUC) guidelines. 370

Murine model CDI was performed on age-and sex-matched conventionally-371 reared Swiss-Webster mice (Taconic) between 8 and 17 weeks of age. 372

To reduce colonization resistance against C. difficile, mice were given a 373 undetectable in all mice prior to inoculation with CDI.

Mice were fed one of eight custom diets (Bio Serv) ad libitum: (1) a MAC-393 deficient control diet containing 68% glucose (w/v), 18% protein (w/v), and 7% fat 394 Two methods were used to quantify SCFAs in cecal contents from mice 439 0.01 mM, 0.001 mM, and 0 mM) were derivatized as above and included in each 505 run to verify sample concentrations were within linear ranges. For samples within 506 linear range, analyte concentration was calculated as the product of the paired 507 internal standard concentration and the ratio of analyte peak area to internal 508 standard peak area. A single product ion was used for each analyte, no 509 secondary or qualifier ions were used. To ensure the highest signal-to-noise 510 ratio, the following steps were taken. First, to ensure that the predicted singly 511 

Raw OD 600 measurements of cultures grown in mRCM (see 'Bacterial strains and 523 culture conditions', above) were exported from Gen5 and analyzed using the 524 growth_curve_statistics.py script (see Code Availability, below). Growth rates 525

were determined for each culture by calculating the derivative of natural log-526 transformed OD 600 measurements over time. Growth rate values at each time point were then smoothed using a moving average over 150-min intervals to 528 minimize artefacts due to noise in OD measurement data, and these smooth 529 growth rate values were used to determine the maximum growth rate for each 530 culture. To mitigate any remaining issues with noise in growth rate values, all 531 growth rate curves were also inspected manually. Specifically, in cases where 532 the growth_curve_statistics.py script selected an artefactual maximum growth 533 rate, the largest local maximum that did not correspond to noise was manually 534 assigned as the maximum growth rate. Additionally, lag time was calculated as 535 

Regulation of short-chain fatty acid 643 production

Prebiotic-646 non-digestible oligosaccharides preference of probiotic bifidobacteria and 647 antimicrobial activity against Clostridium difficile

Formation of propionate and butyrate by the human 712 colonic microbiota

Intestinal Short Chain Fatty Acids and their

A short chain fatty acid-centric 719 view of Clostridioides difficile pathogenesis

Fermentation of fructooligosaccharides and inulin by bifidobacteria: a 724 comparative study of pure and fecal cultures

Restoration of short chain fatty acid and bile acid metabolism following 736 fecal microbiota transplantation in patients with recurrent Clostridium 737 difficile infection

Regulation of bacterial pathogenesis by intestinal 740 short-chain Fatty acids

In vitro 760 immunomodulatory effects of human milk oligosaccharides on murine 761 macrophage RAW264.7 cells. Carbohydr Polym

The multidrug-resistant human pathogen Clostridium difficile has a 784 highly mobile, mosaic genome

An isotope-labeled chemical 787 derivatization method for the quantitation of short-chain fatty acids in 788 human feces by liquid chromatography-tandem mass spectrometry

(blue), and FOS (green) are shown. Chromatograms for FOS