17The lumen of the small intestine (SI) is filled with particulates: microbes, therapeutic particles, and food 18 granules. The structure of this particulate suspension could impact uptake of drugs and nutrients and the 19 function of microorganisms; however, little is understood about how this suspension is re-structured as it 20 transits the gut. Here, we demonstrate that particles spontaneously aggregate in SI luminal fluid ex vivo. We 21 find that mucins and immunoglobulins are not required for aggregation. Instead, aggregation can be controlled 22 using polymers from dietary fiber in a manner that is qualitatively consistent with polymer-induced depletion 23 interactions, which do not require specific chemical interactions. Furthermore, we find that aggregation is 24 tunable; by feeding mice dietary fibers of different molecular weights, we can control aggregation in SI luminal 25 fluid. This work suggests that the molecular weight and concentration of dietary polymers play an 26 underappreciated role in shaping the physicochemical environment of the gut. 27 28
Results
54PEG-coated particles aggregate in fluid from the murine small intestine 55 It has been observed that both bacteria (19)(20)(21)23,25,26) and particles (3,(36)(37)(38) aggregate in the gut.
56Experiments have been performed in which mice are orally co-administered carboxylate-coated nanoparticles, 57 which are mucoadhesive, and PEG-coated nanoparticles, which are mucus-penetrating (3). The carboxylate-58 coated particles formed large aggregates in the center of the gut lumen. In contrast, PEG-coated particles were 59 sometimes found co-localized with carboxylate-coated particles and also penetrated mucus, distributing across 60 the underlying epithelium of the SI as aggregates and single particles.
61To evaluate the distribution of particulate suspensions in the SI, we suspended 1-µm-diameter 62 fluorescent PEG-coated particles (see Materials and Methods for synthesis) in buffers isotonic to the SI and 63 orally administered them to mice. We chose 1 µm-diameter particles because of their similarity in size to 64 bacteria. We collected luminal contents after 3 h and confirmed using confocal fluorescence and reflectance 65 microscopy that these particles aggregated with each other and co-aggregated with what appeared to be digesta 66 (Fig. 1C and D; Materials and Methods). On separate mice, fluorescent scanning was used to verify that 67 particles do transit the SI after 3 h (Fig. 1A and B; Materials and Methods).68 69 109the sizes of all aggregates in solution using confocal microscopy (see Materials and Methods). From these 110 datasets, we created volume-weighted empirical cumulative distribution functions (ECDFs) of all the aggregate 111 sizes in a given solution. We used these volume-weighted ECDFs to compare the extent of aggregation in a 112 given sample ( Fig. 2F and H). To test the variability of aggregation in samples collected from groups of mice 113 treated under the same conditions, we compared the extent of aggregation in p...