The present invention is in the field of methods, devices, and especially compositions for use in sampling micro-organisms. More specifically, the invention relates to devices and methods for sampling the gastrointestinal (GI) tract of a subject for micro-organisms, wherein a composition is employed that preserves the microbial composition and microbial abundance at the time of sampling. The invention is also directed to such preserving compositions.

STATE OF THE ART

Numerous devices for sampling the GI tract of a subject for micro-organisms such as bacteria are known in the art. For instance, WO 2005/112460 describes a method and ingestible devices for taking in vivo biopsies.

Although the device of WO 2005/112460 is mainly directed to taking tissue samples, it may also collect luminal materials such as micro-organisms and preserve them in a hquid which may be a preservative, a sahne or a fixation hquid.

Another device is described in WO 2014/159532, wherein an intestinal microbial flora sampling system is described that can be used to identify microbes present in a subject's digestive tract. The sample is preserved in the device by controlling the temperature in the reservoir for collecting samples. Further devices for sampling the GI tract of a subject for micro-organisms are for instance described in WO 2005/025413, WO

2014/140334 and US 20150112166.

Devices for sampling the GI tract of a subject for micro-organisms, as mentioned hereinbefore, generally contain a reservoir for storing or holding the collected sample, optionally filled with a substance or

composition that has a preserving effect on the microbial composition and microbial abundance in the sample, or having other means for preserving the microbial composition and microbial abundance in the sample, for example by employing means for controlling the temperature in the reservoir in such a way that bacterial growth is inhibited. As the colonic transit time averages about 30 hours (Chaussade et al., Digestive diseases and sciences, 34(8): 1168- 1172 (1989)), and transit takes place at a

temperature of about 37°C, it is evident that precautionary measures have to be taken in order to prevent that the microbial composition and microbial abundance in the sample changes after the sample is taken.

It has already been suggested in the art, as a general solution to this problem, to pre-fill the reservoir of the device with a preserving substance or composition, also called a "quencher", to prevent the micro- organisms from growing in the device and thus to prevent the microbial composition and abundance from changing. WO 2007/061305 A2 describes a device for sampling the GI tract of a subject for micro-organisms, the device optionally comprising in its reservoir a quenching or stabilizing liquid comprising ethanol or methanol. At the moment, however, no extensive research is performed with regard to the type of quencher that is specifically and advantageously suited for employment in reservoirs of devices for sampling the GI tract of a subject for micro-organisms.

The difficulty in developing a suitable preserving composition for use in devices as mentioned hereinabove is that such a composition has to fulfill numerous requirements. One of the most important requirements is that, since the volume of the reservoir of devices for sampling the GI tract is strongly limited as the device has to be ingestible, the preserving

composition has to be effective even when diluted with a sample, for example at a 1:4, 1:6, 1:8 or even higher dilution (composition:sample). Due to the hmited reservoir volume, a device for sampling the GI tract of a subject does not allow for an excess of preserving composition to be filled in the reservoir. This means that the composition, even in its diluted form, should prevent the microbial composition and microbial abundance in the sample from changing. Alternative requirements to be met are (i) low toxicity of the components of the composition towards the human subject and (ii) DNA preservation such that, after sampling, for instance the 16S ribosomal RNA gene-sequence can be analyzed. The latter means that DNA should not be degraded by nucleases during the remaining transit through the gastrointestinal tract.

At this moment, no preserving compositions or "quenchers" have been developed, and compared with one another, for the specific purpose of preserving the microbial composition and abundance in a sample in a device for sampling the GI tract of a subject during colonic transit. With the present invention it was found that only a limited number of preserving compositions are suitable and effective for the purpose of preserving the microbial composition and abundance in a sample in a device for sampling the GI tract of a subject during colonic transit. The present invention therefore provides a device for sampling the gastrointestinal (GI) tract of a subject for micro-organisms, comprising a reservoir for holding a sample of micro-organisms, wherein said reservoir contains an aqueous composition comprising 2.5-30% (w/v) of a

[dodecanoyl(methyl)amino]acetate (lauroyl sarcosine) and 0.007-0.5 M of ethylenediaminetetraacetic acid (EDTA). It was found that the minimum concentration of lauroyl sarcosine, for the aqueous composition to be effective in preserving a sample, is 1.25% (w/v) (0.043 M). The skilled person understands that, depending on the dilution factor, different starting concentrations of lauroyl sarcosine can be used, while, in any case, at least 2.5% (w/v) of lauroyl sarcosine should be present before sampling, as the minimum dilution factor considered here is 1: 1. If it is anticipated to dilute the aqueous composition with sample in a 1: 1 ratio, a concentration of 2.5% (w/v) in the reservoir of a device of the invention (before sampling) is minimally needed for lauroyl sarcosine to be effective. Since the skilled person knows beforehand what the sample volume is that is to be aspirated, he or she can easily calculate how much lauroyl sarcosine is minimally needed to achieve at least a 1.25% (w/v) of lauroyl sarcosine after dilution with sample. The same rationale applies for the other components that are present in an aqueous composition of the invention contained in a device of the invention. For EDTA, the minimum concentration that is to be present in the reservoir of a device of the invention is 0.007 M, which concentration is based on a potential dilution with sample of 1: 1. Thus, if it is envisaged to dilute the aqueous composition with sample in a 1:2 ratio, the skilled person would directly know that the concentration of EDTA in the reservoir of a device of the invention should be at least 0.011 M to achieve the minimum concentration of EDTA of 0.007 M.

It was unexpectedly found that an aqueous composition based on lauroyl sarcosine (also known as sarcosyl) and EDTA can advantageously be used in preserving the microbial composition and abundance in a sample in a device for sampling the GI tract of a subject during GI transit. Even when diluted up to 1:8 times in a sample, the composition still provided for cell lysis and preservation of the microbial composition and microbial abundance in the sample (Figures 5 and 6). In addition, nucleic acids isolated from samples treated with lauroyl sarcosine-based aqueous compositions were of good quality and not degraded by nucleases.

The device of the invention is an ingestible device for sampling the gastrointestinal (GI) tract of a subject for micro-organisms. Ingestible devices for sampling the GI tract of a subject for micro-organisms are known in the art. All ingestible devices that contain a reservoir for holding a sample, said reservoir being functionally and/or operably connected to means for aspirating a sample from the outside of the device to the inside of the device, are compatible with the present invention. Such ingestible devices are generally electronic devices, preferably in the form of a capsule or a pill. The invention inter alia resides in meeting the stringent

requirements under which "quenchers" have to be effective in ingestible devices for sampling the GI tract of a subject for micro-organisms. Ingestible devices with sample aspiration functionality are generally known in the art, and include a device such as described in Sullivan et al., J Pediatr

Gastroenterol Nutr, 7:544-7 (1988) and Meijer et al., Virchows Arch,

442: 124-8 (2003). The device used in the Examples is the IntelliCap® CR (Medimetrics, Eindhoven, NL; Schaar et al., Gastrointest Endosc, 78(3):520- 528 (2013)). Functionality of this device has been extended to aspirate a sample of fluid from the lumen of the GI tract into the reservoir of the device. Further in order to meet the aim of preserving or quenching the contents at the time of sampling, before ingestion the reservoir is partially loaded with an aqueous composition comprises 2.5-30% (w/v) of a

[dodecanoyl(methyl)amino]acetate (lauroyl sarcosine) and 0.007-0.5 M of ethylenediaminetetraacetic acid (EDTA). Preferably, the composition comprises 5-15% (w/v) or 6-14%) (w/v), more preferably 8- 12% (w/v) or 9-11% (w/v), most preferably about 10% (w/v), of lauroyl sarcosine; and 0.01-0.2 M, more preferably 0.02-0.1 M, most preferably about 0.1 or 0.03 M, of EDTA. It is noted that 10% (w/v) of lauroyl sarcosine corresponds to a concentration of 0.34 M.

The term "ingestible" is used herein in the context of a device that can be orally ingested and subsequently can travel through the

gastrointestinal tract until it is expelled from the body through the anus, in a manner analogous to the flow of food through the gastrointestinal tract.

The term "lauroyl sarcosine", as used herein, includes reference to its anionic form or to its solubilized form when it is formulated as a salt in water. Suitable salts of lauroyl sarcosine are inter alia, sodium lauroyl sarcosine and ammonium lauroyl sarcosine. A preferred salt of lauroyl sarcosine is sodium lauroyl sarcosine. The chemical formula of sodium lauroyl sarcosine is displayed hereinbelow.

The term "gastrointestinal (GI) tract", as used herein, includes the entire digestive tract such as the esophagus, stomach, small intestine and the large intestine.

The term "subject", as used herein, refers to a vertebrate, preferably a mammal, more preferably a human.

The term "micro-organism", as vised herein, is defined as a microscopic organism, including bacteria (e.g. intestinal bacteria), fungi and protozoa. While viruses and prions are not minute "living" organisms, for purposes of this disclosure they are included in the term micro-organism. Preferably, the micro-organism is a bacterium, more preferably a bacterium generally known to reside in the gastrointestinal tract, such as bacteria of the phyla Firmicutes, Bacteroidetes, Actinobacteria, and Proteobacteria (gut flora).

The term "reservoir", as used herein, refers to the compartment of a device, for sampling the gastrointestinal (GI) tract of a subject for microorganisms, wherein a sample is held or is to be held. Preferably, the reservoir has a volume of 50-10000 microliter, more preferably 50-5000 or 100-1000 microliter, and most preferably 200-500 microliter or about 300 microliter.

The term "aqueous composition", as used herein, refers to and includes a water-based composition. In particular, it refers to a solution, dispersion or a suspension of matter in an aqueous phase, preferably water, more preferably distilled water. The skilled person understands that a water-based composition employs water as a base ingredient, having dissolved therein components described herein. The composition may be provided or present in a device of the invention in powder form, for instance as a powder coating to the internal surface of the reservoir. The aqueous composition preferably counters changes in microbial composition and microbial abundance in a microbial, preferably bacterial, sample, when contacted therewith.

The terms "quencher", "quench buffer", "quench composition" and "aqueous composition", as used herein, are interchangeable.

The aqueous composition contained in a device of the invention preferably further comprises 2-18 M or 3-15 M, more preferably 5-12 M, even more preferably 6-11 M or 7-10 M, most preferably 7-9 M or about 8 M, of urea. Urea can be dissolved at a temperature of 37 °C, under atmospheric pressure. 2 M urea is a minimum concentration in the reservoir (before sampling) if the sample is to be diluted 1: 1 with sample, as it was found that the lowest effective concentration after dilution is 1 M urea. The skilled person knows how to correct (i.e. increase) the concentration of urea in the reservoir of a device of the invention if the dilution factor of aqueous composition:sample is for instance increased from 1: 1 to 1:4 in order to still arrive at a concentration of urea of at least 1 M. If urea is present in the composition, the composition preferably also comprises 0.05-0.5 M, preferably 0.05-0.4 M, more preferably about 0.2 M of Na2HPO<i. If urea is not explicitly mentioned as present in the composition, the composition preferably comprises 0.1-5 M, more preferably 0.1-1 M, most preferably 0.1- 0.2 M or 0.1 M of NaCl. The statements with regard to the minimum concentrations that were made for lauroyl sarcosine, EDTA and urea, also apply for Na2HP0 and NaCl. More specifically, the minimum concentration of Na2HP04 in the reservoir of a device of the invention is at least 0.05 M when a dilution of 1: 1 with sample is anticipated, making the lowest effective concentration of Na2HPO i 0.025 M. The minimum concentration of NaCl in the reservoir of a device of the invention is at least 0.1 M when a dilution of 1: 1 with sample is anticipated, making the lowest effective concentration of NaCl 0.05 M.

The skilled person understands that the concentrations of components indicated as present in an aqueous composition of the invention are chosen in such a way that they allow for the presence of the other components in at least one concentration within their specified

concentration range. In other words, in an aqueous composition of the invention, the concentrations of components are chosen so that they do not individually, nor in combination, exceed the saturation solubility of the aqueous composition.

The aqueous composition contained in a device of the invention may further comprise 0.07-4 M, more preferably 0.07-0.5 M, most preferably 0.07- 0.15 M of tromethamine (Tris). The statements as regards minimum concentrations that were made for lauroyl sarcosine, EDTA and urea, are also true for Tris. More specifically, the minimum concentration of Tris in the reservoir of a device of the invention is at least 0.07 M when a dilution of 1: 1 with sample is anticipated, making the lowest effective concentration of NaCl 0.035 M. Preferably, the concentration of Tris in the aqueous composition as described herein is 0.07-3 M, 0.07-2.5 M, 0.07-2 M, 0.07-1.5 M, 0.07-1 M, more preferably 0.07-0.4 M, 0.07-0.35 M, 0.07-0.30 M, 0.07- 0.25 M or 0.07-0.20 M, most preferably about 0.1 M.

An alternative aqueous composition contained in a device of the invention comprises lauroyl sarcosine, NaCl and Tris as indicated

hereinabove, and further comprises 30-90%, preferably 60-80%, more preferably about 70% (v/v) of ethanol; and EDTA as indicated hereinabove, preferably to saturation (0.5 M at pH 8).

The pH of the aqueous composition contained in a device of the invention can be 4-11 or 5-10, preferably 6-9, more preferably 6.5-7.5, and most preferably about 6.8 or about 7.5. More specifically, aqueous compositions comprising urea most preferably have a pH of about 6.8 and aqueous compositions comprising Tris and or NaCl most preferably have a pH of about 7.5.

The volume of the aqueous composition contained in a device of the invention is preferably 5-1000 microliter, more preferably 5-500 microliter, and most preferably 10-100 microliter or about 50 microliter. Preferably, the ratio of sample:aqueous composition in the reservoir of the device of the invention is in the range of 1: 1-9: 1, preferably 3: 1-8: 1, more preferably 4: 1- 8: 1. This ratio is achieved when the sample is aspirated from the lumen of the GI tract into the reservoir of the device and contacted with the

composition of the invention. Alternatively, the volume of the aqueous composition in the reservoir of a device of the invention is preferably 5-75%, preferably 10-50%, more preferably 15-25%, and most preferably about 17% of the total volume of the reservoir. Alternatively , the reservoir and aqueous composition have a volume allowing for a dilution of sample:aqeous composition of 1: 1-9: 1, preferably 3: 1-8: 1, more preferably 4: 1-8: 1 during and/or after sample aspiration.

The invention is further directed to an aqueous composition as described hereinabove in the context of an aqueous composition contained in a device of the invention.

More specifically, the invention is directed to an aqueous composition comprising (i) 2.5-30% (w/v), preferably 5-15% (w/v), more preferably 8-12% (w/v), most preferably about 10% (w/v), of lauroyl sarcosine, (ii) 0.007-0.5 M, preferably 0.01-0.2 M or 0.015-0.15 M, more preferably 0.02-0.1 M or 0.02-0.06 M, most preferably about 0.03 M, of EDTA, (iii) 2-18 M, preferably 5-12 M, more preferably 7-9 M, most preferably about 8 M, of urea and (iv) 0.05-0.5 M, preferably 0.05-0.4 M or 0.1-0.3 M, more preferably about 0.2 M of Na2HP04. This composition is preferably buffered at a pH of about 6.8. Alternatively, the invention is directed to an aqueous composition comprising (i) 2.5-30% (w/v), preferably 5-15% (w/v), more preferably 8- 12%) (w/v), most preferably about 10% (w/v), of lauroyl sarcosine, (ii) 0.007- 1 M or 0.007-0.5 M, preferably 0.01-0.2 M, more preferably 0.02-0.1 M or 0.05-0.15 M, most preferably about 0.03 or 0.1 M, of EDTA, (in) 0.07-4 M, preferably 0.07-3 M or 0.07-2.5 M, more preferably 0.07-2 M, 0.07-1.5 M, 0.07- 1 M or 0.07-0.5 M, most preferably 0.07-0.4 M, 0.07-0.35 M, 0.07-0.30 M, 0.07-0.25 M, 0.07-0.20 M, 0.07- 0.15 M or about 0.1 M, of Tris, and (iv) 0.1-5 M or 0.1- 3 M, more preferably 0.1-2 M or 0.1- 1 M, most preferably 0.1-0.2 M or about 0.1 M, of NaCl. This composition is preferably buffered at a pH of about 7.5.

Alternatively, the invention is directed to an aqueous composition comprising (i) 2.5-30% (w/v), preferably 5-15% (w/v), more preferably 8- 12% (w/v), most preferably about 10% (w/v), of lauroyl sarcosine, (ii) 0.007-0.5 M, preferably 0.01-0.5 M, more preferably 0.2-0.5 M, most preferably about 0.5 M or a saturation, of EDTA, (iii) 0.1-5 M or 0.1-2 M, preferably 0.1- 1 M or 0.1-0.5 M, more preferably 0.1-0.3, 0.1-0.2 M or about 0.1 M, of NaCl, (iv) 0.07-4 M, preferably 0.07-3 M or 0.07-2.5 M, more preferably 0.07-2 M, 0.07- 1.5 M, 0.07- 1 M or 0.07-0.5 M, most preferably 0.07-0.4 M, 0.07-0.35 M, 0.07-0.30 M, 0.07-0.25 M, 0.07-0.20 M, 0.07- 0.15 M or about 0.1 M, of Tris, and (v) 30-90%, preferably 60-80%, more preferably about 70%. (v/v), of ethanol.

All these compositions share the presence of lauroyl sarcosine and EDTA, and it was only with these buffers that the beneficial effects of the invention were acquired.

Preferably, a device of the invention contains a composition of the invention.

The invention is further directed to a method for sampling the GI tract of a subject for micro-organisms, comprising the steps of a)

administering a device according to the invention to a subject; and b) sampling the gastrointestinal tract of said subject for micro-organisms. The administration of the device is preferably orally, and can be performed in a manner equivalent to the oral administration of a pharmaceutical

composition suitable for that purpose. The sampling is performed by the device of the invention; the device of the invention preferably having means to communicate with a device external to the subject, wherein said

communication is for providing a stimulus to the device to sample the gastrointestinal tract of the subject. More specifically, and preferably, a device of the invention is tracked during intestinal transit by real-time transmission of pH and/or temperature values measured by a device of the invention. On the basis of pH and temperature values thus obtained, it is possible to determine the position of a device of the invention to a specific location in the intestines and, for instance, specifically sample only the small intestine, such as the ileum, by providing a samphng stimulus via an external device communicating with a device of the invention when a device of the invention is in the small intestines. A device of the invention preferably incorporates means for measuring pH and temperature values and means for transmitting data to an external device having means for receiving such data and, preferably, means for interacting with a device of the invention. Alternatively, a device of the invention can be programmed to sample under predetermined circumstances. When exiting the body of the subject, the device of the invention is preferably cleaned with an aqueous composition, preferably a soap solution, present on a tissue. Further steps may include cleaning a device of the invention with a chlorine solution and/or an alcohol solution, the solution preferably being present on a tissue.. After cleaning a device of the invention, the sampling opening is preferably closed, for instance with parafilm, The sample present in a device of the invention can be retrieved by using a syringe with needle. Subsequently, DNA isolation can be performed by methods known in the art.

The advantage of employing an aqueous composition of the invention in a device for sampling the gastrointestinal (GI) tract of a subject, is that the microbial composition and abundance of the sample is preserved during colonic transit. Even when diluted up to 1:8 times in a sample, the composition still provided for cell lysis and preservation of the microbial composition of the sample (Figures 5 and 6). It was totally unexpected that lauroyl sarcosine-based compositions can be employed in small volumes in a device of the invention, while allowing the reservoir of a device of the invention to be predominantly filled with sample. This is contrary to the fact that buffers such as lysis buffers are generally employed in excess on the sample. After sampling, these effects provide for a reliable and accurate fingerprint of the microbial composition and abundance in the GI tract at a specific sampling location. After sampling, microbial species identification can be performed by methods generally known in the art, including real-time qPCR analysis and (next-gen) sequencing methods. Such analysis is preferably directed to the microbial 16s rRNA gene. A device of the invention, containing a composition of the invention, was with success employed in a human validation study on the effects of diet on intestinal microbial composition.

The invention is further directed to a use of a device of the invention in sampling the gastrointestinal (GI) tract of a subject for micro- organisms. The invention is also directed to a use of a composition of the invention in a device for sampling the gastrointestinal (GI) tract of a subject for micro-organisms.

The invention is also directed to a use of a composition of the invention in counteracting changes in microbial composition and microbial abundance in a microbial sample. Preferably, the microbial sample is a bacterial sample.

The invention also relates to a method of producing an aqueous composition of the invention, comprising the step of a) providing an aqueous composition with 1-30% (w/v) of a [dodecanoyl(methyl)amino]acetate

(lauroyl sarcosine) and 0.005-0.5 M of ethylenediaminetetraacetic acid (EDTA). Preferably, Tris, NaCl, urea, Na2HPO<j, and/or ethanol are added to the composition in the combinations and concentrations as indicated hereinabove for the aqueous composition contained in a device of the invention.

The invention also relates to a method of producing a device of the invention, comprising the step of a) filhng a reservoir, for holding a sample of micro-organisms, of a device, for sampling the gastrointestinal (GI) tract of a subject for micro-organisms, with an aqueous composition of the invention. The reservoir is preferably filled by injecting the composition through a capillary opening of the device, preferably with a syringe.

In the context of a method for sampling of the invention, or a use of a device of the invention for sampling, the subject, before having

administered a device of the invention, may have received dietary or pharmacological intervention - such as a diet or a medicament - , preferably via oral administration.

Preferably, sampling is performed in the small intestine. For the purpose of clarity and a concise description, features are described herein as part of the same or separate aspects and preferred embodiments thereof, however, it will be appreciated that the scope of the invention may include embodiments having combinations of all or some of the features described.

The invention will now be illustrated by the following examples, which are provided by way of illustration and not of limitation and it will be understood that many variations in the methods described and the amounts indicated can be made without departing from the spirit of the invention and the scope of the appended claims.

The content of the documents referred to herein is incorporated by reference. FIGURE LEGENDS

Figure 1. Panels A) and B) show the measurement of OD600 to evaluate the quenchers' (see Table 1) ability to stop the growth of Gram -positive L. plantarum and Gram -negative E. coli, respectively. The growth of L.

plantarum and E. coli was inhibited by all selected quenchers. Water was used as a control.

Figure 2. Panels A)-F) show OD595 measurements of Gram-positive L. plantarum in media (referred to as suspension in the figure) over time in a microtiter plate with different concentrations of quench buffer (referred to as buffer in the figure). In the water control, a steep increase of about 0.6 units OD595 can be observed in all ratios of ":quencher:bacteria and media" (Q:B). The only other increases that are of this magnitude are those observed with the urea buffer and the ethanol control in the 1:8 ratio.

Figure 3. In panels A)-C) a longer incubation time (20h) of the quenchers was tested, again with L. plantarum , with Gram -negative E. faecalis and with a mixture of both species. Bacteria and media were incubated with the selected quenchers in a ratio of 1:4 (Q:B). In this concentration, all of the quenchers prevented the OD595 from increasing significantly, contrary to the water control. In the co-culture sample, the water control shows bacterial growth. This growth was inhibited by all quenchers.

Figure 4. Panels A)-F) show that, in most cases, the addition of a quencher does not affect the bacterial ratio that was retrieved, meaning that a preserving effect on the microbial composition is present, as compared to the water control. Figure 5. Panels A)-F) are based on an experiment that is to a large extent the same as that described for Figure 4, except this time a ratio of Q:B of 1:8 was used. It follows from Panel E) that the urea buffer cannot suitable be used in an 1:8 ratio of Q:B for preserving the bacterial composition of the sample, as L. plantarum started to overgrow E. coli. Instead, the sarkosyl buffer, sarkosyl-urea buffer and control buffer ethanol were able to preserve the bacterial composition to a large extent constant.

Figure 6. Panels A) and B) show the result of an experiment wherein the quenchers' abilities to kill bacterial cells was evaluated. The sarkosyl- ethanol buffer, sarkosyl buffer, sarkosyl-urea buffer and RNAprotect buffer were able to kill all of the L. plantarum in the mixture, in both Q:B ratios. The urea buffer and the control ethanol buffer were unsuccessful in killing all bacterial cells. Lysis of bacterial cells is an important aspect for a quench composition, as it inter alia frees bacterial DNA from cells, which can afterwards be used for 16S rRNA gene analysis.

Figure 7. Panel A) shows the average DNA concentration in

nanogram/microliter obtained from bacterial samples that were quenched. Panel B) shows the average ratio of absorbance on 260 and 280nm

(A260/A280) per sample quenched, which is an indication for purity of the DNA. The purity of the DNA and the amount of DNA isolated were sufficient to perform 16S rRNA gene analysis. Figure 8. Panel A) shows the clustering of capsule derived samples, obtained from sampling the small intestine, and faecal samples with regard to microbial composition. It clearly follows from this figure that capsule derived samples cluster separately from the faecal samples, supporting their distinct microbial composition. Panel B) shows a boxplot with on the y-axis the number of observed species for both the capsule derived sample and faecal samples. It follows from panel B) that species diversity was much higher in faecal samples compared to capsule derived samples.

Table 1: List of quenchers tested.

Quencher Composition

Netherlands) were used as controls.

EXAMPLES

Example 1: General Materials and Methods. Bacterial strains and growth conditions

Lactobacillus plantarum WCFSl was grown in De Man, Rogosa and Sharpe (MRS) medium at 37°C without shaking. Escherichia coli MC1061 was grown in TY medium at 37°C with constant shaking. Enterococcus faecalis JCM 5803 was grown in brain heart infusion (BHI) medium at 37°C without shaking. Cells were used at an optical density at 600nm (OD600) of approximately 1.0, to assure the cultures were in the logarithmic phase of growth for all media. Experiments with mixed cultures were performed with 1: 1 volume mixtures of bacteria in their corresponding medium. Quencher recipes

The quencher compositions are as indicated in Table 1.

Measurement of optical density

Optical density was measured at 600nm with cuvettes in an Ultrospec 2000 (Pharmacia Biotech, Roosendaal, The Netherlands). Microtiter plates were measured at 595nm in a Genios microplate reader (Tecan, Zurich,

Switzerland). Plates were incubated in the plate reader at 37°C and shaken for 10s before each measurement. Quenching experiments: Single microbial cultures

Quenching experiments were performed with bacteria in their

corresponding medium (OD600-1) added to the quench buffers in different volume ratios. Subsequent incubations were at 37°C in closed micro test tubes or in microtiter plates. Samples were frozen immediately after the reported incubation times. Positive controls were frozen immediately at the start of the experiment. Bacterial mixtures were added to the quenchers in ratios 1:8, 1:4, 1:2 and 1: 1. For the co-culture experiment an equal mixture of E. coli and E. faecalis was used. Incubation times were 30min, 6hr, 20hr and 24hr.

DNA isolation

DNA was isolated with standard phenol-chloroform extraction with bead- beating (Jaeggi et al., Gut 64(5):731-42 (2015)). and enzymatic degradation by RNAse A (Qiagen) and proteinase K (Invitrogen, Bleiswijk, The

Netherlands).

RT-qPCR analysis

Real-time qPCR was performed with SYBR® Green PCR Master Mix for L. plantarum (Applied Biosystems, Nieuwekerk a/d IJssel, the Netherlands) and TaqMan® Universal PCR Master Mix for E. coli (Applied Biosystems) in a CFX384 Real-Time PCR System (Bio-Rad Laboratories BV,

Veenendaal, The Netherlands).

L. plantarum forward primer: ATGGTCCCGCGGCG

L. plantarum reverse primer: GTCCCAATGTGGCCGATTAC

E. coli forward primer: CATGCCGCGTGTATGAAGAA

E. coli reverse primer: C GGGT AAC GTC AATGAGC AAA

E.coli probe: TATTAACTTTACTCCCTTCCTCCCCGCTGAA

Primer concentrations used were 200nM and the probe concentration was ΙΟΟηΜ. The PCR program used was 2 min at 50°C and 10 min at 95°C, followed by 15 s at 95°C and 1 min at 60°C for 40 cycles. Flow cytometry: Live-dead assay

After five minutes of exposure to the quench buffers, fluorescent dyes and microspheres from the LIVE/DEAD® BacLight™ Bacterial Viability and Counting Kit (Invitrogen) were added to the cells. Thereafter the cells were fed through a FACSAria II Flow cytometer (BD Biosciences, Franklin

Lakes, US). Data was analyzed with FACSDiva software (BD Biosciences). For setting up the gates, untreated and heat- and ethanol-killed strains were used. Ratios of quencher to bacterial suspension were 1:4 and 1:8. Mixing test

Mixing of the quench buffer inside the Intellicap® and the aspirated sample was evaluated visually by making use of the two phosphorescent

components present in a Cyalume Snaplight® glow in the dark stick

(Cyalume Technologies Inc., West Springfield, MA, USA). When these components mix, the mixture will emit a green light. For the mixing test, an in -vitro test system of Medimetrics' (Eindhoven, The Netherlands)

IntelhCap® was used. This test system has reservoir and actuator configuration identical to the swallowed capsule but has wired power and control connections to ease bench experiments. In addition to mixing, the test system was used to confirm the aspiration capabilities of the system by drawing in glycerol. The device was weighed before and after aspirating glycerol.

Example 2: Bacterial growth is inhibited by quench buffers in 1:2 (Q:B) ratio.

Measurement of OD600 was chosen as the first method to evaluate the quench buffers' ability to stop bacterial growth. We added L. plantarum and E. coli in their corresponding medium to a selection of quench buffers in a quench buffer:bacterial culture ratio (Q:B) of 1:2. The growth of Gram- positive L. plantarum and Gram -negative E. coli was inhibited by all selected quenchers. The increase in OD600 observed with the water control was absent in the other samples (Figure 1).

Example 3: Bacterial growth inhibition by quench buffers in varying Q:B ratio.

In a further experiment, a microtiter plate reader was used to evaluate the efficiency of a wider range of quenchers and ratios with multiple

measurements over a longer period of time. The effects of quench buffers sarkosyl, urea, sarkosyl-urea and controls ethanol, water and RNAprotect on bacterial growth were assessed (Figures 2 and 3). In the water control, a steep increase of about 0.6 units OD595 can be observed in all ratios of Q:B. The only other increases that are of this magnitude are those observed with the urea buffer and the ethanol control in the 1:8 ratio. These both show an increase in OD of 0.4 units. In other samples the OD595 also increased over time. For example in the sarkosyl buffer and sarkosyl-urea buffer, wherein the samples show an increase in OD595 of 0.1 after 200 minutes. Looking at the results of the other experiments (shown in Figures 3-6)), it is however unhkely that the observed increase in OD is due to bacterial growth. Example 4: Quantification of bacterial growth by 16s rRNA gene analysis. In another experiment, bacterial growth was quantified by targeting and amplifying the 16S rRNA gene of bacterial species. Different 16s rRNA gene primers provide a good way to assess the presence and prevalence of different bacteria in a mixed sample. Primer sets were designed to have a nucleotide sequence that only matches the sequence 16s RNA genes of one specific prokaryotic taxonomic group. With the use of real-time PCR, one can determine the presence of a bacterial group in the total DNA extracted from the sample. Depending on the specificity of the primers, detection can range from the domain level up to the species level. For the following experiment we made use of primer sets specific for the bacterial species that were added in an equal volume mixture to the series of quenchers in the ratios 1:4 and 1:8. DNA was collected from the mixtures after incubation at 37° for 30 minutes, 6 and 24 hours. In each graph (vide Figures 4 and 5), the starting ratio is indicated (as Start), which was obtained from control samples containing the bacterial mixture immediately frozen without quench buffer. The starting ratio indicates that there was 60% E. coli and 40% L. plantarum before incubation. The 0 minutes sample shows the ratio retrieved from samples that were immediately frozen after the bacteria were mixed with the quench buffer. It follows from Panel E) of Figure 5 that urea cannot suitable be used in an 1:8 ratio of Q:B for preserving the bacterial composition of the sample, as L. plantarum started to overgrow E. coli.

Instead, the sarkosyl buffer, sarkosyl-urea buffer and control buffer ethanol were able to preserve the bacterial composition and keep the total amount of bacterial DNA to a large extent constant.

Example 5: Flow cytometry: Live-dead assay

To further evaluate the quench buffers' abilities to kill bacterial cells, we subjected quenched bacterial cultures to flow cytometry analysis. The bacteria (L. plantarum) were exposed to the different quenchers for 5 minutes before being added to a mixture containing fluorescent dyes that distinguish hve from dead bacteria. The ratios (Q:B) used were 1:4 and 1:8, as the quench buffer:sample ratio in a reservoir of a device for sampling the GI tract of a subject, such as the IntelliCap device of Medimetrics, is designed to be in this range. In this experiment, we further included a new self-designed quench buffer sarkosyl-ethanol. The sarkosyl-ethanol buffer, sarkosyl buffer, sarkosyl-urea buffer and RNAprotect buffer were able to kill all of the L. plantarum in the mixture, in both Q:B ratios, whereas the urea buffer and the control ethanol buffer were unsuccessful in killing all bacterial cells (Figure 6). Example 6: DN A isolation and quencher effects on yield

For the 16s rDNA primer experiments as described hereinbefore, numerous samples were collected. This allowed us to compare two DNA isolation methods. DNA of samples was isolated either by columnar extraction using QIAamp® DNA mini kit, or by standard phenol-chloroform (PC) extraction. The isolates were measured with a Nanodrop spectrophotometer to determine the DNA quantity and quality. Firstly, this allowed us to compare the yields of the different extraction protocols. Isolation with the QIAamp® DNA mini kit gave an average yield of 2.16 ng/μΐ with a high standard deviation (SD) of 1.32 ng/μΐ. The PC extraction yielded an average of 263.77 ng/μΐ with an SD of 46.78 ng/μΐ. The concentrations in the extracts obtained by the QIAamp® DNA mini kit were so low that the absorbance at 260 and 280nm could not be measured appropriately because the values were close to or beyond the measurement threshold of the

spectrophotometer. Some samples even gave a negative A280 measurement. The A260 and A280 measurements of the PC extracted samples were within a normal range and gave an average A260/A280 ratio per sample of 2.06 with an SD of 0.05. These values indicate that there was not much protein contamination in these samples. While performing PCR analysis on the samples, we noticed that good PCR results could be obtained with both the QIAamp® DNA mini kit and by PC extraction. The samples obtained with the QIAamp® DNA mini kit could be used undiluted or lOx diluted, while the PC samples needed to be diluted lOOOx before stable PCR results could be obtained.

Considering that the average DNA concentration in the PC samples has a lower relative SD and that the samples have measurable A280 values, we chose these samples to evaluate the effects of the quenchers on DNA quahty and yield. Except for the samples with water, all samples yielded a significantly lower DNA yield than the control without quencher (Figure 7). When compared with water, only the DNA yield of the sample treated with sarkosyl buffer was not significantly lower. Looking at the A260/A280 ratio, samples treated with sarkosyl buffer, urea buffer and sarkosyl-urea buffer gave values significantly lower than the untreated samples. When compared with the water control, all the quenchers have a lower A260/A280 value. However, these values were still sufficient to perform 16S rRNA gene analysis.

Example 7: Experiments with IntelhCap® in-vitro test system

A further experiment was performed with the IntelliCap® in-vitro test system. A mixing test was conducted with the use of two chemoluminescent fluids retrieved from a Cyalume Snaphght® glow stick. When these fluids mix, the mixture emits light. This concept was used to assess visually whether quencher and sample would mix well in the IntelliCap®. The capsule was filled with 50μ1 of one of the fluids resembling the quench buffer and submerged in the other to aspirate it. Within 1 minute after aspiration started, the first light was emitted. Until the completion of the aspiration, the emitted light increased in strength. A second test was performed to confirm the aspiration capabihty of the IntelliCap® system. Glycerol was chosen as fluid to be aspirated because of its viscosity of 1.412 Pa · s, which is significantly higher than the viscosity of the luminal content of the intestine. The IntelliCap® was weighed before and after aspirating glycerol from a container. The differences in mass were 0.223 and 0.169 g, corresponding to aspirated volumes of 185 and 134μ1 respectively. Example 8: Human validation study: Predominant diet-microbiota

interactions in the human small intestine.

Study design

A randomized cross-over fully controlled feeding trial was performed on human subjects. Two intervention diets were used to induce a temporary changes in microbiota composition: a 4-day low carbohydrate/high-protein diet versus a high-carbohydrate/low protein diet. AH volunteers also received a medium protein/medium carbohydrate diet three days prior to the first intervention and three days between the interventions. These periods were included as a run-in and washout period. Faecal and capsule samples were taken at the end of both intervention periods (effectively at day 7 and 14). The study aimed to recover faecal and capsule samples from 10 volunteers. The 10 participants followed the full protocol. The study was approved by the Medical Ethics Committee of Wageningen University and performed according to the principles of the Declaration of Helsinki an accordance with the Medical Research Involving Human Subjects Act (WMO).

Diets

Diets were strictly controlled during the entire 2-wk study period.

Participants visited the research facility every working day during lunch. They consumed a hot meal, which was weighed to the nearest gram by the research dieticians. Breakfast, evening bread meals, snacks, beverages, and all meals for the weekends were provided in pre-calculated take-home packages. Participants were carefully instructed how to prepare take-home meals. Participants consumed foods covering 100% of their designated needs. Participants were instructed to eat all the provided food and not to change their physical activity pattern for the duration of the study.

Participants reported all deviations of study guidelines and were not allowed to drink alcohol during the complete trial. Body weight was measured every day on a cahbrated scale at the research facility and diets were adjusted if needed. Participants did not change in bodyweight during the intervention (p=0.64, paired t-test, data not shown). Diet quantification

All diets were designed for macronutrient composition and energy content by using Compleat (Food calculation program developed by the Division of Human Nutrition; Wageningen University, using NEVO 2013 online, RIVM, the Netherlands; http://nevo-online.rivm.nl; RIVM, the Netherlands). Three types of diets were designed, namely a high-protein/low-carbohydrate diet (26,7En% protein, 38.2En% carbohydrate), a low-protein/high-carbohydrate diet (7En% protein, 59.6En% carbohydrate) and a medium diet (16En% protein, 44.3En% carbohydrate). Diets were individually tailored to meet each volunteer's energy requirement (± 1 MJ/d). Total energy expenditure for each participant was estimated from BMI, physical activity level and basal metabolic rate was estimated by the Schofielcl equation.

All diets were provided as a 3-4-day menu cycle and consisted of normal food, enriched with commercially available protein powder (whey protein isolate; Syntrax Nectar Protein - Pink Grapefruit) or carbohydrate

supplements (syrup concentrate). Duplicate portions of a mean daily energy amount of 13MJ of each intervention diet were collected each day, pooled per intervention, homogenized and analysed for energy, macronutrient and dietary fiber composition as previously described (Altorf-van der Kuil, W. et al. Br J Nutr. 2013; 110(5):810-22).

Sample collection

After each of the intervention diets, all 10 participants came fasted to the research facility and swallowed the capsule (IntelliCap® system,

Medimetrics Personalized Drug Delivery B.V., Eindhoven, The Netherlands)

- containing 50 microliter of the sarkosyl-urea buffer as indicated in Table 1

- together with 250 mL of water. After ingestion of the capsule, pH and temperature data was monitored through the gastro-intestinal tract via a portable unit (Koziolek, M. et al. J Pharm Sci 2015; 104:2855-2863; Maurer, J. M. et al. PLoS One 2015; 10: eO 129076.)· Pyloric passage (range 3-168 min) was determined by a sharp rise in pH and was confirmed by ingestion of a small volume of ice cold water that did not affect the emitted

temperature. The subjects received a standardized liquid breakfast as soon as pyloric passage of the capsule was confirmed in order to stimulate the production of gastro-intestinal juices. A luminal sample was taken from the small intestine when the capsule was located in the ileum, which was defined as 30 min after reaching a pH plateau (range 78 -261min after pyloric passage). After actuation of the capsule for luminal fluid sampling, subjects were required to stay at the research facihty until pH data from the IntelhCap® system indicated passage of the capsule via the ileocecal valve into the cecum. From that points onwards, participants kept the portable unit as close to the body as possible and collected all faecal samples until recovery of the capsule.

Participants were instructed to collect all stool samples using a Fecotainer® (AT Medical BV, Enschede, The Netherlands). A portion of the stool sample without the IntelliCap® capsule was labelled and frozen within 4 hours after defecation. The remaining stool sample was stored in a portable refrigerator and delivered at the research facihty within 24 hours. Upon delivery, capsule recovery was checked and participants were informed as soon as the capsule was retrieved from the stool sample. Most capsules (15/20) were excreted around 1 day after ingestion of the capsule, and the other 5 capsules were excreted 2-4 days after ingestion. Faecal samples from the faeces in which the capsule was excreted, were frozen within 4 hours after defecation, and were used for the microbiota analysis. No adverse events regarding swallowing of the capsule were reported during the study. Results

Little is known about the microbiota composition and function in the small intestine (SI) and its responsiveness to (digestible-) dietary ingredients, which is largely due to the invasiveness of sampling procedures that target this region of the human intestinal tract. The use of a minimally-invasive capture device (the IntelliCap® CR system, Medimetrics Personalized Drug Delivery B.V., Eindhoven, The Netherlands) for sampling the SI luminal content and assessing its microbial composition was validated. Using this device, it was possible to study the impact of dietary changes on SI microbial composition in vivo in healthy human subjects.

The microbiota composition of the capsule, and the faecal sample in which the capsule was recovered, was analyzed by 16S rRNA gene sequencing using Illumina MiSeq technology (Illumina Inc, San Diego CA, USA).

Microbiota composition was compared between sample locations and between diets.

Capsule derived samples clustered separately from the faecal samples, supporting their distinct microbial composition (Figures 8 A and 8B). As anticipated, species diversity was much higher in faecal compared to capsule samples. Analysis of these differences revealed that several bacterial groups contributed to the distinction between capsule and faecal samples, irrespective of the diet (Table 2). Many of the microbial famihes found to be present at higher relative abundance in the capsule samples relative to the faecal samples have previously been detected as inhabitants of the SI, including Streptococcus, Veillonella, Gemella, Rothia,

Granulicatella, and Heamophilus. Conversely, the microbial groups with significantly higher relative abundance in faecal samples are well known members of the large intestinal microbiota, like the genus Bifidobacterium and specific members of the Lachnospiraceae, although other genera of this family were enriched in the capsule samples (Lachnospira, Roseburia and Anaerostipes). These observations indicate that the capsule in combination with the sarkosyl-urea quencher as described in Table 1 can be used to reliably and safely sample and preserve the SI chyme in healthy human volunteers, thereby enabhng the investigation of the human SI microbiota using a minimally invasive strategy.

Table 2. LDA score of the significantly different relative abundances of bacterial taxa in capsules (small intestine) as compared to faeces (large

Table 2 lists the bacterial taxa with a significantly different relative abundance between the capsule derived samples and fecal samples, irrespective of the diet, as analysed by Linear discriminative analysis Effective Size (LEfSe), at different taxonomic levels. The direction (faeces or capsule) indicates the location of the sample with the highest relative abundance. The LDA score (logarithmic value) is an indication for the effect size. The p-value indicates the significance of the effect. Bacterial taxa mentioned in this results section are indicated in grey. It follows from Table 2 that the microbial population in the small intestine (capsule) contains microbial groups that have previously been detected as inhabitants of the SI. Conversely, the microbial groups with significantly higher relative abundance in faecal samples are examples of well known members of the large intestinal microbiota. Comparison of the impact of the two diets on the small and large intestine microbiota identified several bacterial groups that were differentially modulated by the LC-HP and HC-LP dietary regimes. In particular the LC- HP diet appeared to be associated with a significantly increased relative abundance of the phylum Firmicutes in the SI, although a similar increase was observed in faecal samples (not significant). Within the Firmicutes phylum, the genera Lactobacillus and Coprococcus were significantly enriched in the SI in the LC-HP diet, whereas Dorea and Streptococcus were significantly enriched in the faecal microbiota. However, although not in all cases meeting the significance criteria, the changes of these genera appeared to be strongly conserved in both SI and faecal microbiota, illustrating a high congruency of the diet induced microbiota composition changes in the small and large intestine microbiota. Moreover, only a single family of bacteria (F16; associated with the TM7 phylum) displayed an opposite dietary response in SI and faecal microbiota, although the enrichment of this family during the LC-HP was only significant in faecal microbiota. The HC-LP diet appeared to be associated with the significant modulation of only a single genus (Selenomonas) in the SI, and although not significant, the direction of change of this genus was conserved in faecal samples. These results imply that at taxonomic level the microbiota composition changes induced by the diets is strongly conserved in the SI and faecal samples.

To further analyse the diet induced changes, we predicted the differential functional composition of the microbiota in the samples using the freely available bioinformatics software package PICRUSt (Phylogenetic

Investigation of Communities by Reconstruction of Unobserved States version 1.0.0), focused on microbial functions associated with "metabolism". These analyses strongly confirmed the high level of congruency of the diet- induced microbiota changes in the SI and faecal populations at the functional level. The microbial functions that were predicted to be enriched by the LC-HP and HC-LP dietary regimes were displaying highly consistent changes in both the SI and faecal microbiota, although for individual functions this observation was not always meeting the significance criteria in both samphng locations. Moreover, the diet-induced function differences that were significant in both locations were consistently assigned to the same OTUs in the SI and faecal microbiota.

Intriguingly, a quite distinct enrichment of bacterial primary carbohydrate transport functions, belonging to the phosphotransferase system (PTS) was observed when comparing the HC-LP and LC-HP diets. PTS import of carbohydrates is highly efficient and is directly linked to metabohc activity of the cell through phosphoenolpyruvate derived phosphorylation of the incoming carbohydrate. The LC-HP diet appeared to induce an enrichment of PTS functions associated with the import of cyclic mono- and

disaccharides (mannose, glucose, sucrose, lactose and trehalose) in both the SI as well as the feacal microbiota. In contrast, the HC-LP diet led to a significant enrichment of PTS functions related to the import of hnear polyols (glucitol, sorbitol, mannitol, and sorbose) in only the faecal microbiota, although these functions also appeared to be enriched in the SI, but were not significant in that location. These findings imply that bacterial groups with PTS functions that import mono- and disaccharide sugars have a selective advantage under conditions where these sugars are in low abundance in the diet (LC-HP), while this 'advantage' is no longer selective under dietary regimes that encompass high levels of these 'simple sugars' (HC-LP). Conversely, diets rich in carbohydrates (HC-LP) also contain higher levels of linear polyols that are not absorbed by the small intestine mucosa, and remain available for microbial fermentation in the distal regions of the ileum as well as the colon, selecting for bacterial groups that can effectively import these polyols.

Taken together, these findings strongly indicate that the diet modulates the microbiota species as well as function composition predominantly in the SI. Although similar diet-induced changes can be detected in the faecal microbiota (both in terms of composition as well as predicted function), these faecal changes are very likely due to effects that were elicited in the SI. The detection of these effects in the faecal microbiota is probably due to progression of the SI microbiota to the large intestine, eventually ending up in the faecal material. This conclusion is in good agreement with the notion that that majority of the dietary ingredients of the human diet are digestible and absorbable in the SI, and do not directly interact with the large intestinal microbiota.