HAFNIA ALVEI FORMULATIONS

Abstract

A composition essentially made of a Hafnia alvei probiotic strain expressing the ClpB protein; wherein the ClpB protein is in an amount of at least 0.7% (w/w) in weight relative to the total weight of the composition; and the ratio of the total number of Hafnia alvei Colony Forming Units to the total Hafnia alvei cell number ranges from 10.sup.−4 to 0.8. Also, oral dosage forms, namely gastro-resistant capsules including the composition of essentially made of a Hafnia alvei probiotic strain expressing the ClpB protein.

Claims

1-15 (canceled)

16. A composition essentially consisting of Hafnia alvei probiotic strain; said strain expressing the ClpB protein; wherein: the ClpB protein is in an amount of at least 0.7% (w/w) in weight relative to the total weight of the composition; and the ratio of the total number of Hafnia alvei Colony Forming Units to the total Hafnia alvei cell number ranges from 10.sup.−4 to 0.8.

17. The composition according to claim 16, wherein the number of Hafnia alvei Colony Forming Units cells is equal or superior to 10.sup.6 per gram of composition.

18. The composition according to claim 16, wherein the total number Hafnia alvei cell number is equal or superior to 10.sup.10 per gram of composition.

19. The composition according to claim 16, wherein the Hafnia alvei strain is freeze-dried.

20. A pharmaceutical or nutraceutical composition, comprising from 5 to 30% (w/w) of the composition according to claim 16, said pharmaceutical or nutraceutical composition further comprising at least one pharmaceutically or nutraceutically acceptable excipient.

21. The pharmaceutical or nutraceutical composition according to claim 20, wherein said at least one pharmaceutically or nutraceutically acceptable excipient is selected from a group consisting of at least one anti-adherent, at least one texturizing agent, and combinations thereof.

22. The pharmaceutical or nutraceutical composition according to claim 21, wherein said at least one anti-adherent is magnesium stearate.

23. The pharmaceutical or nutraceutical composition according to claim 21, wherein said at least one texturizing agent is a modified starch.

24. The pharmaceutical or nutraceutical composition according to claim 20, further comprising zinc and/or chrome.

25. The pharmaceutical or nutraceutical composition according to claim 24, wherein the zinc and/or chrome are in the form of organic salts. 26 (New). The pharmaceutical or nutraceutical composition according to claim 20, said composition comprising: from about 10% to about 15% (w/w) of a Hafnia alvei composition essentially consisting of Hafnia alvei probiotic strain; said strain expressing the ClpB protein; wherein: the ClpB protein is in an amount of at least 0.7% (w/w) in weight relative to the total weight of the composition; and wherein the ratio of the total number of Hafnia alvei Colony Forming Units to the total Hafnia alvei cell number ranges from 10.sup.−4 to 0.8; from about 80 to about 85% (w/w) of modified starch; from about 0.5 to about 1.5% (w/w) of magnesium stearate; from about 2.0 to about 3.0% (w/w) of a zing organic salt selected from zinc bisglycinate; and from about 0.01 to about 0.03% (w/w) of a chrome organic salt selected from chrome picolinate; in weight relative to the total weight of the composition.

27. An oral dosage form selected from capsules and tables, said dosage form comprising the pharmaceutical or nutraceutical composition according to claim 20.

28. The oral dosage form according to claim 27, said oral dosage form being coated with an enteric coating.

29. The oral dosage form according to claim 27, said oral dosage form being in the form of capsules.

30. The oral dosage form according to claim 27, said enteric coating comprising hydroxypropyl methyl-cellulose and gellan gum.

31. A blister comprising at least one oral dosage form according to claim 27.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0186] FIG. 1 is a diagram showing the pH-profile during the experiments under fed conditions with the Simulator of the Human Intestinal Microbial Ecosystem. The pH of the medium was controlled automatically. Arrows indicate the time and corresponding pH of samples taken during the stomach incubation phase (ST0 and ST2) and small intestine incubation phase (SI1, SI2, and SI3).

[0187] FIG. 2 is a graph showing the average log (CFU)±stdev (n=3) obtained through spread plating on LB agar (A). Average log (count)±stdev (n=3) of viable bacterial cells (B), non-viable bacterial cells (C) and total bacterial cells (D) obtained through flow cytometry. Data are representative for samples collected from the bacterial powder (Product) and those collected during passage in the stomach (ST0 and ST2) and small intestine (SI1, SI2, and SI3) under fed conditions. Differences in samples (ST0/ST2/SI1/SI2/SI3) as compared to their preceding sample were indicated with *.*: statistically significant change (p<0.05).

[0188] FIG. 3 is a graph presenting the difference in log (CFU) obtained through spread plating on LB agar (A) and difference in log (count) of viable bacterial cells (B) over three different time spans, i.e. the stomach, small intestinal, and overall GIT incubation during fed conditions. Statistical differences between ST2-Product, SI3-ST2, and SI3-product were calculated. *: statistically significant change (p<0.05).

[0189] FIG. 4 is a graph presenting Average log (CFU)±stdev (n=3) obtained through spread plating on LB agar (A). Average log (count)±stdev (n=3) of viable bacterial cells (B) obtained through flow cytometry. Data are representative for samples collected from the mucin beads after 1 h (SI1), 2 h (SI2), and 3 h (SI3) of small intestinal incubation under fed conditions with mucin beads.

[0190] FIG. 5 is a graph showing the High fat diet (HFD) validation. A. Body weight (in g) in mice fed with a high fat/high carbs diet (HFD) (n=67) and in mice fed with a control diet (Ctrl) (n=8).

[0191] FIG. 6 is a graph showing the ClpB levels in plasma (above, 6A) and feces (below, 6B) measured after the administration of treatment A and the comparative treatment.

[0192] FIG. 7 is a graph showing the relative hormone-sensitive lipase protein (pHSL) expression rate (against actine expression rate as a standard) in obese HFD mice treated with composition A, comparative treatment and control treatment. hormone-sensitive lipase protein and actine expression rates were measured by western blot.

[0193] FIG. 8 is a graph presenting the fat mass gain (in g) in high fat diet (HFD)-induced obese mice treated with composition A, comparative treatment and control treatment.

[0194] FIG. 9 is a graph showing the efficiency in the inhibition of bodyweight gain of Ob/Ob mice by the G2 treatment (4 10.sup.9 CFU of Hafnia alvei per gram of the composition) according to the present invention. 2-way ANOVA, p=0.041, Bonferroni post-test, Control vs H. alvei, p<0.05

[0195] FIG. 10 is a graph showing the dose-dependent improvement of the lean mass to fat mass ration by the treatment of Ob/Ob mice with compositions of the present invention. Kruskal-Wallis, Dunn's post-test, $$$ p<0.001, $$ p<0.01

[0196] FIG. 11 is a graph showing the pixel density of the ClpB protein (96 kDa) and bioactive fragments thereof (70 kDa, 40 kDa, 37 kDa and 25 kDa) past the Gastro-Duodeno-Ileal Model simulation within the oral dosage form according to the invention.

EXAMPLES

[0197] The present invention is further illustrated by the following examples.

Example 1

Hafnia alvei Survival and Activity in the Simulator of the Human Intestinal Microbial Ecosystem

[0198] This example shows the evaluation of the intestinal fate of a strain of Hafnia alvei during passage through the complete gastrointestinal tract (GIT). First, the viability and functionality of the bacterial strain during passage through the upper GIT under fed conditions, when dosed as a powder formulation, was determined. To do this, under very controlled simulated conditions, the Simulator of the Human Intestinal Microbial Ecosystem (SHIME®) was used. Two sets of upper GIT experiments were performed. During the first set of experiments the passage of H. alvei through the fed upper GIT in the absence of a mucosal layer was tested. During the second set of experiments, a mucosal layer was introduced in the small intestine which allowed to study the capacity of the strain to adhere to the gut wall under relevant physiological conditions. The main end-points were the quantification of culturable bacterial cells (CFU) through spread plating and the quantification of viable and non-viable bacterial cells through live/dead flow cytometry and this both on the luminal samples and mucosal samples. After passage of H. alvei through the upper GIT under fed conditions in the absence of a mucosal layer, the small intestinal suspension was transferred to sterile colonic incubations. This allowed to study the growth and metabolic activity of this bacterial strain under proximal colon simulating conditions. The main end-points were the quantification of culturable bacterial cells through spread plating and the quantification of viable and non-viable bacterial cells through flow cytometry. The metabolic activity of the bacteria strain was assessed by measuring of the pH of the medium and by quantifying the concentrations of short chain fatty acids (SCFA), branched chain fatty acids (BCFA), ammonium, and lactate.

[0199] The reactor setup was adapted from the SHIME, representing the gastrointestinal tract (GIT) of the adult human, as described by Molly et al. (Molly, K., M. V. Woestyne, et al. 1993 Applied Microbiology and Biotechnology 39: 254-258.). The SHIME consists of a succession of five reactors simulating the different parts of the human gastrointestinal tract.

[0200] The first two reactors are of the fill-and-draw principle to simulate different steps in food uptake and digestion, with peristaltic pumps adding a defined amount of SHIME feed and pancreatic and bile liquid, respectively to the stomach and small intestine compartment and emptying the respective reactors after specified intervals. The last three compartments—continuously stirred reactors with constant volume and pH control—simulate the ascending, transverse and descending colon. Retention time and pH of the different vessels are chosen in order to resemble in vivo conditions in the different parts of the gastrointestinal tract.

Test Product

[0201] The strain of Hafnia alvei was tested to assess its survival and the production of a target protein, while passing through the stomach and small intestine. To each stomach reactor 10.sup.10 CFU of H. alvei, formulated as a powder, was added. The ratio of the of Hafnia alvei CFU to the total Hafnia alvei cell number was of 0.32.

[0202] All experiments were performed in biological triplicate to account for biological variability.

Upper GIT Study

[0203] Gastric Phase (Fed State) [0204] Incubation during 2 h at 37° C., while mixing via stirring, with sigmoidal decrease of the pH profile from 5.5 to 2.0 (FIG. 1). [0205] Pepsin is supplied with the activity being standardized by measuring absorbance increase at 280 nm of TCA-soluble products upon digestion of hemoglobin (reference protein). [0206] Addition of phosphatidylcholine. [0207] Addition of the SHIME® nutritional medium containing arabinogalactan, pectin, xylan, starch, glucose, yeast extract, peptone, mucin, and L-cystein-HCl. The salt levels recommended by the consensus method (NaCl and KCl) were implemented. [0208] Sampling: t=0 and 2 h; at these time points the pH of the medium was equal to 5.5±0.05 and 1.99±0.05, respectively.

[0209] Small Intestinal Phase (Fed State) [0210] While mixing via stirring, the pH initially automatically increases from 2.0 to 5.5 within a period of 5 minutes after which a gradually increasing pH from 5.5 to 7.0 during an incubation of 3 h at 37° C. is controlled automatically by the software (as shown in FIG. 1). [0211] Regarding pancreatic enzymes both a raw animal pancreatic extract (pancreatin) containing all the relevant enzymes in a specific ratio as well as defined ratios of the different enzymes can be used. [0212] Regarding bile salts, 10 mM bovine bile extract is generally supplemented (bovine bile is a closer match to human than porcine in terms of tauro- and glycocholate) [0213] Addition of mucin coated microcosms to simulate the small intestinal mucus layer (only during one set of upper GIT experiments). [0214] Sampling: t=1 h, 2 h, and 3 h; at these time points the pH of the medium was equal to 6.5±0.05, 7.0±0.05, and 7.0±0.1, respectively.

CFU Counting

[0215] Samples were collected at different stages of the experiment for stomach and small intestine to determine the number of colony forming units of H. alvei by spread plating. The number of colony-forming units of H. alvei contained in the dry powder (product) were also determined. During the experiments with mucus beads, beads were harvested from the reactor. The mucin beads were first washed in a PBS solution. Subsequently, the mucin beads were incubated for 30 min at 37° C. in PBS containing 1% Triton X-100. Ten-fold dilution series were prepared from these samples in phosphate.

Quantification of Viable and Non-Viable Bacterial Cells by Flow Cytometry

[0216] Samples were collected at different stages of the experiment for stomach and small intestine to determine the number of viable and non-viable H. alvei cells by flow cytometry. The number of viable and non-viable H. alvei cells, present in the dry powder (product), were also determined. During the experiments with mucus beads, beads were harvested from the reactor. The mucin beads were first washed in a PBS solution. Subsequently, the mucin beads were incubated for 30 min at 37° C. in PBS containing 1% Triton X-100. A ten-fold dilution series was initially prepared in phosphate buffered saline. Assessment of the viable and non-viable population of the bacteria was done by staining the appropriate dilutions with SYTO 24 and propidium iodide. Samples were analyzed on a BDFacs verse. The samples were run using the high flow rate. Bacterial cells were separated from medium debris and signal noise by applying a threshold level of 200 on the SYTO channel. Optimization of the proper PMT settings and construction of appropriate parent and daughter gates allowed to determine all populations of interest. Results are reported as average log (counts)±stdev of the three independent biological replicates.

[0217] Statistically significant differences between the number of CFU and counts of viable and non-viable bacterial cells were determined in between each sampling point and its preceding one during the experiments under fed conditions to demonstrate changes in function of time.

[0218] The same differences were determined for the results obtained for the product and the first sampling point of the experiment to demonstrate the immediate influence of the environmental conditions on the survival of the bacterial strain. Furthermore, statistically significant differences were determined for the bacterial strain over three different time spans, i.e. the stomach, small intestinal, and overall gastrointestinal incubation to clearly demonstrate the effect of the residing environmental conditions on the culturability and viability of H. alvei. In terms of statistics, the differences for all data discussed and indicated by “p<0.05” or “*” were significant with a confidence interval of 95%, as demonstrated using a Student's t-test.

Upper GIT Results

[0219] Sampling of the stomach reactor immediately after dosing of the bacterial strain indicated that the number of culturable bacterial cells was equal to the number of culturable bacterial cells present in the product. The same results were obtained for the number of viable, non-viable, and total bacterial cells (FIG. 2). This revealed that there was no immediate effect on the “culturability” and viability of H. alvei once this bacterial strain came into contact with the gastric juice. Under fed conditions the initial pH of the stomach is high (pH value of 5.5) due to the buffering capacity of the ingested food. The continuous secretion of hydrochloric acid in the stomach environment surpasses this initial buffering effect resulting in a sigmoidal decreasing pH of the stomach till a final value of 2.0 over a two-hour period. These low pH values can impose a major stress on bacterial survival in the stomach. Indeed, after 2 h of stomach incubation the number of culturable bacterial cells of H. alvei was decreased and the number of non-viable bacterial cells increased.

[0220] After passage through the stomach the bacterial cells enter the small intestinal incubation phase which is marked by a sharp increase in the environmental pH till a value of 5.5. Throughout the small intestinal incubation phase the pH further increases till a final value of 7.0.

[0221] Notwithstanding the presence of these beneficial pH conditions in the small intestine, the secretion of bile acids in the small intestinal lumen generally imposes a major challenge on bacterial survival.

[0222] Surprisingly, the high concentration of bile acids did not result in a decrease in the number of culturable and viable bacterial cells of H. alvei. Indeed, after 1 h of small intestinal incubation the number of culturable and viable bacterial cells of this strain increased till the end of the stomach incubation. This indicated that H. alvei was not sensitive to bile acids and was even capable to consume the carbohydrate substrates, present in the fed upper GIT, finally resulting in the growth of this bacterial strain.

[0223] Throughout the experiment the total number of bacterial cells remained constant indicating that no cell lysis occurred during passage through the fed upper GIT.

[0224] In view of the assessment of the altered levels of the bacterial cells during the stomach (ST2-Product), small intestinal (SI3-ST2) and overall GIT incubation (SI3-Product) are presented in FIG. 3. These results indicated that the number of culturable and viable bacterial cells decreased over the course of the stomach transit. Hence, the strain of H. alvei is sensitive to the low pH conditions of the stomach. During the upper GIT, a net increase in the number of culturable and viable bacterial cells occurred indicating that H. alvei was not sensitive to the high concentrations of bile acids in the small intestine and was capable to grow during the small intestinal transit. As such, the number of culturable and viable bacterial cells of H. alvei was not significantly decreased after a full passage through the upper GIT under fed conditions.

[0225] The SHIME assay repeated in the present of mucin beds confirmed the above results. The determination of the number of culturable bacterial cells and viable bacterial cells from samples collected from the initially sterile beads revealed that H. alvei was capable to adhere to the mucin beads and this already after 1 h of small intestinal incubation. Furthermore, this bacterial strain remained attached to the mucin beads after 2 h and 3 h of small intestinal transit (FIG. 4).

Growth and Metabolic Activity in the Colon

[0226] Short-term Colonic Batch Incubations

[0227] Short-term colonic batch incubations were performed using a representative colon medium containing host- and diet-derived compounds. To all colonic batch incubations, a centrifuged and autoclaved SHIME suspension was added to provide the bacteria with relevant colonic metabolites. After passage of H. alvei through the upper GIT under fed conditions (in the absence of mucus beads), a part of the small intestinal liquid phase was transferred to colonic reactors containing the colon medium and the sterile SHIME suspension All bottles were incubated for 48 h at 37° C. under anaerobic conditions.

[0228] Samples were taken at the start of the incubations (0 h) and after 24 h and 48 h of incubation. The growth of H. alvei under these sterile colonic conditions was determined through quantification of the number of CFU by spread plating and the quantification of the number of viable and non-viable bacterial cells through flow cytometry.

[0229] The fermentative activity of H. alvei in the colon was studied by determining the pH of the medium in the colonic reactors. Furthermore, concentrations of acetate, propionate, butyrate, branched chain fatty acids, lactate and ammonium were determined. The experiments were performed in biological triplicate to account for possible biological variability.

[0230] Results

[0231] During the short-term colonic incubations, H. alvei was capable to grow under proximal colon simulating conditions since during the first 24 h the number of culturable and viable bacterial cells increased (log CFU: T.sub.0: 7.21; T.sub.24 h: 8.92 and T.sub.48 h: 8.02). In between 24 h and 48 h of colonic incubation the number of culturable and viable bacterial cells decreased. This was mainly due to a conversion of H. alvei from a culturable into a VBNC (viable but non culturable) state since the number of viable bacterial cells decreased less than the number of culturable bacterial cells (log[count of viable cells]: T.sub.0: 7.46; T.sub.24 h: 9.06 and T.sub.48 h: 8.75). This conversion could be due to the lowering of the pH of the medium (T.sub.0: 6.23; T.sub.24 h: 5.96 and T.sub.48 h: 5.77) or due to the absence of carbohydrates which were completely consumed during the first 24 h of incubation.

[0232] The metabolic activity of H. alvei in the colon was confirmed by the increase of lactate concentrations throughout the colonic incubations (T.sub.0: 0.7 mM; T.sub.24 h: 1.77 mM and T.sub.48 h: 2.11 mM)

[0233] The results of Example 1 show that intact H. alvei cells can vectorize ClpB past the acid conditions of the stomach, since no lysis was observed, and ensure the short-term delivery of ClpB.

[0234] Surprisingly, contrary to other bacterial strains that are inactivated by the gastric conditions, H. alvei CFUs attain the proximal intestine where they proliferate and ensure the colonization of the distal parts of the GIT. Hence, the composition according to the invention shall further ensure a more prolonged secretion of ClpB via the CFUs having attained the stationary bacterial growth phase in the colon.

Example 2

In Vivo Effect of the Invention's Composition on HFD Mice

[0235] This example demonstrates the effect of the Hafnia alvei CFU/total cell ratio on high fat diet-induced obese mice.

[0236] One-month-old male C57B16 mice (Janvier Laboratories) were induced with high fat/high carbs diet for 4 weeks. Induction of obesity by high fat diet was validated by measurement of mean body weight (FIG. 5) in a group induced and a group non-induced for obesity.

[0237] Mice were then intragastrically gavaged with as follows:

TABLE-US-00002 TABLE 1 Treatment groups of the in vivo experiment 2. CFU/total Treatment Strain CFU Total Cells cells ratio A H. alvei 4.0 10.sup.7 8.4 10.sup.8 0.04 Comparative Inactivated 0 4.8 10.sup.8 0 treatment H. alvei Control (MH culture — — — medium)

[0238] The presence of ClpB in the treatment compositions in treatment A and in the comparative treatment was confirmed by Western-Blot.

[0239] Mice were placed individually into the BioDAQ cages (Research Diets) and intragastrically gavaged daily for 6 weeks. At the end of the treatment the mice were euthanized and tissue samples (plasma, colic fecal, epididymal fat) were collected.

[0240] The inventors showed that the comparative treatment did not induce the presence of ClpB in the mice plasma (FIG. 6A) or feces (FIG. 6B), despite the presence of ClpB in such treatment

[0241] Furthermore, contrary to the composition according to the invention, the comparative composition failed to induce the pHSL expression (FIG. 7) and improve the body composition (fat mass gain inhibition, FIG. 8)

Example 3

In Vivo Effect of the Invention's Composition on Ob/Ob Mice

[0242] Batch N0717031A of Hafnia alvei composition according to the invention was used to prove the in vivo effects of the composition according to the present invention.

[0243] In view of obtaining a negative control, a sample of the batch was inactivated by pasteurization (negative control). The ClpB quantification in the negative control showed a ClpB concentration of 1.56 mg per gram of freeze-dried composition and the CFU/total Hafnia alvei cells ratio was 2.2 10.sup.−4.

[0244] Genetically obese ob/ob mice (n=5×15) were acclimated to the animal facility for 1 week. Mice were intragastrically gavaged twice a day for three weeks with the probiotic treatments presented in table 1. At the end of the experiment, mice were euthanized and intestinal and epididymal fat tissue samples were collected.

TABLE-US-00003 TABLE 2 Treatment groups of the in vivo experiment 3. mg Treatment Strain CFU Total Cells CFU ClpB/g G1 H. alvei 4597 .sup. 4.0 10.sup.10 .sup. 5.5 10.sup.10 0.72 9.3 G2 (Batch 31A) 4.0 10.sup.9 5.5 10.sup.9 Negative Inactivated 8.7 10.sup.3 4.0 10.sup.9 2.2 10.sup.−7 1.56 control 1 H. alvei 4597 (Batch 32A) Control (vehicle) — — — —

[0245] At the end of the treatment the G2 was evaluated for the body weight gain control as a proof of concept. Indeed, Ob/Ob mice presented a significant reduction in weight gain compared to the control group, as presented in FIG. 9.

[0246] Further analyses were carried out on the body weight analysis. As presented in FIG. 10, the present results confirmed that the treatment with G1-G2 induce a reduction of the body fat mass percentage and the amelioration of the lean mass to fat mass ratio. Interestingly, the improvement in the body composition occurred in a CFU and total cell number dose-dependent manner (G1>G2).

[0247] Even more interestingly, the dose dependent effect correlates not only with the ClpB concentration but also with the number of the administrated Hafnia alvei CFU and the ratio of the CFU count over the total Hafnia alvei cells count.

Example 4

Batch Production with Improved CFU Count

[0248] Given the results of Example 3, different Hafnia alvei bioreactor conditions were tested in view of optimizing the CFU count relative to the total Hafnia alvei cell count.

[0249] Preliminary assays showed that the incubation period had no or little effect the CFU count. Furthermore, among the tested conditions, oxygen stress had no beneficial effect on the CFU count. On the contrary, continuous heat stress (38° C.) reduced the CFU count in the bioreactor of laboratory scale. Surprisingly, as presented in table 1 initial 38° C. heat stress or terminal 15° C. stress while the rest of the cell culture maintained at 35° C., presented an improvement of the CFU count.

TABLE-US-00004 TABLE 3 Tested bioreactor conditions and CFU count. CFU during Week: Assay 0 2 4 Control 1.16 10.sup.12 6.00 10.sup.11 5.90 10.sup.11 A. Terminal 15° C. stress 1.03 10.sup.12 7.05 10.sup.11 5.70 10.sup.11 B. Initial 43° C. stress 1.32 10.sup.12 6.75 10.sup.11 5.20 10.sup.11 C. Initial 38° C. stress 1.16 10.sup.12 6.90 10.sup.11 6.60 10.sup.11 D. Continuous 38° C. stress 2.70 10.sup.11 1.10 10.sup.10 1.65 10.sup.10

TABLE-US-00005 TABLE 4 CFU and total cells per gram of obtained compositions under the tested laboratory scale bioreactor conditions. CFU/g of Cells/g of CFU/total composition composition cell ratio Control 1.16 10.sup.12 1.94 10.sup.12 0.60 A. Terminal 15° C. stress 1.03 10.sup.12 1.33 10.sup.12 0.77 B. Initial 43° C. stress 1.32 10.sup.12 1.78 10.sup.12 0.74 C. Initial 38° C. stress 1.16 10.sup.12 2.06 10.sup.12 0.56 D. Continuous 38° C. stress 2.70 10.sup.11  4.6 10.sup.11 0.58

[0250] Bioreactor conditions A and B were retained since they presented an improved CFU/total cell ratio as presented in table 4.

[0251] The quantity of the ClpB protein and fragments thereof was quantified by means of pixel densitometry on the immunoblotted freeze-dried material (Loaded proteins: 50 μg; primary anti-α-MSH primary antibody (polyclonal rabbit; Delphi Genetics)—dilution 1/1000; secondary anti-rabbit-HRP secondary antibody (Dako)—dilution 1/5000; Diluant & blocking buffer TBST-BSA 5%; Exposition time=8 seconds. Films were scanned using ImageScanner III (GE Healthcare) and analyzed for the band pixel density using the ImageQuant TL software 7.0 (GE Healthcare)).

[0252] Stress assays A, B improved the CFU/total cell ratio in addition to the ClpB concentration.

[0253] Based on those results, two batches were prepared on an industrial scale bioreactor.

[0254] CED01 batch characterized by: [0255] about 9 mg of ClpB per gram of lyophilized composition; [0256] 5 10.sup.11 CFU per gram of lyophilized composition; and [0257] 9.7 10.sup.11 total Hafnia alvei cells per gram of lyophilized composition.

[0258] CED02 batch characterized by: [0259] about 15 mg of ClpB per gram of lyophilized composition; [0260] 2.7 10.sup.11 CFU per gram of lyophilized composition; and [0261] 1.12 10.sup.12 total Hafnia alvei cells per gram of lyophilized composition.

Example 5

Oral Dosage Form

[0262] Four oral dosage forms we prepared comprising the CED01 batch as hereinbefore described and presented in table 5.

TABLE-US-00006 TABLE 5 Screened oral dosage forms. Mass % Mass per Form Capsule Constituents (w/w) capsule(g) M+/HPMC HPMC H. alvei CED01 10.6 0.05 Capsule Pregeflo ® 81.4 0.3852 Methocell K100M ® 7.0 0.0331 Magnesium stearate 1.0 0.0047 Total 100 0.4730 M−/HPMC HPMC H. alvei CED01 10.6 0.05 Capsule Pregeflo ® 88.4 0.4183 Magnesium stearate 1.0 0.0047 Total 100 0.4730 M+/DrCaps ® DrCaps ® H. alvei CED01 10.6 0.05 Pregeflo ® 81.4 0.3852 Methocell K100M ® 7.0 0.0331 Magnesium stearate 1.0 0.0047 Total 100 0.4730 M−/DrCaps ® DrCaps ® H. alvei CED01 10.6 0.05 Pregeflo ® 88.4 0.4183 Magnesium stearate 1.0 0.0047 Total 100 0.4730

Example 6

Gastro-Duodeno-Ileal Model (GDIM)

[0263] The oral dosage forms of example 5 were subjected to a GDIM assay. The results of this example allow the selection of the excipients as well as the coating agent or capsule used for an efficient oral administration of the composition according to the present invention.

[0264] In brief, humified dosage forms of example were subjected to incubation through three compartments, each simulating the gastric, duodenal and ileal content: [0265] Gastric compartment T.sub.0-T.sub.30  44 mL of a solution consisting of NaCl (3 g/L), KCl (1.1 g/L), CaCl.sub.2 (0.15 g/L), pepsin (951 U/mg; 0.003% w/w) Citrate/phosphate buffer 32 mM, pH 3.5. 37° C., rotation at 150 rotations per minute. [0266] Duodenal compartment T.sub.30-T.sub.60  65 mL of a solution consisting of NaCl (3 g/L), KCl (1.1 g/L), CaCl.sub.2 (0.15 g/L), bile (0.3% w/w), Trypsin (7500 U/mg 0.007% w/w), Di-sodium hydrogen phosphate dihydrate buffer 200 mM. pH 6.5, 37° C., rotation at 150 rotations per minute. [0267] Ileal compartment T.sub.60-T.sub.120  65 mL of a solution consisting of NaCl (3 g/L), KCl (1.1 g/L), CaCl.sub.2 (0.15 g/L), bile (0.3% w/w), Trypsin (7500 U/mg; 0.0056% w/w), Di-sodium hydrogen phosphate dihydrate buffer 200 mM. pH 7.0, 37° C., rotation at 150 rotations per minute.

[0268] Past the GDIM assay, samples (n=3) were homogenized in aseptic conditions and the viability of Hafnia alvei past the GDIM was assessed by flow cytometry. The latter analysis showed that all four dosage forms sufficiently protected Hafnia alvei through the gastric pass.

[0269] The obtained samples were then lyophilized and the ClpB content was measured in the obtained powders, by densitometry of the Western-blot as previously detailed.

[0270] Surprisingly, the formulation of the invention coated with a hydroxypropylmethylcellulose and gellan gum coating (DrCaps®) yielded the best stability of ClpB during the simulated digestion.

[0271] More in particular, as shown in FIG. 11, the formulation devoid of texturizing agent (hydroxypropylmethylcellulose, Pregeflo®) showed the a pronounced gastric-resistance not only for the ClpB protein (˜96 kDa) but also for the bioactive ClpB fragments (˜96 kDa, ˜70 kDa, ˜40 kDa, ˜37 kDa and ˜25 kDa).