In vitro evaluation of the anti-pathogenic activity of Okoubaka aubrevillei on the human gastrointestinal tract.

BACKGROUND: Okoubaka aubrevillei is used in traditional West African medicine and in homeopathy for treatment and prevention of several gastrointestinal problems. The aim of this in vitro study was to evaluate the effect of repeated doses of two Okoubaka products (10 % ethanolic tincture, mother tincture (MT); 3rd decimal potency, 3X) on the microbial activity of physiological human colon microbiota using a Simulator of the Human Intestinal Microbial Ecosystem (SHIME®) and to investigate any preventive effect against infections with diarrhea-causing pathogens.
METHODS: Upon inoculation with fecal microbiota from a healthy donor, 4 parallel proximal colon compartments of the SHIME were treated either with Okoubaka MT, Okoubaka 3X, ethanol control or blank control for 7 days. Using the Okoubaka-adapted microbial community from SHIME, 48 h challenge tests were performed with enterotoxigenic Escherichia coli (ETEC) and Salmonella enteritidis in 4 different doses (103-108 colony forming units as typical in vivo infectious doses). Pathogen concentrations, short-chain fatty acids (SCFAs) and branched SCFA production were measured in triplicate at 0, 24 and 48 h.
RESULTS: In the challenge tests, both Okoubaka products were able to restrict the colonization of ETEC and Salmonella at 3 of the 4 pathogen doses (except the highest doses), with a stronger anti-pathogenic effect for MT, which included a reduction of 2.0 log-units of ETEC (p < 0.0001) and 1.1 log-units of Salmonella (p < 0.0001). Total SCFA levels remained unaffected, but butyrate increased during the first 24 h (p < 0.0001 for ETEC), accompanied by decreased acetate production.
CONCLUSION: We observed in vitro a systemic activating effect of Okoubaka on intestinal microbiome resistance, which resulted in an anti-pathogenic effect, especially against ETEC. We hypothesize that the mode of action in vivo is also based on systemic regulative effects. Thieme. All rights reserved.

Introduction

Okoubaka ( Okoubaka aubrevillei Pellegr. et Normand) is a West African tree from the plant family Santalaceae. The bark was initially used as a remedy in traditional West African medicine, e. g., to prevent food poisoning 1 . As an herbal preparation, it is an ingredient of malaria remedies used in African ethnomedicine 2 .

Magdalena Kunst, a German doctor, brought Okoubaka bark from one of her journeys in 1972 back to Germany 1 . Okoubaka developed into an established and registered homeopathic medicinal product over the last 40 years. The raw material is monographed in the German Homeopathic Pharmacopoeia (GHP) 3 , and a commission D (advisory board of the German competent authority, Federal Institute for Drugs and Medical Devices) monograph was published in 1989 4 .

Okoubaka is known to act on the gastrointestinal tract 5 6 7 8 and is used, usually as mother tincture (MT, 10 % ethanolic tincture of the bark, prepared according to the European Pharmacopoeia method 1.1.8) or in low homeopathic potencies (e. g., 3X, corresponding to a 10 −3 dilution) in the prevention and treatment of traveler’s diarrhea 9 .

Due to its tannins and catechins content, the bark acts as a tanning agent with its typical astringent effect and probably also by adsorbing toxins and bacterial enterotoxins. Further immunomodulating effects, e. g., an increase of about 20 % of phagocytosis in granulocytes tests, were observed for the macerates in water and ethanol and for Okoubaka MT 10 11 . Kreutzkamp 10 tested various fractions of the separated bark (watery macerate; ethyl acetate fraction of a complex extraction) against gram-positive and gram-negative bacteria (e. g., E. coli ) in an agar dilution test. The author concluded that probably no antibiotic substances were present in Okoubaka as the evaluated minimum inhibitor concentrations (between 256 and 1024 µg/ml) were too high compared with typical antibacterial drugs (approximately 1 µg/ml) 10 .

Because of the unique homeopathic manufacturing methods (i. e., potentization by stepwise dilution and succussion), the final concentration of active substances in Okoubaka potencies is rather low (dilution of active substances 10 −3 for 3X) and cannot explain its adsorbing effects and its resulting prophylactic property – even for the MT, which includes 10 % of starting material. It is assumed that homeopathic medicinal products do not act in a pharmacological dose-dependent way but as a stimulation and regulation therapy by activating the body’s self-healing capacities 12 13 14 15 . According to the similarity principle (“Similia similibus curentur”), homeopathic substances treat complaints characterized by signs and symptoms that are similar to those that these substances would induce in a healthy subject 16 .

Okoubaka is used according to its homeopathic “drug picture” for the treatment of several health problems and diseases 6 8 17 . However, the main focus of the remedy is on gastrointestinal problems, as mentioned above 4 .

Our study aimed to evaluate the effect of homeopathically prepared Okoubaka as MT and potency (3X). For this purpose, an established in vitro model of the human intestinal tract was used (Simulator of the Human Intestinal Microbial Ecosystem, SHIME ® ) 18 . Our working hypothesis was that the intestinal microbial community works as a self-regulating system and that treatment with Okoubaka products increases intestinal microbiome resistance and the bacterial defense/immune mechanism against pathogens.

Changes in microbial activity result in changes in the production of short-chain fatty acids (SCFAs, concentrations of acetate, propionate, and butyrate) and branched SCFAs (b-SCFAs). Therefore we measured propionate and butyrate, which contribute to salt and water homeostasis in the colon and a healthy large bowel epithelium 19 . Furthermore, acetate and lactate, which some bacteria convert to butyrate (“cross-feeding”) 20 , as well as b-SCFAs (the sum of isobutyrate, isovalerate, and isocaproate) and ammonium, which are signs of the proteolytic activity (protein fermentation) of the gut microbiota 21 , were measured.

In a second step, we aimed to simulate, by challenge tests, an infection that might occur, for instance, during acute diarrhea while traveling. Therefore the Okoubaka -adapted microbial community from the SHIME ® was used and challenged with Salmonella enteritidis or enterotoxigenic Escherichia coli (ETEC), as both pathogens are known for causing traveler’s diarrhea 22 23 . Pathogen concentrations and SCFA and b-SCFA production were measured during the 48-hour challenge tests.

Material and methods

All experiments were performed in the laboratories of ProDigest (Gent, Belgium), a research organization specializing in pre-clinical gastrointestinal research.

In vitro modeling of the gastrointestinal tract

The SHIME system simulates the different parts of the gastrointestinal tract of the adult human in 5 reactors and was first described by Molly et al. 24 . The first 2 reactors represent the stomach and small intestine. Special conditions of food intake and digestion are simulated with peristaltic pumps, and a defined amount of SHIME feed and pancreatic and bile juices are added. Further, the reactors are emptied after defined time intervals (“fill and draw” principle). The next 3 reactors simulate the 3 colon compartments of the large intestine and contain a fecal sample from a healthy donor. The reactors are continuously stirred, contain a constant volume, and are pH-controlled. Inoculum preparation, retention times, pH, temperature settings, and SHIME feed composition represent in vivo conditions and have been previously described by Possemiers et al. 25 .

For the present experiments, an adapted SHIME system with a simulation of 4 parallel running proximal colon compartments was used ( Fig. 1 ). This part of the colon was chosen because it is closest to the small intestine, where the infections with the pathogens causing diarrhea ( Salmonella and ETEC) occur.

Fig. 1

 Experimental design.

The colon reactors of each arm were inoculated with a fresh fecal sample from a healthy donor (who had not taken antibiotic treatment during the previous 6 months, did not have a history of gastrointestinal disorders, and had a BMI below 25) and simulated the conditions in the proximal colon (pH 5.6–5.9; retention time 20 hours; volume 500 ml; 12 % concentration of fecal inoculum).

During the experiment, the following parameters were monitored to identify changes in microbial community composition and activity:

Acid/base consumption

pH was continuously monitored during the SHIME experiment to ensure the pH range required in the SHIME reactors (pH 5.6–5.9 for proximal colon). During the treatment period, bacteria might produce increased amounts of SCFAs, resulting in acidification of the reactor. To keep the pH within the pre-set intervals, a base was added that further led to increased base consumption. As a possible rough indicator for changes in the microbial fermentative activity, acid/base consumption and overall fermentative activity were continuously monitored during the SHIME experiment by the built-in SHIME software. Data were exported and used for further evaluation at the time points on which samples for the other parameters mentioned below were taken to get corresponding consumption values for these time points.

SCFA and lactate concentration

Changes in microbial activity result in changes in SCFA production and lactate concentration. SCFAs were monitored by gas chromatography with flame ionization detection (GC-FID, as described by De Weirdt et al. 26 ) as a concentration of acetate, propionate, and butyrate on days 2, 4, and 7 of the treatment period and in triplicate. Lactic acid determination was conducted at the same time points using a commercially available enzymatic assay kit (Enzytec™ D-/L-Lactic Acid; R-Biopharm, Darmstadt, Germany), according to the manufacturer’s instructions.

b-SCFA and ammonium concentration

b-SCFAs (sum of isobutyrate, isovalerate, and isocaproate) were monitored by GC-FID as described by De Weirdt et al. 26 on days 2, 4, and 7 of the treatment period, in triplicate. Ammonium concentrations were determined by a steam distillation approach followed by titration, as previously described by Duysburgh et al. 21 .

Experimental design

The following experimental procedure was performed ( Fig. 1 ):

2-week stabilization period;

1-week “prophylactic” treatment period in SHIME system, followed by

48-hour challenge tests with 2 pathogens in different doses.

Stabilization period

After a 2-week stabilization period, a steady-state status in the SHIME system was reached, controlled by acid/base consumption (pH) and SCFA production (acetate, propionate, and butyrate).

A minimum 20 % coefficient of variation (CV), corresponding to 80 % reproducibility (100 % – CV) 25 of SCFA levels (acetate, propionate, and butyrate) between the different colon units, indicated that the starting microbial community among them was similar, which was important for comparison purposes between the 4 different compartments. We aimed to achieve a reproducibility above 85 %.

“Prophylactic” treatment period

After the stabilization period, one proximal colon compartment of the SHIME reactor was treated with Okoubaka MT and one with Okoubaka 3X (3 rd decimal dilution [potency]) in doses of 5 drops (corresponded with 120 µl) 3 times per day. This dosage regimen corresponds to the recommended dose, adapted to the conditions in the SHIME system, which correspond in turn to half an adult human (the dosage for 1 human consists of a maximum of 5 drops 6 times per day according to the package leaflet). In the remaining 2 proximal colon compartments, 1 control arm with no treatment (blank control) and 1 ethanol control (62 % [m/m]) were included. In all 4 compartments, treatments/controls were given for 1 week.

Challenge tests

After the treatment period, aliquots were taken from both the “ Okoubaka adapted” (MT and 3X) and “ethanol-controlled” microbial community in SHIME and used for challenge tests with 2 different pathogens in small reactors. An initial dose of 3 drops Okoubaka MT/3X or ethanol was added at the beginning of the 48-hour challenge tests in the corresponding sample from the SHIME system.

A 2 % inoculum of overnight grown Salmonella enteritidis (brain heart infusion broth as growth medium) or ETEC (Lysogeny broth as growth medium) – both prepared by Oxoid (Cambridge, UK) – were used.

The pathogens were added immediately after the initial dose of the test products and ethanol control in 4 different concentrations: 10 3 , 10 5 , 10 6 , and 10 8 CFUs for ETEC and 10 3 , 10 4 , 10 5 , and 10 6 CFUs for Salmonella enteritidis . The typical in vivo infectious doses were extrapolated to the conditions in the small reactors (50 ml; 37°C; under shaking and anaerobic conditions), which correspond to the colon of 1/10 of an adult human.

All experiments were done in triplicate. Anti-pathogenic activity and metabolic activity were determined at time points 0, 24, and 48 hours after taking samples of 2 ml volume. Changes in concentrations of both pathogens were measured during the experiment using different methods of colonization:

Salmonella enteritidis colonization

For the assessment of Salmonella enteritidis colonization, selective enumeration on MacConckey agar was used, containing 100 ppm streptomycin, which is selective for gram-negative bacteria. Briefly, samples were collected after 0, 24, and 48 hours to determine the number of CFUs of the bacterial strain tested by spread plating. Ten-fold dilution series were prepared from these samples in PBS, subsequently transferred to Petri dishes containing selective agar media and incubated aerobically at 37 °C for at least 24 hours.

ETEC colonization

For the quantitative assessment of ETEC colonization, the quantitative polymerase chain reaction (qPCR) using a QuantStudio 5 Real-Time PCR system (Applied Biosystems, Foster City, CA USA) was applied, as plating alone is not selective enough for ETEC in the colonic medium. Therefore, CFUs correspond to 16S RNA copies for ETEC.

Cell pellets were generated after 0, 24, and 48 hours of incubation and were subjected to propidium monoazide (PMA) treatment. PMA is an ionic substance that can only cross the damaged cell membranes of non-viable cells and prevent, after UV-induced crosslinking, the amplification of this DNA in qPCR assays. Subsequently, DNA was extracted from PMA-treated cells followed by quantification of their concentration through qPCR using primers that were selective for ETEC 27 .

Heat-labile toxin produced by ETEC

ETEC is capable of producing 2 different types of enterotoxins (proteins): 2 heat-labile toxins (LT1 and LT2) and a heat-stable toxin (ST). In our experiments, we measured the production of the heat-labile ETEC-toxin LT1, which acts like the cholera toxin, via ganglioside GM1-enzyme-linked immunosorbent assay (GM1 ELISA). The method described by Salimian et al. 28 was used with the following minor modifications. Both the cell wall-bounded LT1 and the one present in the fermentation liquid (supernatant) were determined. To do so, we centrifuged 1 ml of sample, after which the supernatant and the cell pellet were separated. The cell pellet was resuspended in phosphate-buffered saline (PBS) and sonicated before toxin quantification through GM1 ELISA. All experiments were done in triplicate.

Test products

Okoubaka MT was prepared according to the European Pharmacopoeia method 1.1.8 and method 4a of GHP using ethanol 62 % (m/m) 3 29 . For the raw material, the “dried bark of the aerial parts of Okoubaka aubrevillei Pellegr. et Normand” 3 was used. Okoubaka 3X was prepared according to the European Pharmacopoeia monograph 2371, starting from MT (corresponding to the 1 st decimal dilution) stepwise to 3X 29 . The potentization steps were done in ethanol 62 % (m/m) according to Hahnemannian principles with “10 vigorous downwards strokes” 30 against an elastic body as described in the original literature 31 . For the ethanol-control, ethanol 62 % (m/m) was used.

All test materials were prepared by Deutsche Homöopathie-Union DHU-Arzneimittel GmbH & Co. KG (Karlsruhe, Germany).

The greatest possible distance between the test products and the ethanol control was ensured during the experiments to prevent interactions between them. Also, the Okoubaka test products and the ethanol control were stored in different rooms.

Endpoints

Treatment period

To detect changes in the microbial community during treatment with Okoubaka products, the following parameters were measured and normalized to the ethanol control on days 2, 4, and 7 of the treatment period in triplicate: acid/base consumption, SCFAs (acetate, propionate, and butyrate), lactate concentration, b-SCFAs (sum of isobutyrate, isovalerate, and isocaproate), and ammonium concentration.

The following parameters were evaluated in the challenge tests with the 2 different pathogens (normalized to the ethanol control):

Direct anti-pathogenic activity of the Okoubaka -adapted microbial community was assessed via colonization of both pathogens ( Salmonella enteritidis and ETEC) 24 and 48 hours after challenge with the pathogens.

Concentration of the heat-labile ETEC toxin LT1 was evaluated at the time points 0, 24, and 48 hours.

Changes in microbial metabolic activity were investigated at time points 0, 24, and 48 hours of incubation using an analysis of SCFA production (acetate, propionate, and butyrate) and b-SCFA production (sum of isobutyrate, isovalerate, and isocaproate).

Data management/statistical analysis

All variables are described groupwise by mean, median, standard deviation, range (min, max), and 95 % confidence interval (CI), as appropriate.

Data for the anti-pathogenic activity of the challenge tests were statistically analyzed by 2-way analysis of variance (ANOVA) with repeated measurements, with and without interaction terms. Group-specific differences were estimated and expressed as mean ± 95 % CI. A p-value of < 0.05 indicated statistical significance. P-values were not adjusted in terms of multiple testing.

Analyses were performed with software SAS 9.4; in addition, MS Excel was used for some graphical presentations.

Results

Stabilization period: stability of the SHIME setup

SCFA levels of each of the 4 arms of the SHIME system had a reproducibility of 90.1 % (100 % – CV) on average (acetate 91.9 %, propionate 92.7 %, butyrate 85.7 %) at the end of the stabilization period, which was higher than the targeted minimum value of 85 %.

Treatment period: “ Okoubaka -adapted” microbial community activity

Overall fermentative activity: acid/base consumption

A temporary trend toward increased base consumption was observed at the start (day 2) of the treatment by both Okoubaka products (compared to the ethanol control), followed by a decrease at the end (day 7) of the treatment period: 19.4 ml/day for Okoubaka MT and 10.4 ml/day for Okoubaka 3X on day 2 of the treatment period; −2.0 ml/day for Okoubaka MT and −1.9 ml/day for Okoubaka 3X on day 7 of the treatment period.

SCFA production and lactate concentration

Some – but not statistically relevant – changes in the “ Okoubaka -adapted” microbial community compared to the ethanol control were observed. The changes – determined at days 2, 4, and 7 of the treatment period – concerned the following parameters: acetate increase, trends toward increased propionate levels and decreased butyrate levels, and lactate increase (data not shown).

b-SCFA and ammonium

The levels of the sum of b-SCFAs and ammonium were not strongly affected by the Okoubaka treatment. The coefficient of variation (CV) was 5.4 % for b-SCFA levels and 4.5 % for ammonium levels between all 4 arms of the SHIME system (ethanol control, blank control, Okoubaka MT, and 3X).

Challenge Tests: anti-pathogenic activity

Fig. 2 shows the colonization of ETEC (using qPCR) after 24 and 48 hours of incubation. For graphical representation, test data (done in triplicate) were normalized to the ethanol control.

Fig. 2

 Results of the quantitative colonization of ETEC (using qPCR) normalized to ethanol control after 24 and 48 hours of incubation, upon quantification via plating on MacConkey agar supplemented with 100 ppm streptomycin; 16 s RNA copies correspond to CFU.

There was a reduction of colonization compared with ethanol control in 3 out of the 4 pathogen doses for Okoubaka MT and 2 out of the 4 pathogen doses for Okoubaka 3X. The strongest effects were observed for the lowest ETEC dose (10 3 CFU) with a reduction of 2.0 log-units of ETEC for treatment with MT and of 1.6 log-units of ETEC for 3X after 24 hours. The anti-pathogenic effects were almost similar after 48 hours. At both time points, the effects decreased for higher pathogen doses in a dose-dependent way. No anti-pathogenic effects and even a trend towards increased colonization of ETEC were observed at the highest ETEC dose (10 8 CFU). The values are shown in Table 1 .

Table 1 also shows the values estimated with a 2-way repeated measurement ANOVA with treatment and dose as fixed factors for ETEC. The estimated effect on colonization reduction for all pathogen doses collectively was significant for Okoubaka MT (p < 0.0001) and 3X (p = 0.0041).

Toxin production by ETEC

There were no significant effects of the Okoubaka test products on heat-labile ETEC toxin LT1 (results not shown). The measured enterotoxin concentrations after application of either Okoubaka MT or Okoubaka 3X were in the range of toxin levels of an active ETEC culture (21.4 ng/mL in the supernatant).

Fig. 3 shows the colonization of Salmonella enteritidis after 24 and 48 hours of incubation.

Fig. 3

 Results of the quantitative colonization of Salmonella enteritidis normalized to ethanol control after 24 and 48 hours of incubation, upon quantification via plating on MacConkey agar supplemented with 100 ppm streptomycin.

For graphical representation, test data (done in triplicate) were normalized to the ethanol control.

Both test products ( Okoubaka MT and 3X) compared with the ethanol control restricted the growth of Salmonella enteritidis , except at the highest Salmonella concentration (10 6 CFU), with the strongest anti-pathogenic effect during the first 24 hours of incubation. A stronger anti-pathogenic effect was observed for Okoubaka MT with a reduction of 1.1 or 1.0 log-units of Salmonella for the 2 lower pathogen doses (10 3 or 10 4 CFU, respectively). The values are shown in Table 2 .

Table 2 also shows the values estimated with a 2-way repeated measurement ANOVA with treatment and dose as fixed factors for Salmonella . The estimated effect on colonization reduction for all pathogen doses collectively was significant for Okoubaka MT (p < 0.0001).

SCFA and b-SCFA production in ETEC

Total SCFA production – evaluated at time points 0, 24, and 48 hours during the challenge tests – remained essentially unaffected after treatment with both Okoubaka products, although there were changes in single fatty acid composition.

The changes for butyrate, acetate, propionate, and b-SCFAs during the time intervals 0–24 hours and 24–48 hours are shown in Fig. 4 . Increased production of butyrate during the first 24 hours was observed. The butyrate increase for Okoubaka MT and 3X, normalized to ethanol control, was statistically significant for the 0–24 hours time interval; data and p-values are shown in Table 3 . Butyrate levels remained high during the 24–48 hours time interval. Overall acetate levels increased at the time interval 0–24 hours, followed by a decrease at the 24–48 hours time interval. Normalized to ethanol, acetate levels of Okoubaka MT and 3X decreased during the 0–24 hours time interval, i. e., reflecting acetate consumption. The decrease was statistically significant for lower doses of ETEC; data and p-values are shown in Table 3 .

Fig. 4

 Values for butyrate, acetate, propionate, and b-SCFAs [mmol/l] for MT, 3X, and EtOH during challenge with ETEC for the time intervals 0–24 hours and 24–48 hours.

Levels of butyrate increase and acetate decrease (difference 24–0 hours) normalized to EtOH during 0–24-hour time interval for MT and 3X during challenge with ETEC.

ETEC dose[CFU]	butyrate levels normalized to EtOH0–24 hours time interval (Diff. 24 hours–0 hours) [mmol/l]	lower CL[mmol/l]	upper CL[mmol/l]	p-value	acetate levels normalized to EtOH0–24 hours time interval (Diff. 24 hours–0 hours)[mmol/l]	lower CL[mmol/l]	upper CL[mmol/l]	p-value
MT and 10 3	0.64	0.48	0.79	< 0.0001	−1.64	−2.08	−1.84	< 0.0001
MT and 10 5	0.62	0.46	0.78	< 0.0001	−1.99	−2.43	−1.54	< 0.0001
MT and 10 6	0.51	0.36	0.67	< 0.0001	−0.44	−0.89	0.00	0.0544
MT and 10 8	0.60	0.44	0.78	< 0.0001	0.21	−0.24	0.65	0.3590
3X and 10 3	0.83	0.67	0.98	< 0.0001	−0.59	−1.04	−0.15	0.0105
3X and 10 5	0.55	0.40	0.71	< 0.0001	−2.28	−2.73	−1.83	< 0.0001
3X and 10 6	0.56	0.40	0.71	< 0.0001	−1.18	−1.62	−0.73	< 0.0001
3X and 10 8	0.42	0.27	0.58	< 0.0001	0.15	−0.30	0.59	0.5121

SCFA/b-SCFA production in Salmonella enteritidis

Total SCFA production remained unaffected upon treatment with both Okoubaka products, but there were changes in single fatty acid composition.

The changes for butyrate, acetate, propionate, and b-SCFAs during 0–24 hours and 24–48 hours time intervals are shown in Fig. 5 . Increased production of butyrate during the first 24 hours was observed. The values of butyrate increase for Okoubaka MT and 3X, normalized to ethanol control, are shown in Table 4 and were less pronounced as compared to ETEC. Butyrate levels remained high during the 24–48 hours time interval. Overall, acetate levels increased at the time interval 0–24 hours, followed by a decrease at the 24–48 hours time interval. Normalized to ethanol, acetate levels of Okoubaka MT and 3X slightly decreased during the 0–24 hours time interval, i. e., reflecting acetate consumption. The postulated cross-feeding effect from acetate to butyrate was less pronounced; data are shown in Table 4 .

Fig. 5

 Values for butyrate, acetate, propionate, and b-SCFAs [mmol/ l] for MT, 3X, and EtOH during challenge with Salmonella enteritidis for the time intervals 0–24 hours and 24–48 hours.

Levels of butyrate increase an acetate decrease (difference 24–0 hours) normalized to EtOH during 0–24-hour time interval for MT and 3X during challenge with Salmonella enteritidis .

Salmonella dose [CFU]	butyrate levels normalized to EtOH0–24 hours time interval (Diff. 24 hours–0 hours) [mmol/l]	lower CL[mmol/l]	upper CL[mmol/l]	p-value	acetate levels normalized to EtOH0–24 hours time interval (Diff. 24 hours–0 hours)[mmol/l]	lower CL[mmol/l]	upper CL[mmol/l]	p-value
MT and 10 3	0.23	0.00	0.46	0.0530	−0.64	−1.12	−0.16	0.0094
MT and 10 4	0.46	0.25	0.67	< 0.0001	−0.21	−0.63	0.22	0.3308
MT and 10 5	0.35	0.15	0.56	0.0013	−0.44	−0.87	−0.02	0.0415
MT and 10 6	0.07	−0.14	0.28	0.4869	0.14	−0.29	0.56	0.5192
3X and 10 3	0.24	0.04	0.45	0.0223	−0.55	−0.98	−0.13	0.0123
3X and 10 4	0.14	−0.07	0.35	0.1858	−0.30	−0.72	0.13	0.1690
3X and 10 5	0.40	0.19	0.60	0.0004	−0.87	−1.29	−0.44	0.0002
3X and 10 6	0.32	0.11	0.53	0.0035	−0.08	−0.50	0.35	0.7161

Discussion

During the experiments, the effect of Okoubaka MT and 3X on the human gut microbiota in terms of microbial activity was tested using a long-term SHIME setup to simulate a prophylactic treatment. Subsequently, in short-term challenge tests, the anti-pathogenic activity of the Okoubaka -adapted microbial community was evaluated against 2 diarrhea-causing pathogens (ETEC and Salmonella enteritidis ) in 4 different concentrations.

SHIME system: reproducibility (comparability) of the 4 arms of the SHIME system

With a 90.1 % reproducibility (calculated as 100 % – CV) of the SCFA levels between the 4 proximal colon units of the SHIME system at the beginning of the treatment period, there were similar starting conditions between the Okoubaka treatments, ethanol, and blank control arms. It was higher than the 80 % value generally accepted for the SHIME system established by Possemiers et al. 25 and even higher than the targeted 85 % for the present study. To rule out unequal baseline conditions as far as possible, the sample with the highest butyrate levels combined with the lowest acetate levels was allocated to the ethanol control. Hence, the ethanol control had the best starting conditions with regard to health-related butyrate of all 4 SHIME units.

In addition, ammonium and b-SCFAs remained quite stable during the 1-week treatment period. These parameters were essentially unaffected by the Okoubaka treatment and matched well in the 4 reactors at the 3 measurement time points (4.5 % CV for ammonium and 5.4 % CV for b-SCFAs).

Challenge tests: anti-pathogenic activity

Both Okoubaka products were capable of restricting the colonization of ETEC and Salmonella enteritidis at 3 pathogen concentrations up to 2.0 log-units for ETEC and 1.1 log-units for Salmonella . A reduction of 2.0 log-units implies a 99 % reduction of the pathogen, while a reduction of 1.0 log-unit corresponds to a 90 % reduction of the pathogen.

Some differences in the intensity of the effects shall be discussed in the following.

A stronger anti-pathogenic effect was observed for the MT. One explanation could be that the MT (defined as 1X potency according to European Pharmacopoeia method 1.1.8 29 ) contains 100-fold more starting material ( Okoubaka bark) compared with 3X. Such a concentration dependence was also observed in in vitro tests with other potentiated substances 32 33 . It could also be explained by the fact that in vitro tests – even if the bacterial community tested is a complex self-regulating system – should be interpreted within the boundaries of the models and are therefore not able to reflect the integrated conditions of an entire organism.

The anti-pathogenic effect was stronger for the lower pathogen concentrations, with a reduction of up to 2 log-units of ETEC and 1.1 log-units of Salmonella for the lowest pathogen doses (10 3  CFU/10 4 CFU). With increasing doses, the anti-pathogenic activity decreased, however. For ETEC, typical infectious doses between 10 3 and 10 8 germs are described 22 34 , ingested, for instance, by contaminated water or food. Both ETEC and Salmonella are very sensitive to the acid milieu in the stomach. Therefore patients with hypoacidity of the stomach or those treated with proton pump inhibitors are much more sensitive to infections with these pathogens 22 35 . For those patients, infectious doses of 1000 germs for ETEC and lower are possible. For Salmonella , the minimum infectious doses are even lower, between 100 (described for contaminated fatty food) and 10 6 germs 36 . To compare the SHIME in vitro model with a human, the typical infectious doses must be extrapolated to the conditions in the proximal colon: taking into account that, in the present in vitro model, the acidic gastric compartment has been bypassed and the conditions in the reactors corresponded to 1/10 adult, 100 to 10 000 CFU for ETEC, and 10 to 10 000 CFU for Salmonella are needed for them to be infectious. Therefore, the lower pathogen concentrations tested in the study – for which prophylaxis through Okoubaka was observed – reflect the typical range of bacterial infections.

For concentrations above about 10 6 to 10 7 CFU for ETEC and about 10 5 CFU for Salmonella , which under realistic conditions can only be reached when visibly spoiled food is consumed, the prophylactic administration of Okoubaka had no anti-pathogenic effect.

For the highest pathogen doses (10 8 for ETEC and 10 6 for Salmonella ), a trend for increased growth of the pathogens compared to the ethanol control was observed in the test model. Such stress-dependent effects have also been observed in other experimental systems with potentized remedies, e. g., in the treatment of arsenic- or mercury-stressed duckweeds with potentized Arsenicum album or Mercurius solubilis . In slightly stressed duckweeds, treatment with the isopathic remedy resulted in a growth enhancement (improvement), whereas in severely stressed duckweeds, the same treatment resulted in growth reduction 37 . Such inverse effects might be related to the specific mode of action of homeopathic remedies: each medicine produces an “artificial illness” (primary action, phase 1), to which the organism (secondary action, phase 2) reacts accordingly. If the artificial illness is (homeopathically) similar to the disease’s symptoms to be treated, there will be an improvement 13 . Thereby the regulation ability of the organism (by activating self-healing capacities) is a prerequisite. If the disease (in our model, the number of pathogens) is too strong, the system’s regulation ability may be overstrained. It is also conceivable that phase 1 and phase 2 are delayed: since the time frame in our system was limited to 48 hours, it could be that the secondary action would be observed after 48 hours.

Challenge tests: production of ETEC toxin

The Okoubaka -adapted microbial community was able to inhibit the growth of ETEC but could not help the degradation of already existing toxins. Toxin elimination via adsorbing effects through tannins is one of the discussed modes of action for Okoubaka bark. However, in the diluted preparations and even in the MT, the tannin concentration is too low to exhibit adsorbing effects.

As the exact tannin content of the tested MT is unknown and analytical testing not available, the tannin content was estimated as follows: by macerating 100 g dried Okoubaka bark in 1 liter methanol for 5 hours, Kreutzkamp isolated 3.2 g of the dry residual of catechins corresponding to a yield of 3.2 %. After extracting this residual in ethyl acetate following by distillation, 800 mg of tannins were obtained 10 11 . In the investigated Okoubala MT (10 % tincture of the bark in 62 % [m/m] ethanol, prepared by percolation), the proportion of dry residual is much lower (i. e., 0.4 %) 3 . A usual tannin-based preparation against diarrhea contains 500 mg tannin albumin per tablet or 250 mg tannins. The recommended dose is 1–2 tablets 4–6 times per day. This corresponds to a minimum of 1000 mg and a maximum 3000 mg tannins per day 38 . The amount of tannins in 3 drops MT is approximately 0.01 mg (rough estimation based on dry residual) and therefore 100 000 orders of magnitude lower. Therefore, our results are plausible and were to be expected.

Mode of action

We observed no relevant effects in the microbial community during treatment with Okoubaka products, as monitored on days 2, 4, and 7 of the treatment period regarding acid/base consumption, SCFAs, b-SCFAs, and ammonium. In the challenge tests, both Okoubaka products restricted the colonization of ETEC (up to 2.0 log-units) and Salmonella enteritidis (up to 1.1 log-units), but no effect was observed on the ETEC toxin.

We essentially see 2 main possibilities to explain these overall findings: by direct phytopharmacological effects and/or by systemic regulative effects on the bacterial community in the SHIME system. As discussed above, direct pharmacological effects by tannins seem rather unlikely as an overall explanation. A further argument for a systemic regulative effect is the comparably small difference for Okoubaka MT and 3X in ETEC colonization reduction and the comparable effects on SCFAs and b-SCFAs after 24 hours, though these 2 products differ by a factor of 100 in compound concentration. Though direct pharmacological effects cannot be excluded, we hypothesize based on the results obtained that the primary mode of action of Okoubaka in the present in vitro SHIME system is a systemic regulative effect on the bacterial community.

Possible implications for the application of Okoubaka in practice: anti-pathogenic activity

Based on the study results concerning anti-pathogenic activity, an extrapolation of the in vitro results obtained to in vivo conditions would lead to the following considerations. Reducing the pathogen concentration up to 99 % (2 log-units) would give the organism the possibility and time to eliminate the remaining pathogens. Therefore, symptoms of acute diarrhea would possibly not occur, or the course of the disease would be milder. Therefore, our experiments may deliver or be a part of an explanation for the traditional, empirically-based use of Okoubaka as prophylaxis before traveling.

Okoubaka seems to work better for prophylaxis of ETEC than for Salmonella enteritidis infections. This applies to the extent of the reduction of colonization of the pathogens (log-units), the level of significance of the results, the differences between MT and 3X, and the differences between the first and the second time interval of the challenge tests. Better effects for ETEC were also observed concerning the metabolic activity, especially for butyrate concentration.

One reason for these differences could be the different pathogenic mechanisms of the 2 bacteria. ETEC intoxications are caused by 2 different types of enterotoxins: 2 heat-labile toxins (LT1 and LT2) and a heat-stable toxin (ST). The toxins act in the small intestine supported by “colonization factor antigens” that are responsible for the adhesion of the bacteria to the mucosa of the small intestine. The heat-labile enterotoxin LT1 acts like the cholera toxin in the small intestine by stimulating adenylate cyclase, which leads to the secretion of water and chloride via cAMP-mediated secondary messenger reactions. The heat-stable enterotoxin works as a guanylate cyclase activator in the intestinal mucosa, resulting in cGMP-mediated loss of water and electrolytes 39 . These reactions result in symptoms like watery diarrhea with abdominal cramps; nausea and sub-febrile temperatures are also possible 39 .

Salmonella invades the mucosal cells of the small intestine and is phagocytosed by macrophages, which leads to an inflammatory reaction resulting in the excretion of water and electrolytes into the intestinal lumen 39 . Histological investigations show a massive immigration of neutrophils into the intestinal tissue 40 . Elimination of Salmonella from the body is difficult due to its penetration into the cells of the intestinal mucosa and its subsequent phagocytosis by the macrophages. Symptoms of Salmonella enteritidis infections are diarrhea, headaches, abdominal pain, nausea, and sometimes vomiting. These symptoms are often accompanied by fever. In rare cases, sepsis with high fever has been described 36 .

According to the homeopathic similarity principle, based on the various symptoms caused by these 2 pathogens, a different relationship between the 2 different types of enteritis and Okoubaka could be assumed, possibly explaining the different effect of Okoubaka on the 2 pathogens. For example, “cramping pain in abdomen; diarrhea; loose and frequent stools, as with a virus” were symptoms experienced during a homeopathic drug proving of Okoubaka 12C by Riley in 1990 41 , which fits better to ETEC than to Salmonella infections.

Metabolic activity: possible effects of SCFA increase

Changes in microbial activity result in changes in SCFA production. SCFA production of the microbial community is related to various positive health effects.

For SCFAs, diverse effects in the human intestinal tract have been discussed: acetate, butyrate, and also propionate are considered as favorable to health, i. e., with protective effects against inflammation and prevention of the colonization of bacterial pathogens 42 . Butyrate is produced, for example, by members of the Clostridium clusters XIVa (Firmicutes), Faecalibacterium prausnitzii, and Oscillibacter 43 , via cross-feeding, which is a process by which some bacteria convert acetate to healthful butyrate along with other substrates 20 . Butyrate plays an important role in the maintenance of the barrier function of the intestine. It reduces intestinal permeability through various mechanisms, closes tight junctions, and initiates repair mechanisms 44 45 .

An increased butyrate production at the expense of acetate (cross-feeding process) was observed during the challenge tests, even for the higher pathogen concentrations. However, there was no significant change in butyrate production during the 7-day prophylaxis. This increase of butyrate, produced by the Okoubaka -adapted microbial community during the pathogen challenge, might initiate butyrate-mediated host interactions against diarrhea. These effects might act in addition to the direct anti-pathogenic activity.

Strengths and limitations of this study

One strength of this study is the use of the SHIME technology, which is a validated experimental system 18 24 25 and which has been used for an extensive number of research projects in the past, as shown by more than 100 publications in PubMed when searching for the “Simulator of the Human Intestinal Microbial Ecosystem” 46 . In addition, by using a complex in vitro model of the human intestinal tract, the intestinal microbial community can be tested as a self-regulating system, thereby exceeding the limits of conventional in vitro experiments, such as using single cell lines.

This study also had limitations. Firstly, our results were based on the microbiota of only 1 donor. Further experiments with the microbiome of several donors are needed to investigate the generalizability of the results. Also, even if the SHIME technology is a comprehensive in vitro method, it cannot measure the complex interactions of an entire host organism. Furthermore, the time frame for the challenge tests in our system was limited to 48 hours, so possible delayed reactions could not be observed. As the evaluation’s main focus was to investigate the influence of the prophylactic administration of Okoubaka , only an initial dose (3 drops) of the Okoubaka products was applied at the beginning of the challenge tests. In reality, patients would continue treatment in situations of acute diarrhea past the stage of prophylaxis. This may be a plausible explanation for better anti-pathogenic and metabolic activity of Okoubaka in our in vitro system during the first 24 hours. Further experiments in which Okoubaka products are repeatedly given to the test system during challenge tests may address this hypothesis.

Conclusions

Both Okoubaka products had only limited effects on the metabolic activity of the human gut microbiome during the 1-week treatment period in the in vitro SHIME system. Statistically significant anti-pathogenic activity against ETEC and Salmonella enteritidis was observed in the in vitro challenge tests after prophylactic treatment of the microbial community with Okoubaka MT and 3X. The effects were more potent for lower pathogen doses, which correspond to typical infectious doses.

The results were concentration-dependent and slightly more pronounced for Okoubaka MT compared to the diluted Okoubaka 3X preparation. Okoubaka treatment seems to better fit for prophylaxis of ETEC infections than for Salmonella infections. Slightly positive effects on metabolic activity (change in SCFA production, especially increases in butyrate) were observed. These effects might act in addition to the direct anti-pathogenic activity and initiate effects, such as butyrate-mediated host interactions against diarrhea.

Based on the results obtained in the complex in vitro microbial community system, we hypothesize an activating effect on intestinal microbiome resistance for both Okoubaka products, which may provide – or be a part of – an explanation for the traditional, empirically-based use of Okoubaka as prophylaxis and treatment of traveler’s diarrhea in homeopathy.

Contributions

Susann Buchheim-Schmidt had the lead on the conception and the design of the study, contributed to the assembly, analysis, and interpretation of the data, and wrote the manuscript.

Uwe Peters contributed to the conception and the design of the study and data interpretation and reviewed the manuscript.

Cindy Duysburgh, Pieter Van den Abbeele, and Massimo Marzorati had the lead on the generation, collection, and assembly of data; contributed to the analysis and interpretation; and reviewed the manuscript. Cindy Duysburgh had the lead on the practical implementation of the study.

Thomas Keller led the analysis of data, contributed to the interpretation, and reviewed the manuscript.

Stephan Baumgartner, Petra Klement, and David Martin contributed to the conception and the design of the study, the data interpretation, and reviewed the manuscript.

All authors approved the final manuscript.

Funding

This investigation was funded by institutional resources only (DHU-Arzneimittel GmbH & Co KG) and received no further third-party funding.