A randomized trial to evaluate the impact of copra meal hydrolysate on gastrointestinal symptoms and gut microbiome

Preparation of copra meal hydrolysate and placebo drinks

Copra meal hydrolysate (CMH) was prepared following Rungruangsaphakun & Keawsompong (2018). The CMH powder contained 9.7% crude protein, 1% crude fat, 14.5% crude fiber, 12.3% ash and 53% nitrogen-free extract. The nitrogen-free extract was composed of 3-6 molecules of MOS, mainly mannose, linking together. Amount of total oligosaccharide of CMH was 14.77%; mainly mannooligosaccharide with 3-6 degrees of polymerization. The CMH drinks were prepared in three formulas: the placebo group (maltodextrin 10 g), the 3 g CMH group (3 g CMH and 7 g maltodextrin) and the 5 g CMH group (5 g CMH and 5 g maltodextrin). The products were packaged in individual daily-use sachets of identical appearance and completely dissolved in room temperature drinking water before consumption.

Study design

This study was a randomized double-blinded placebo-controlled trial with a three-arm parallel group design in 1:1:1 ratio as the placebo group, CMH 3 g (3CMH group) and CMH 5 g (5CMH group) (Fig. 1). Healthy adults, aged between 18 and 40 years with a body mass index (BMI) of 18.5 to 25.0 kg/m², were enrolled (Table S1). This study excluded those who (1) received any antibiotics or anti-viral agents within 3 months before the study, (2) received probiotics/prebiotics within 1 month before the study, and (3) were allergic to coconut and soybean which were components of the study product. Recruitment was conducted through poster and web-based advertising. The study was carried out at the outpatient clinic of King Chulalongkorn Memorial Hospital, Bangkok, Thailand. Physicians and study nurse who involved in the study explained the protocol and enrolled participants to the study. Written informed consent was obtained from all participants. The study was approved by the Ethics Committee of King Chulalongkorn Memorial Hospital, Bangkok, Thailand, under approval reference number 388/61 and registered on the Thai Clinical Trials Registry (TCTR20190426003).

Sample size calculation

The sample size was determined using the tests for two independent means (two-tailed test) as formula described below. On the basis of previously published study of the effect of MOS from coffee mannan on fecal microflora (Asano et al., 2006), the mean change of stool Bifidobacterium from baseline of subjects received MOS 3 g (µ1) was 20%, while for those received MOS 1g (µ2) was 10%. The difference (Δ) was used to calculate, with a power of approximately 80% (beta 0.2) and a significance level of alpha 0.05. The ratio (r) was 1. SD was estimated to be 8. The numbers of study participants to be included was 11 subjects per group. With the anticipation of 20% drop out rate, at least 14 subjects per group should be included. So, in our study, we included 20 subjects per group. ${\displaystyle n_1=\frac{{\left(z_{1-\frac{\alpha }{2}}+z_{1-\beta }\right)}^2\left[{\sigma }_1^2+\frac{{\sigma }_2^2}{r}\right]}{{\Delta }^2}}$ ${\displaystyle r=\frac{n_2}{n_1},\Delta ={\mu }_1-{\mu }_2}$

Randomization and allocation concealment

Each eligible participant was randomized to one of three groups and assigned a unique study number. A blocked randomization list was created using Sealed Envelope Ltd. (https://www.sealedenvelope.com), with block sizes of 6 by the statistician who was not involved in the study. Study nurse who did not aware of the code group assigned participants to interventions.

Intervention

After randomization, participants were subjected to a 2-week baseline period, 3-week treatment period and 2-week washout period with instruction to use the products once daily during the treatment period (Fig. 2). Participants were asked to fill out a daily stool record chart and a 24-hour diary record 3 times per week (2 weekdays and 1 day during weekend) throughout the study period. Dietary restrictions were imposed throughout the entire study period (total 49 days), with prohibition of consumption of other fermented products, probiotics, prebiotics and synbiotics supplements to limit potential interference with evaluation of the testing product. The participants were excluded from the study if (1) received dietary supplements described above during the study, (2) received any antibiotics or anti-viral agents during the study and (3) failed to collect stool samples according to the study protocol. Each participant visited the clinic weekly throughout the study. Stool samples were collected at the baseline visit, the end of treatment, and washout periods. Subjects and researchers who had access to the outcomes of the study were blinded to the code group assigned to participants.

Stool record chart

The stool record charts, adapted from Walton et al. (2010), were recorded daily by all participants throughout the study, in order to investigate whether the products affected the stool frequency, diarrhea, constipation, and gastrointestinal discomfort (for examples bloating symptoms) of the subjects (Table S2). Each stool characteristic was scored as a number and calculated as amean (±SD) per week.

24-hour dietary record

The 24-hour dietary records were filled in by the subjects three times per week (2 weekdays and 1 day during weekend) throughout the study period. Energy and nutritive values were calculated using the nutrient calculation computer software INMUCAL-Nutrients V4.0 database NB. 4 (Food Composition Database for INMUCAL Program, 2018) (Table S3).

Stool collection and processing

Fresh stools were collected at the baseline, at the end of the treatment period and the end of the washout period for analysis of microbiota, SCFAs and secretory immunoglobulin A (IgA). A stool sample of 20 g was collected in a 76 × 20 mm sterile container and immediately placed on ice for transfer to storage at −80 °C. Stool samples were prepared following the method of Kisuse et al. (2018) for microbiota and SCFA analyses. The fecal samples were diluted 10-fold with phosphate-buffered saline (pH 8.0) using a stomacher blender (Stomacher^® 80 Biomaster; Seward, Worthing, UK) for 5 min. A 1 ml sample of fecal slurry was then centrifuged at 13,000×g for 5 min., and the supernatant was stored at −80 °C for subsequent SCFA analysis. For microbial DNA extraction, 1 ml of fecal slurry was placed in a 1.5 ml centrifuge tube and then stored at −80 °C.

Microbial DNA extraction and 16S rRNA gene sequencing

Microbial DNA was extracted according to the modified method of Mirsepasi et al. (2014), utilizing a bead meter and a QIAamp^® DNA stool mini kit (Qiagen GmbH, Germany). Qualification and quantification of DNA were determined using a Nanodrop Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Extracted DNA samples were immediately stored at −20 °C.

The variable region of V3-V4 of 16S rRNA genes was amplified using the forward primer Imina-V3-V4-F (5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG- CCTACGGGNGGC WGCAG-3′) and reverse primer Imina-V3-V4-R (5′- GTCTCGTGGGCTCGGAGATGTGTAT AAGAGACAGGACTACTACHV- GGGTATCTAATCC-3′). Cycling conditions consisted of an initial denaturation at 94 °C for 2 min, followed by 25 cycles of denaturation at 94 °C for 20 s, annealing at 57 °C for 30 s, extension at 72 °C for 30 s, and a final extension at 72 °C for 10 min. The amplified products were purified using NucleoSpin^® Gel and PCR Clean-up (Macherey-Nagel Inc., USA) according to the manufacturer’s protocol and sent to the Omics Sciences and Bioinformatics Center (Faculty of Science, Chulalongkorn University, Bangkok, Thailand) for further sequencing using the Illumina MiSeq platform (USA).

Processing of 16S rRNA gene sequences and data analysis

The raw sequences were processed using bioinformatics tools. Paired-end reads were first quality trimmed using BBDUK (read quality >15 at 3′), and the primer at the 5′ end was removed using seqtk (https://github.com/lh3/seqtk). Any obtained sequence shorter than 150 bp was excluded along with its pair. Filtered pairs of sequences were corrected, merged and chimeras were removed using the DADA2 pipeline (v1.10) (Callahan et al., 2016). High-quality sequences were obtained at 41,843 ± 10,553 reads per sample. Taxonomy was identified using the QIIME2 classifier (v2019.1) (Bolyen et al., 2019) with the Greengenes version 13.8 database. Cutoff confidence was indicated at 0.7.

The PICRUSt2 pipeline package (v2.1.0_b) was used to predict the functional potential of gut microbiomes from observable 16S rRNA sequences. The KEGG pathway levels were inferred according to the instructions from the PICRUSt2 website (Douglas & Maffei, 2020).

Alpha and beta diversity analysis

Alpha diversity (Chao1, Shannon) and beta diversity were calculated using the Vegan package of R software (version 2.5.6). Statistical differences in the diversity indices among the three groups were identified using the Kruskal–Wallis H test. Beta-diversity among participants was visualized via non-metric multidimensional scaling (NMDS) ordination based on Bray–Curtis dissimilarity matrices.

Short-chain fatty acid analysis

Lactic acid, acetic acid, butyric acid, and propionic acid from the fecal slurries were analyzed using high-performance liquid chromatography (HPLC) (Water 1525, USA). Samples were prepared following the modified method of Wang et al. (2019). The supernatant was collected using centrifugation at 13,000×g at 4 °C for 5 min and mixed with 0.2% v/v tartaric acid as the internal standard (ratio 3:1). Before injection, the mixed solution was filtered through a 0.2 µm PVDF filter (Verical, Thailand). SCFAs were isolated using an Aminex HPX-87H column (300 × 7.8 mm; Bio-Rad, Hercules, CA, SA) at 50  °C, with 8 mM sulfuric acid used as the mobile phase at a flow rate of 0.60 ml/min for 60 min and wavelength 210 nm using a UV detector (Waters 2489). The injector was set with a split ratio of 20:1. Agilent Technologies 7890A equipment (Santa Clara, USA) was used for data collection and calculation of all parameters. Standard solutions were prepared from lactic acid, acetic acid, propionic acid and butyric acid at concentrations of 1, 2, 2.5, 5, 10 and 20 mmol/ml.

Stool secretory IgA

Analysis of secretory immunoglobulin A (IgA) in stool samples was performed in 55 out of the 56 subjects who completed the study (19 in 3CMH, 17 in 5CMH, 19 in placebo group). Samples were collected at the baseline and after the treatment period and stored at −80 °C. All fecal samples were thawed at room temperature and weighed to 0.1 g. Phosphate-buffered saline was added to the samples, then they were mixed in a stomacher machine and centrifuged at 14,000 rpm at 4 °C for 15 min. A sample of 200 µl (10×) of supernatant was collected and diluted to 200× for the single plex human isotyping assay. IgA was measured using a Bio-Plex Pro™ Human Isotyping kit (cat no. 171A3101M; Bio-Rad). The Bio-Plex Pro™ assays were essentially immunoassays formatted on magnetic beads by captured antibodies directed against the desired biomarker coupled to the beads. The coupled beads reacted with samples containing the IgA biomarker of interest, and biotinylated detection antibody was added to create a sandwich complex. The final detection complex was formed with addition of streptavidin-phycoerythrin conjugated using the Bio-Plex^® 200 system reader (Bio-Rad). The concentration of analyte bound to each bead was proportional to the median fluorescence intensity of the reporter signal.

Statistical analysis

Statistical analyses were performed using the statistical software package SPSS version 17. Statistical significance among the three groups was determined using one way ANOVA and the Kruskal-Wallis test for parametric and non-parametric tests, respectively. Comparisons of the two groups were determined using the independent t-test and Mann–Whitney U-test for parametric and non-parametric tests, respectively.

Linear mixed models of bacterial relative abundance were performed for each participant using R package (2.15.0) including age, gender, carbohydrate intake, fiber intake, dose of CMH supplement and sampling period.

Demographic characteristics of study subjects

Of the 227 subjects screened, 60 were enrolled in the project at visit 1 and 56 completed the study (Fig. 1). The drop out occurred randomly due to the use of antibiotics and anti-viral treatment of subjects, which happened to occur in CMH group. Another drop-out subject was due to the non-compliance of stool collection. Therefore, the higher drop out in CMH groups was unlikely related to the testing products. The participants were randomly assigned to the three subgroups, each of which had similar demographic characteristics (Table 1). Participants in the 3CMH group were older than those in the placebo group. During the testing period, energy, and nutrient consumption values of the placebo, 3CMH and 5CMH groups were calculated (Table 2). Results showed that all subjects consumed similar carbohydrate, protein, fat, fiber, and sugar contents. Therefore, the energy consumption was also similar in all groups.

Influence of CMH on defecating conditions and gastrointestinal symptoms

Defecating conditions and gastrointestinal symptoms in the baseline period of all groups were similar (Table S4). Interestingly, stool size of participants in the 3CMH group was significantly smaller than that in the placebo group during the testing period (p < 0.05). The effect was still apparent even after the products were discontinued (washout period). Stool size from the 5CMH group was also smaller than that from the placebo group but did not reach statistical significance. Participants in the 3CMH and 5CMH groups showed improvement in defecating conditions including less strong smell, softened stools and reduced gastrointestinal discomfort (such as bloating, flatulence and abdominal discomfort) compared to the placebo group during the testing and washout period but without statistical significance (Table S4).

Median changes of defecating and gastrointestinal symptoms between the baseline and testing period or washout period were analyzed. Two factors including size of stool and flatulence/bloating symptoms showed significant difference between sample groups (Fig. 3). Comparing with the baseline, size of stool and flatulence/bloating symptoms of the 3CMH group reduced and reduction level was significantly lower than the placebo group (p = 0.059). Similarly, reduction of stool size and flatulence/bloating symptoms in the 3CMH group was observed in the washout period when compared to the baseline. Median level change was also significantly lower than the placebo group (p = 0.045). Unlike the 3CMH group, median changes of stool size and flatulence/bloating symptoms from baseline to testing period in the 5CMH group were similar to the placebo group. Flatulence and bloating symptoms were significantly greater in the 5CMH group at washout period compared with the baseline and significantly lower than the placebo group (p ≤ 0.05).

Influence of CMH on microbial richness and diversity

Microbial richness indices of stool samples in each group were monitored during the baseline, testing and washout periods (Fig. S1). The Chao1 index of the 3CMH group during the testing and washout periods increased compared to the baseline, while the Chao indices for the placebo and 5CMH groups decreased and were stable, respectively.

Similarly, the Shannon diversity index for the 3CMH group showed an increasing trend during the testing and washout periods compared to the baseline (Fig. S1), while Shannon diversity indices of the 5CMH and placebo groups during the baseline, testing and washout periods were stable. These results suggested that the administration of CMH at 3 g/d promoted microbial diversity and richness.

The microbial communities from all groups and time points were intermingled in non-metric multidimensional scaling plots using Bray–Curtis dissimilarity (Fig. S1). Similar to the result of alpha diversity, beta-diversity showed no differences or clusters due to CMH intake or sampling period.

Influence of CMH on fecal microbiota

To assess specific changes in gut microbiota caused by CMH, gut microbiota from different taxa were identified from the three sampling periods (baseline, testing, and washout) (Tables S5–S7). Six microbiota phyla (Actinobacteria, Bacteroides, Firmicutes, Lentisphaerae, Proteobacteria and Verrucomicrobia) were found in all groups (Table S8). The most dominant phyla in each period were Firmicutes, Bacteroides and Actinobacteria. Dynamic changing patterns during the baseline, testing and washout periods of all phyla from the three groups were similar. No significant changes were recorded in the microbial phyla in the CMH treatment. It was observed that Firmicutes and Bacteroidetes ratios at the testing period for both the 3CMH and 5CMH groups were lower than the placebo, although they were not significantly different (Table S8).

At the family level, changes in gut microbiota during testing period were similar in all groups (Fig. 4A). Supplementation with 3CMH promoted various microbes including Desulfovibrionaceae, Bifidobacteriaceae, Barnesiellaceae, Veillonellaceae, Lachnospiraceae, Clostridiaceae, Rikenellaceae, Prevotellaceae, Porphyromonadaceae, Paraprevotellaceae, Odoribacteraceae, and Bacteroidaceae, while eight families of gut microbiota in the placebo and 5CMH groups were stable (mean change = 0) and six families decreased. Ruminococcaceae was reduced in the 3CMH group, whereas it increased in the placebo and 5CMH groups. Median change in Ruminococcaceae in the placebo group was significantly greater than in the 3CMH group (p ≤ 0.05).

After the washout period, median changes in gut microbiota were similar in all groups (Fig. 4B). However, induction of Bifidobacteriaceae was observed in the 3CMH group even after the products were discontinued, while showing reduction in the placebo and 5CMH groups. The amount of Enterobacteriaceae reduced in the 3CMH and placebo groups, with significantly different median change compared with the 5CMH group (p ≤ 0.05), that showed an increase during the washout period.

To confirm the relationship of changes of three families including Ruminococcaceae, Bifidobacteriaceae and Enterobacteriaceae with CMH intake, the linear mixed effect model was performed (Table S9). Results revealed that gender, age, sampling period, carbohydrate and fiber consumption had no significant effect on changes in the three families. Supplementation with 3CMH showed a significant effect on Ruminococcaceae observed during testing period, while with 5CMH significantly affected Bifidobacteriaceae and Enterobacteriaceae during the washout period (p ≤ 0.05)

At the testing period, predictions of functional analysis of gut microbiota were focused on carbohydrate-related metabolism (Fig. 5). Among the fifteen pathways, median change in relative abundance of the two pathways, ascorbate/aldarate and fructose/mannose metabolism showed significant differences between the three groups. The gene function associated with ascorbate/aldarate and fructose/mannose metabolism in the 3CMH group significantly increased more than in the placebo group (p ≤ 0.05). However, median changes in the 5CMH group were similar to those in the placebo group, indicating that CMH intake enhanced the microbial pathways of ascorbate/aldarate and fructose/mannose metabolism.

Influence of CMH on fecal short-chain fatty acids

The major fecal metabolite in all samples were acetic acid, followed by butyric acid, propionic acid, and lactic acid (Table 3). The SCFAs in all subjects at the baseline were similar. Intake of CMH had no effect on the four SCFAs and they were similar during the testing period; however, significant differences among the three groups were observed during the washout period. Highest abundance of lactic acid was recorded in the 5CMH group followed by the 3CMH and placebo groups (p < 0.05). The amount of lactic acid in the 5CMH group was significantly higher than in the placebo group but similar to the 3CMH group. The level of butyric acid in the placebo group was significantly greater than in the 3CMH group.

Influence of secretory immunoglobulin A in stools

Results revealed no significant differences in secretory IgA levels in stools among all groups both at the baseline and after the testing period (Table 4). Forty-two percent of the participants who received 3CMH treatment showed increased stool secretory IgA compared to 20% of the participants in the 5CMH and placebo groups. The increase in IgA correlated with increased stool frequency (r = 0.303 to 0.039) and reduced abdominal discomfort symptoms (r = −0.238) during the treatment period.