Influence of resistant starch on the SCFA production and cell counts of butyrate-producing Eubacterium spp. in the human intestine


Andreas Schwiertz Deutsches Institut für Ernährungsforschung, Abteilung Gastrointestinale Mikrobiologie, Arthur-Scheunert-Allee 114-116, D-14558 Bergholz-Rehbrücke, Germany (e-mail:


Aims:  The genus Eubacterium, which is the second most common genus in the human intestine, includes several known butyrate producers. We hypothesized that Eubacterium species play a role in the intestinal butyrate production and are inducible by resistant starch.

Methods and Results:  In a human pilot study species-specific and group-specific 16S rRNA-targeted, Cy3 (indocarbocyanine)-labelled oligonucleotide probes were used to quantify butyrogenic species of the genera Eubacterium, Clostridium and Ruminococcus. Following the intake of RS type III a significant increase in faecal butyrate but not in total SCFA was observed. However, increase in butyrate was not accompanied by a proliferation in the targeted bacteria.

Conclusions:  The tested Eubacterium species have the capacity to produce butyrate but do not appear to play a major role for butyric acid production in the human intestine.

Significance and Impact of the Study: In view of the fact that the bacteria responsible for butyrate production are largely unknown, it is still difficult to devise a dietary intervention to stimulate butyrogenic bacteria in a targeted way.


Dietary fibre, which is not hydrolysed by host enzymes in the small intestine, undergoes bacterial degradation in the large intestine. The main fermentation products are the short-chain fatty acids (SCFA) acetate, propionate and butyrate. The latter has been shown to be the preferred energy substrate of the colonocytes (Roediger 1980). Butyrate has also been implicated in providing protection against cancer (Hill 1995). Since Englyst et al. (1987) reported that starch exhibits in vitro the strongest butyrogenic effect of all carbohydrates tested, several studies have demonstrated that resistant starches stimulate bacterial butyrate formation (Jenkins et al. 1998; Le Blay et al. 1999). Despite this, however, little is known about the predominant butyrate-producing bacteria in the human intestinal tract. The most obvious reason lies in the fact that there is no simple way to enrich and isolate butyrate producers. Each random isolate has to be tested for its ability to produce butyrate. So far, only a few reports have contributed to the description of the human butyrate-producing flora (Barcenilla et al. 2000; Sharp and Macfarlane 2000).

The choice of bacteria in this study was based on reports by Moore and Holdeman (1974), Finegold et al. (1983), Simmering et al. (1999) and Schwiertz et al. (2000), and included known butyrate producers (Moore and Holdeman Moore 1986): Eubacterium barkeri, Eu. biforme, Eu. cylindroides, Eu. dolichum, Eu. hadrum, Eu. limosum, Eu. moniliforme, Eu. multiforme, Eu. rectale, Eu. ramulus, Eu. saburreum, Eu. tortuosum and Eu. ventriosum. The species were chosen on the basis of their numerical importance and availability of probes for whole cell in-situ hybridization experiments. The objective of our study was to test the hypothesis that resistant starch (RS) type III stimulates the growth of the butyrogenic species of the genera Eubacterium, Clostridium and Ruminococcus and thereby induces higher faecal SCFA concentrations.


Oligonucleotide probes

16S rRNA-targeted probes were chosen to detect Eubacterium species known to produce butyrate, or bacterial groups known to include butyrogenic species (Moore and Holdeman Moore 1986; Franks et al. 1998). These probes included species-specific probes for the detection of Eubacterium species (Simmering et al. 1999; Schwiertz et al. 2000) and bacterial group probes for the detection of Ruminococcus, Eubacterium and Clostridium species (Erec482, Franks et al. 1998). In addition, species-specific probes for the detection of Eu. multiforme, Eu. rectale, Eu. saburreum and Eu. tortuosum were designed and validated for whole cell in-situ hybridization experiments as described earlier (Schwiertz et al. 2000). All used probes are summarized in Table 1.

Table 1.   Probes used in the study and their corresponding sequences Thumbnail image of


The resistant starch (RS) product was made from native pea starch (Cosucra, Fontenoy, Belgium). The starch was enzymatically debranched and a 15% gel (w/w) was retrograded at room temperature for 24 h. The RS type III-content of the product was 50%, determined according to Englyst et al. (1992). The daily intake was 15 g RS type III product.

Subjects and pilot study design

Five healthy subjects (four females and one male) participated in a 28-d pilot study. The age ranged from 25 to 64 years. None of the subjects took antibiotics during the preceding three months and during the pilot study. The volunteers maintained their usual lifestyles and dietary intakes throughout the study period. During the study period of 28 d each subject consumed in addition to its unrestricted diet 15 g of RS type III product. The Ethical Committee of Brandenburg approved the protocol (no. eK 170–5/kmo, 8.10.1998) and all persons were asked to give their informed consent before the beginning of the study.


Fresh faecal samples were collected from each subject at the beginning of the study, on the last day of RS-intake, and finally on the 14th day following the end of RS-intake.

Hybridization and analysis of faecal flora

For the analysis by whole cell in-situ hybridization a 0·2 g (wet weight) faeces specimen was added to 1·8 ml of sterile resuspension buffer containing 50% (v/v) ethanol-phosphate-buffered saline (139 mol l−1 NaCl, 10 mol l−1 NaH2PO4-Na2HPO4 at pH 7·2). The suspension was mixed by inverting and vortexing the tube for 5–10 min. The fixed samples were stored at –20 °C until used. Hybridization and enumeration were performed as described earlier (Schwiertz et al. 2000).

Data analysis and statistics

The data for each time point of the human pilot study are given as both individual data and the means and S.D. of five subjects. Differences between the means were checked for significance by the paired t-test, as described by Lorenz (1992), and are indicated as P.


Probe design and specificity testing

Several 16S rRNA-targeted oligonucleotide probes for the detection of Eu. multiforme, Eu. rectale, Eu. saburreum and Eu. tortuosum were designed based on comparative analysis using the ARB software program (Strunk and Ludwig 1996) and checked with the RDP- and the EMBL-databases. The probes designed for the detection of Eu. multiforme (S-S-Emul-0183-a-A-18) and Eu. saburreum (S-S-Esab-1467-a-A-18) hybridized to the corresponding target organism but not to the approximately 100 intestinal organisms used in the specificity testing (Schwiertz et al. 2000). The optimal hybridization temperatures of the designed probes for the hybridization experiments were 48 °C for Eu. saburreum and 51 °C for Eu. multiforme, respectively. None of the probes designed for the detection of Eu. rectale and Eu. tortuosum, respectively, worked in whole cell in-situ hybridization experiments.

Enumeration of total and specific bacterial populations

To determine the impact of RS on the bacterial flora, five subjects were tested for their faecal bacterial composition in response to the consumption of RS. Cell counts and incidences of target organisms and groups in the five subjects are summarized in Table 2. The lower detection limit was 107 cells g−1 dry weight faeces. Subjects harboured at least one butyrate-producing Eubacterium species in each of their collected faecal samples. Eubacterium hadrum was detected in all five subjects. Eubacterium ramulus was found in four subjects while Eu. biforme was detected in only three subjects. No other Eubacterium species (Eu. barkeri, Eu. cylindroides, Eu. dolichum, Eu. saburreum, Eu. limosum, Eu. moniliforme, Eu. multiforme and Eu. ventriosum) could be detected throughout the study in the faeces of any of the subjects. Organisms detected with the REC-cluster probe were found in all subjects at concentrations of 1·0 × 1010 to 3·7 × 1011 cells g−1 dry weight of faeces. RS-intake did not lead to a significant change in any of the bacterial groups and species detected. Fourteen days after the intervention period the cell numbers of Eu. hadrum and Eu. ramulus increased significantly in comparison to the beginning of the study. Changes in the cell counts of Eu. biforme, the REC-cluster and total bacteria were not significant at this time.

Table 2.   Bacterial counts determined with whole cell in-situ hybridization for the different subjects during the study period Thumbnail image of

SCFA analysis in faecal samples

The major SCFA found in the faecal samples were acetate, propionate and butyrate (Table 3); the concentrations of iso-butyrate and iso-valerate, were very low (< 16 μmol g−1 dry weight) and are therefore not shown. The differences in the faecal SCFA concentrations between the five subjects were considerable. During the period of RS intake the faecal concentration of total SCFA increased in four of the five subjects. With the exception of one subject, the increase in SCFA was accompanied by a rise in the butyrate concentration. Following the intervention period, a significant increase in the faecal butyrate concentration was observed, while the relative proportions of the single SCFA remained unchanged. At the 14th day after the last RS-intake, the mean SCFA concentrations were lower than during the period of RS consumption (Table 3) and not different from those measured at the beginning of the study.

Table 3.   SCFA concentrations and relative proportions of the SCFA for the human pilot study Thumbnail image of


Short-chain fatty acids (SCFA) can be derived in large quantities from bacterial fermentation of dietary fibre in the large bowel (McIntyre et al. 1993). Although not proven, a high fibre intake has been associated with a reduced risk of colon cancer (McIntyre et al. 1993; Hill 1995). One of the SCFA, butyrate, is known to play a key role in the energy metabolism of the colonic epithelial cells and is thought to be important in the maintenance of colonic health in humans (Mortensen and Clausen 1996). In particular, its ability to modify nuclear architecture and induce death by apoptosis in colon cancer cells is of great interest (Heerdt et al. 1997). It is known that bacterial butyrate production is stimulated particularly by resistant starches (Wang et al. 1999; Sharp and Macfarlane 2000).

Since butyrate-producing strains of the genus Eubacterium are found in rather high cell counts (Finegold et al. 1983; Simmering et al. 1999; Schwiertz et al. 2000) we hypothesized that these species might play a role in the reported increase of butyrate following the intake of resistant starch. To test this assumption we performed a human pilot study, in which we investigated the influence of resistant starch on butyrate-producing Eubacterium species and on bacterial groups known to include butyrogenic species detectable with the Erec482-probe (Finegold et al. 1983; Franks et al. 1998). The enumeration of the species was done by whole cell in-situ hybridization. The cell counts obtained in our study are essentially in agreement with the published data, which were obtained with either classical methods (Finegold et al. 1983) or whole cell in-situ hybridization (Franks et al. 1998; Simmering et al. 1999; Schwiertz et al. 2000). No correlation between cell counts of the targeted organisms, RS intake and the increase in the butyrate concentration could be drawn. It may be speculated that other bacterial population groups were responsible for the observed butyrate increase after RS intake. A significant increase in cell numbers was observed only for Eu. hadrum and Eu. ramulus at day 14 after the last RS intake. This phenomenon may be due to dietary compounds not investigated in this study, as the increase was only significant 14 d after the last RS intake. RS intake cannot be held responsible for the detected increase in Eu. hadrum and Eu. ramulus cell counts.

The fact that not all Eubacterium species could be detected during this study can be explained by the detection limit of whole cell in-situ hybridization, which is approximately 107 cells g−1 faeces dry weight (Schwiertz et al. 2000). However, it cannot be excluded that the cell counts for Eu. rectale and Eu. tortuosum increased during RS intake. No specific probe could be developed for either of the two organisms. This may be due to inaccessibility of the target site (Fuchs et al. 1998).

Sharp and Macfarlane (2000) showed that butyrogenic species require a high supplementation with RS. Furthermore, their study indicates that saccharolytic clostridia are best adapted to fast growth rates and high substrate concentrations and that resistant starch granules are advantageous for their growth. As several of the Eubacterium species are close relatives of the clostridia, it could be speculated that such conditions might also be favourable for the eubacteria, while our chosen intake of 15 g RS III d−1 per person might have been insufficient to stimulate the growth of the bacterial species investigated.

Earlier studies with resistant starch reported an increase in the SCFA and butyrate concentrations (Jenkins et al. 1998; Le Blay et al. 1999). The results of this human pilot study are consistent with these reports as we found a rise in total SCFA after RS intake in four of the five subjects and a significant increase in the butyrate concentration. However, there was no increase in the proportions of butyrate relative to the other SCFA.

So far, no key butyric acid producing bacteria have been identified, although numerous species belonging to the cluster XIVa of the clostridia (Collins et al. 1994) are able to utilize starch and produce butyrate (Barcenilla et al. 2000; Sharp and Macfarlane 2000). Since species-specific probes for these bacteria are not available, the use of the REC probe, which detects most of the species from cluster XIVa, is presently the best way to describe the impact of resistant starch on butyrogenic bacteria from this cluster. However, the intake of resistant starch did not result in an increase in cell numbers detectable with this probe.

In summary, we have shown that the intake of RS type III leads to an increase in the faecal concentrations of SCFA and butyrate, but does not stimulate a number of selected butyrogenic Eubacterium species. This indicates that these organisms do not play a major role in butyrate formation from starch.


This work has been carried out with financial support from the Commission of the European Communities, specific RTD programme `Quality of Life and Management of Living Resources', QLK1-2000-108, “Microbe Diagnostics”. It does not necessarily reflect its views and in no way anticipates the Commission's future policy in this area.