Diet shapes the ability of human intestinal microbiota to degrade phytate – in vitro studies


  • L.H. Markiewicz,

    Corresponding author
    • Department of Immunology and Food Microbiology, Institute of Animal Reproduction and Food Research of the Polish Academy of Sciences, Olsztyn, Poland
    Search for more papers by this author
  • J. Honke,

    1. Department of Chemistry and Biodynamics of Food, Institute of Animal Reproduction and Food Research of the Polish Academy of Sciences, Olsztyn, Poland
    Search for more papers by this author
  • M. Haros,

    1. Cereal Group, Institute of Agrochemistry and Food Technology, IATA-CSIC, Valencia, Spain
    Search for more papers by this author
  • D. Świątecka,

    1. Department of Immunology and Food Microbiology, Institute of Animal Reproduction and Food Research of the Polish Academy of Sciences, Olsztyn, Poland
    Search for more papers by this author
  • B. Wróblewska

    1. Department of Immunology and Food Microbiology, Institute of Animal Reproduction and Food Research of the Polish Academy of Sciences, Olsztyn, Poland
    Search for more papers by this author


Lidia Hanna Markiewicz, Department of Immunology and Food Microbiology, Institute of Animal Reproduction and Food Research, Tuwima Str. 10, 10-748 Olsztyn, Poland.




Investigation of intestinal bacterial groups involved in phytate degradation and the impact of diets with different phytate contents on phytase activity.

Methods and Results

Faecal samples of adults on conventional (n = 8) or vegetarian (n = 8) diets and breastfed infants (n = 6) were used as an inoculum for modified media supplemented with phytate. Populations of Gram-positive anaerobes (GPA), lactic acid bacteria (LAB), Proteobacteria–Bacteroides (P-B), coliforms and anaerobes were studied. The PCR-DGGE analysis revealed a random distribution of DGGE profiles in the dendrograms of GPA, P-B and coliforms, and a partially diet-specific distribution in the DGGE dendrograms of LAB and anaerobes. The degradation of phytic acid (PA) was determined with HPLC method in supernatants of the cultures. Regardless of the diet, the Gram-positive anaerobes and LAB displayed the lowest ability to degrade phytate, whereas the coliforms and P-B cultures produced higher amounts of intermediate myo-inositol phosphates. Bacterial populations grown in a nonselective medium were the most effective ones in phytate degradation. It was the vegetarians' microbiota that particularly degraded up to 100% phytate to myo-inositol phosphate products lower than InsP3.


A diet rich in phytate increases the potential of intestinal microbiota to degrade phytate. The co-operation of aerobic and anaerobic bacteria is essential for the complete phytate degradation.

Significance and Impact of the Study

This study provides insights on the effect of diet on specific metabolic activity of human intestinal microbiota.


A diet provides nutrients both for the human organism and the bacteria inhabiting the human gastrointestinal tract. All compounds of a diet are potentially biologically active and may have both beneficial and harmful effects on the organism and intestine health. It has been also proved that components of the diet (e.g. prebiotics, proteins and glycoproteins or fat) influence the composition and metabolic activity of intestinal microbiota (Bielecka et al. 2002; Brunser et al. 2006; Barbier de La Serre et al. 2010; Świątecka et al. 2010, 2011).

Phytic acid (PA) and its salts (phytates, myo-inositol hexakis phosphates or InsP6) are components of cereal grains and legume seeds. Its chemical properties make it both an antinutrient (decreases protein digestion and bioavailability of nutritionally important cations such as Ca2+, Mg2+, Fe2+and Zn2+; Wodzinski and Ullah 1996) and a beneficial (strong antioxidant) food compound. A degradation of PA leads to a release of inorganic phosphate and formation of lower inositol phosphates: penta-, tetra-, tri-, di- and mono-myo-inositol phosphates (InsP5, InsP4, InsP3, InsP2 and InsP1) of which only InsP5 displays the ability to decrease the availability of minerals (Sandberg et al. 1989). Other derivatives of PA have no antinutritional properties; moreover, InsP3 and InsP4 play an important role in the intracell signal transduction pathways (Suzuki and Hara 2010; Suzuki et al. 2010).

A partial degradation of PA may occur during food processing (Afify et al. 2011), but most of the food phytate remains nondegraded and reaches the gastrointestinal tract. The human intestinal tissue expresses a very low phytase activity, which is insufficient to degrade all the dietary phytate and cannot contribute to any physiologically relevant degradation of inositol phosphates in the small intestine (Iqbal et al. 1994). Ingested phytate is subjected to a chemical (low pH in the stomach) or enzymatic degradation carried out by the intrinsic food phytases (mainly 6-phytases of plant origin) or intestinal microbiota (3-phytases and 6-phytases) (Sandberg 2002), and the efficacy of phytate degradation in the intestine is higher due to longer colon transit times (Joung et al. 2007).

Studies on the ability of particular groups of intestinal microbiota to decompose phytates are scarce. In fact, phytases are produced by Klebsiella and Escherichia coli (Gammaproteobacteria) (Lim et al. 2000; Bohm et al. 2010). Haros et al. (2005, 2007) have proved that some Bifidobacterium strains of the human intestinal origin manifest the phytase activity and the phytases from Bifidobacterium pseudocatenulatum and Bifidobacterium longum subsp. infantis have been recently characterized (Tamayo-Ramos et al. 2012). However, studies of Steer et al. (2004) conducted in the chemostat model indicated that this bacterial group is not involved in the phytate degradation.

This study aimed at examining which bacterial groups participate in the hydrolysis of PA and whether diets containing different amount of phytates (conventional, vegetarian and breastfeeding) influence the specific metabolic activity (phytate-degrading ability) of the intestinal microbiota.

Materials and methods


Adult volunteers on the conventional (C) diet [six women and three men, average age 35·5 years, average BMI 22·9 (min. 19·5, max. 28)], adult vegetarians and vegans [V; three women and five men, average age 26 years, average BMI 21·1 (min. 18·6, max. 24·1)] and breastfed infants (BF; two girls and four boys, average age 17·5 weeks) were included in the study. The adult volunteers declared that products containing live bacteria (i.e. fermented milk or fermented vegetable products and bacterial preparations) had been excluded from their diets at least 1 week before the sampling. The conventional diet was a western-type diet without any restrictions other than the aforementioned. The diet of the vegetarians contained no meat, fish and seafood (four volunteers) and, additionally, no animal-originating food (i.e. milk and milk product and eggs; four volunteers).


Fresh faecal samples of the adults were transported in a sterile tube to the laboratory in anaerobic conditions (AnaeroGen System; Oxoid, Poznan, Poland) within one hour and immediately subjected to the analyses. Faecal samples of the infants were transferred to a sterile tube and frozen at −20°C, transported to the laboratory within 2 days and subjected to the analyses. The experimental design is presented in Fig. 1.

Figure 1.

Schema of the study design. Modified media used in the study: Garche's (G), de Man Rogosa medium (MRS), neomycin brilliant green taurocholic acid medium with blood (NBGT), MacConkey (MC) and Wilkins–Chalgren medium with horse's blood (W-Ch). For the detailed description see: 'Materials and methods'.

Microbiological analyses

The faecal samples were transferred to the Anaerobic Workstation (MG500; DonWhitley Scientific, Shipley, West Yorkshire, UK; atmosphere composition: 80% N2, 10% CO2, 10% H2). All the media and reagents were incubated under anaerobic conditions for at least 24 h before their use. To investigate selected groups of intestinal bacteria, the following modified liquid media were used: Garche's medium for Bifidobacterium (Haros et al. 2005), de Man Rogosa medium (MRS) for the lactic acid bacteria (LAB), the MacConkey medium for Enterobacteriaceae, the neomycin brilliant green taurocholic acid medium with blood (NBGT) for Bacteroides (Yasui et al. 1979) and the nonselective Wilkins–Chalgren medium with blood for the anaerobes. All the media were modified by a supplementation of 1 mmol l−1 sodium phytate (Sigma, Poznan, Poland), an elimination of inorganic phosphorus and lowering the concentration of protein components and a yeast extract in order to generate low-phosphate conditions, which promote a synthesis of the enzymes responsible for the phytate degradation (Haros et al. 2005). For the detailed composition of the microbiological media, see the Supporting Information. The modified media without the inoculum were incubated and processed in the same way as the cultures of the faecal bacteria and were taken as controls for the HPLC analysis.

The faecal samples (1 g) were diluted in 1% peptone water (49 ml) and used for the media inoculation. Fifteen millilitres of medium were inoculated with approximately 107 cells, estimated with DAPI, and incubated at 37°C either in the anaerobic conditions (Garche's medium, NBGT and anaerobes; 40 h) or in aerobic ones (MacConkey, 24 h; MRS, 40 h). After the incubation, the bacterial cell number was assessed on appropriate nonmodified media using the plate-count method. Results of the cell counts were expressed as log10 CFU ml−1.

DNA extraction, PCR-DGGE analysis and band sequencing

The DNA from 1·5 ml of the faecal bacteria cultures was isolated using the GeneMATRIX Bacterial and Yeast Genomic DNA Purification kit (Eurx, Gdansk, Poland) applying the protocol delivered by the producer. The amplification of the V6-V8 fragment of 16S rRNA gene was carried out using the extracted DNA as a template and universal primers 1404-r (cggtgtgtacaagaccc) and 968-GC-f (GC-aacgcgaagaacctta) (Satokari et al. 2002). The reaction mixtures (30 μl) comprised of 3 μl of a reaction buffer, 1 U of Taq DNA polymerase (Fermentas, Gdansk, Poland), 5 mmol l−1 MgCl2, 200 μmol l−1 each: dATP, dCTP, dGTP and dTTP, the template DNA (2 μl), sterile deionized water for filling up to 30 μl. The reactions were run in a MasterGradient Cycler (Eppendorf, Warsaw, Poland) using the following program: 95°C for 5 min, 35 cycles at 95°C for 20 s, 58°C for 20 s, 72°C for 20 s and the final 72°C for 20 min. The PCR product (20 μl) was separated in a polyacrylamide gel (8%, acrylamide/bisacrylamide 37·5 : 1) with denaturing gradient ranging from 25 to 60%. The 100% concentration of denaturants corresponded to 40% (v/v) of formamide and 7 mol l−1 urea (Satokari et al. 2002). The electrophoresis was carried out in the 0·5× TAE buffer at 60°C, and the steps of pre-electrophoresis (10 min, 200 V) and separating electrophoresis (18 h, 85 V) were applied. The gels were stained with SybrGreen I dye (Sigma) in the 1× TAE buffer according to the supplier's recommendations.

Selected DGGE bands were cut out from a gel with a sterile scalpel, incubated overnight in 40 μl of the TE buffer at 4°C. Two microlitres of the aliquots were used as a template in the re-amplification reactions conducted in the conditions described previously. The re-amplified DNA was purified using the GeneMATRIX PCR/DNA Clean-Up Purification Kit (Eurx). The DNA was commercially sequenced at Genomed (Warsaw, Poland). The obtained sequences were identified using a BLASTn tool and deposited in the GenBank database (accession numbers: KC108628-KC108644, KC137555-KC137570, KC147729-KC147781).

Determination of myo-inositol phosphates by HPLC

Five millilitres of the faecal bacterial cultures were centrifuged (15 min, 10 000 g, 10°C), and the supernatant was mixed with 10 ml of 0·025 mmol l−1 HCl. The samples were transferred to plastic minicolumns filled with the Dowex AG 1-X8 resin, from which inositol phosphates were eluted using 2 mmol l−1 HCl (5 × 2 ml). After desolventizing by evaporation, the dry residue was dissolved in a mixture of methanol/0·05 mmol l−1 formic acid and 15 ml l−1 TBA-OH (tertrabutylammonium hydroxide) and analysed by HPLC according to the methods by Sandberg and Ahderinne (1986) and Sandberg et al. (1989) using a Shimadzu chromatograph (LC-10 AD pump, refractometric detector RID-6A, CTO 6A column oven) and the Nova-Pak C18 column. The mobile phase was composed of the following: methanol/0·05 mmol l−1, formic acid (51 : 49 v/v) and 15 ml l−1 TBA-OH. The flow rate was 0·7 ml min−1. Sodium phytate (Sigma) was the external standard, and the injections were made with a 20 μl loop. The concentrations of InsP3, InsP4, InsP5 (intermediate hydrolysis product; INT-InsPs) and InsP6 determined in the supernatant were expressed as percentage of the initial amount of phytic acid in the media. The amount of the lower derivatives (InsP1, InsP2 and myo-inositol; LD) was calculated as a difference between the initial content of phytate and the sum of InsP6 and INT-InsPs.

Data analysis

The DGGE banding profiles were processed using the BioNumerics software (Applied Maths, Sint-Martens-Latem, Belgium). The gels were normalized to one sample (as an external standard) run for the each gel set. Similarities of the profiles were calculated using the Pearson's correlation moment coefficient, and the dendrograms were constructed using the UPGMA method. The average similarity of a composite data set and a dendrogram made of similarity of profiles of the bacteria grown in Garche's, MRS, MacConkey and NBGT media was also calculated.

The differences in the bacterial cell numbers of in vitro cultures and the concentrations of myo-inositol phosphates were analysed with the nonparametric Kruskal–Wallis test (for nonuniformly distributed variables) or the variance analysis (F-test; for variables with normal distribution and equal variances; Statistica, Crakow, Poland). The means (medians) of inositol phosphate percentage distribution for all the samples were calculated to compare the potential of the investigated bacterial groups to degrade phytate. The relative proportions of LD, INT-InsPs and nondegraded phytic acid (NDPA) were calculated on the basis of inositol phosphate percentage distribution. The (F) factor was introduced to characterize numerically the ability of the bacterial cultures to degrade phytate. The general aptitude to degrade phytate (to release at least one phosphate group from InsP6) was defined as Ftotal [(initial amount of phytate − NDPA) × (LD + INT-InsPs)]/100. The ability of the bacterial cultures to generate intermediate inositol phosphates (FINT-InsPs) was defined as ((initial amount of phytate − NDPA) × INT-InsPs)/100. The ability of the bacterial cultures to hydrolyse phytate to the lowest inositol phosphates (FLD) was defined as ((initial amount of phytate − NDPA) × LD)/100. The values of LD, INT-InsPs and NDPA were medians of inositol phosphate percentage distribution, whereas the initial amount of phytate was 100%. The maximum value for each F factor (corresponding to the highest possible ability to degrade phytate or generate particular group of inositol phosphates) was 100.


Counts of faecal bacteria cultures

The mean for the faecal bacteria culture counts reached the levels ranging from 6·9 in the NBGT medium for infants' microbiota to 9·1 log CFU ml−1 in the cultures of the anaerobes from the vegetarians (Fig. 2). Due to a high intragroup variability, in most cases, no differences in the in vitro growth of investigated groups of microbiota were observed when compared to the both tested media and the volunteer groups counts (Fig. 2). A significantly lower bacterial number of infant's microbiota grown solely in NBGT medium when compared to both adult groups (P < 0·05) was observed.

Figure 2.

Counts (log10 CFU ml−1) of faecal bacterial cultures in modified media: Garche's (G), de Man Rogosa medium (MRS), MacConkey (MC), the neomycin brilliant green taurocholic acid medium with blood (NBGT) and Wilkins–Chalgren medium with horse's blood (W-Ch). The media were inoculated with diluted faecal bacteria of the adults on conventional or vegetarian diet, as well as the breastfed infants. Medium: (□) Garche's; (image_n/jam12204-gra-0001.png) MRS; (image_n/jam12204-gra-0002.png) NBGT; (image_n/jam12204-gra-0003.png) MacConkey; (image_n/jam12204-gra-0004.png) Wilkins-Chalgren.

Composition of faecal bacteria cultures

In order to determine the composition of bacterial cultures in the tested media, a PCR-DGGE assay (with primers universal for the bacteria) combined with band sequencing was applied. This also made it possible to confirm the selectivity of the particular medium (targeted bacteria) and to identify the growth of the other bacteria.

The analysis of the DGGE profiles for the faecal bacterial cultures grown in modified Garche's, MacConkey and NBGT media provided dendrograms in which the profiles were distributed randomly and characterized by their relatively high similarities (Supporting Information). The DGGE profiles of the cultures grown in the modified MRS medium (LAB) and in the modified Wilkins–Chalgren (anaerobes) medium were partially clustered in a diet-dependent manner. Particularly, the DGGE profiles of the MRS cultures were grouped into three clusters in which the profiles of vegetarians (cluster V), omnivores (cluster O) and infants (cluster I) predominated (Supporting Information). Also, the advanced cluster analysis of a composite including comparison of the DGGE profiles for all the examined bacterial groups did not demonstrate a relationship between the diet and the similarity of the DGGE patterns in the cultures. The analysis resulted in a dendrogram consisting of two clusters of relatively high similarity (60%), cluster I–V (comprising microbiota the profiles for the infants and vegetarians) and cluster V-C (comprising the profiles for vegetarians and all the profiles for the conventional diet group microbiota) (Fig. 3).

Figure 3.

Dendrogram of the composite data set consisting of DGGE patterns of faecal bacteria cultures in the modified media: neomycin brilliant green taurocholic acid medium with blood (NBGT), MacConkey, MRS and Garche's. The media were inoculated with the faecal bacteria (107 cells) retrieved from the adults on the conventional diet (C1–C8), the adult vegetarians (V1–V8) and the breastfed infants (I1-I6).

Sequencing of the DGGE bands showed that apart from the targeted bacteria, the growth of which improved in the particular medium, other bacterial groups were also detected (Table 1). The DGGE profiles of the cultures grown in Garche's medium consisted of bands representing mostly the Gram-positive anaerobic bacteria of Clostridia and Actinobacteria and some representing microaerophilic Gram-positive bacteria (Lactobacillales) (Table 1). Therefore, although Garche's medium is usually dedicated for the Bifidobacterium cultivation, in the conditions applied in this study (liquid cultures of complex microbiota), that medium was also appropriate for the Gram-positive anaerobes and was described as the GPA. In the cultures grown in the modified MRS medium, apart from the lactic acid bacteria (genera: Lactobacillus, Lactococcus, Streptococcus) that constituted the majority of the identified bands, there were also two bands identified as E. coli detected. In the modified MacConkey medium, only Gammaproteobacteria (genera: Enterobacter, Escherichia, Klebsiella, Citrobacter) were detected. In the NBGT medium, dedicated for Bacteroides, one sequence was classified as Bacteroides vulgatus species, whereas the rest of the bands represented Proteobacteria (genera: Escherichia, Shigella, Citrobacter, Desulfovibrio). Thus, the cultures grown in the NBGT medium were described as Proteobacteria–Bacteroides. The highest diversity was observed in the nonselective Wilkins–Chalgren medium, where representatives typical for the intestinal microbiota were detected. Those bacteria belonged to the families of Clostridiaceae, Lachnospiraceae, Eubacteriaceae, Enterobacteriaceae (Gammaproteobacteria), Sutterellaceae (Betaproteobacteria), Veillonellaceae and Bacteroidaceae (Table 1).

Table 1. Identification of DGGE bands taken from the DGGE profiles of bacterial cultures cultivated in the nonselective modified Wilkins–Chalgren medium (anaerobes, A), modified MRS medium (MRS), modified MacConkey medium (MC), modified NBGT (N) and modified Garche's medium (G)
Band numberaClosest relativeSimilarity %Class, Family
  1. a

    To see the band position in DGGE gels see Supporting Information.

A2, A6, A9, A13, A14 Escherichia coli 98–99Proteobacteria, Enterobacteriaceae
A3Enterobacter sp.99Proteobacteria, Enterobacteriaceae
A5 Dorea longicatena 97Clostridia, Lachnospiraceae
A7 Sutterella wadsworthensis 99Betaproteobacteria/Burkholderiales
A8 Clostridium lactatifermentans 96Clostridia, Clostridiaceae
A10-A11 Dialister invisus 98Negativicutes, Veillonellaceae
A12 Eubacterium eligens 96Clostridia, Eubacteriaceae
A15 Allisonella histaminiformans 99Negativicutes, Veillonellaceae
A16 Sutterella stercoricanis 98Proteobacteria, Sutterellaceae
A17 Citrobacter freundii 100Proteobacteria, Enterobacteriaceae
A18,A20 Clostridium perfringens 98–100Clostridia, Clostridiaceae
A19 Bacteroides vulgatus 99Bacteroidia, Bacteroidaceae
MC1-MC4, MC13, MC18 Escherichia coli 99–100Proteobacteria, Enterobacteriaceae
MC5, MC10, MC11 Enterobacter cloacae 99Proteobacteria, Enterobacteriaceae
MC7-MC9Enterobacter sp.100Proteobacteria, Enterobacteriaceae
MC10, MC12 Enterobacter cloacae 100Proteobacteria, Enterobacteriaceae
MC11, MC14-MC16 Citrobacter freundii/Klebsiella oxytoca 100Proteobacteria, Enterobacteriaceae
MC6, MC15, MC17, MC20, MC21 Escherichia coli/Enterobacter cloacae 99–100Proteobacteria, Enterobacteriaceae
MC19 Klebsiella pneumoniae 99Proteobacteria, Enterobacteriaceae
MRS1, MRS7, MRS16 Lactobacillus salivarius 99–100Lactobacillales, Lactobacillaceae
MRS2, MRS10, MRS11, MRS15 Streptococcus infantarius 99–100Lactobacillales, Streptococcaceae
MRS3, MRS14 Escherichia coli 99–100Proteobacteria, Enterobacteriaceae
MRS5 Lactococcus garvieae 100Lactobacillales, Streptococcaceae
MRS6 Streptococcus salivarius 100Lactobacillales, Streptococcaceae
MRS8 Lactobacillus mucosae 100Lactobacillales, Lactobacillaceae
MRS9, MRS17 Lactobacillus gasseri 99Lactobacillales, Lactobacillaceae
MRS12, MRS18 Lactobacillus fermentum 100Lactobacillales, Lactobacillaceae
MRS13Lactococcus lactis subsp. lactis99Lactobacillales, Streptococcaceae
N1, N2, N5-N7, N10, N12 Escherichia coli 98–100Proteobacteria, Enterobacteriaceae
N4 Citrobacter freundii 100Proteobacteria, Enterobacteriaceae
N8 Bacteroides vulgatus 97Bacteroidia, Bacteroidaceae

Escherichia coli

Enterobacter cloacae subsp. cloacae

100Proteobacteria, Enterobacteriaceae
N11 Desulfovibrio piger 98Deltaproteobacteria

Shigella flexneri

Escherichia coli

100Proteobacteria, Enterobacteriaceae
N16 Klebsiella oxytoca 100Proteobacteria, Enterobacteriaceae
G1, G9 Mitsuokella jaluladini 98–99Clostridia, Veillonellaceae
G2Clostridium sp.99Clostridia, Clostridiaceae
G3 Ruminococcus torques 100Clostridia, Lachnospiraceae
G4, G16 Eubacterium hadrum 99Clostridia, Eubacteriaceae
G5 Bifidobacterium pseudocatenulatum 99Actinobacteria, Bifidobacteriaceae
G6 Clostridium ramosum 100Erysipelotrichi, Erysipelotrichaceae
G7 Streptococcus infantarius 100Lactobacillales, Streptococcaceae
G8,G11 Dorea longicatena 99–100Clostridia, Lachnospiraceae
G10 Ruminococcus lactaris 98Clostridia, Ruminococcaceae
G12, G13 Bifidobacterium adolescentis 99–100Actinobacteria, Bifidobacteriaceae
G14, G15 Clostridium perfringens 100Clostridia, Clostridiaceae

Phytate degradation ability in selected groups of intestinal microbiota in vitro

The HPLC method was used to determine the level of InsP6, its intermediate derivatives (InsP3, InsP4, InsP5; INT-InsPs) and its lower derivatives (InsP1, InsP2 and myo-inositol; LD). The ability to degrade phytate by particular populations of the intestinal microbiota was analysed independently from the diets. The cultures of the Gram-positive anaerobes and lactic acid bacteria appeared to be the least effective in the hydrolysis of phytic acid (Fig. 4). No more than 20% of the phytic acid was degraded to lower myo-inositol phosphates (INT-InsPs and lower derivatives) by those bacteria. In the case of the Proteobacteria–Bacteroides cultures, the amount of INT-InsPs and lower derivatives in the culture medium reached up to 65 and 12%, respectively. Bacteria belonging to E. coli group (MacConkey medium) appeared to be more efficient in the decomposition of phytic acid and generated mainly InsP3 and InsP4 (23 and 65%, respectively), so that together they constituted the majority of all the myo-inositol phosphates. The lower derivatives made the remaining amount of the myo-inositol phosphates. The highest degree of phytate degradation was, however, obtained in the nonselective Wilkins–Chalgren medium, where all phytate was hydrolysed either to the lower inositol phosphates (65%) or to INT-InsPs (35%) (Fig. 4).

Figure 4.

Ability of selected intestinal groups of bacteria to degrade phytate in vitro. GPA, Gram-positive bacteria; LAB, lactic acid bacteria; P-B, Proteobacteria–Bacteroides culture. The graph shows means (medians) for the percentage distribution for the determined inositol phosphates (InsP3, InsP4, InsP5), nondegraded phytic acid (NDPA) and lower derivatives of phytate degradation (LD) in relation to the initial content of phytic acid (1 mmol l−1). (image_n/jam12204-gra-0005.png) NDPA; (image_n/jam12204-gra-0006.png) InsP5; (image_n/jam12204-gra-0007.png) InsP4; (image_n/jam12204-gra-0008.png) InsP3; (image_n/jam12204-gra-0009.png) LD.

Phytate degradation – influence of the diet on the bacterial phytase activity

When comparing the phytate degradation by Gram-positive anaerobes, the only significant difference observed was the higher level of INT-InsPs generated by the infant-originating bacteria than the vegetarians one (P < 0·05; Fig. 5). The analysis of single intermediate derivatives InsP3, InsP4 and InsP5 revealed a nonsignificantly higher concentration of InsP3 and InsP4 (P = 0·059 and P = 0·051, respectively) in the GPA cultures from infants (Supporting Information).

Figure 5.

Degradation of InsP6 [nondegraded phytic acid (NDPA)] to intermediate derivative inositol phosphates (INT-InsPs, the sum of InsP3, InsP4 and InsP5;) and the lower inositol phosphates (InsP1 and InsP2; LD) in modified media: (a) Garche's medium; (b) de Man Rogosa medium (MRS) medium; (c) MacConkey medium; (d) neomycin brilliant green taurocholic acid (NBGT) medium and (e) nonselective Wilkins–Chalgren medium inoculated with faecal bacteria of the adults on conventional diet, the adult vegetarians and the breastfed infants. The brackets show values different at P < 0·05 or P < 0·01. (image_n/jam12204-gra-0010.png) 25–75%; (image_n/jam12204-gra-0011.png) Non-outliers range; (○) Outliers; (■) Extreme.

In the LAB cultures (MRS medium), a higher proportion of INT-InsPs in the conventional diet group when compared to the vegetarian group was observed (P < 0·01; Fig. 5). This resulted from a significantly higher concentration of InsP5 (P < 0·01; see Supporting Information). No significant differences in myo-inositol phosphates concentrations were observed for the E. coli cultures (MacConkey medium). In this case, the bacteria from all the volunteer groups showed a high phytase activity generating mainly INT-InsPs (Fig. 5). In the Proteobacteria–Bacteroides cultures, the concentration of the lower myo-inositol phosphate derivatives generated by the bacteria in the conventional diet group was significantly higher than that in the vegetarians (P < 0·05) and infants (P < 0·05; Fig. 5). The InsP3 content in Proteobacteria–Bacteroides cultures of the conventional-diet adults was lower when compared to the infants' microbiota (P < 0·05; see Supporting Information), which manifested as a significantly lower total content of INT-InsPs (P < 0·05; Fig. 5).

In the cultures of the anaerobes (nonselective Wilkins–Chalgren medium), some significant differences were observed between the vegetarians' and infants' microbiota (Fig. 5). In the vegetarians' bacterial cultures, lower concentrations of INT-InsPs (P < 0·05) and NDPA (P < 0·01) as well as a higher content of lower myo-inositol phosphates (P < 0·01) were determined. The low content of INT-InsPs resulted from the decreased level of InsP4 (P < 0·01) and InsP5 (P < 0·05) (Supporting Information).

The differences in contents of NDPA, INT-InsPs and LD in supernatant of the cultures are presented in Table 2 as relative proportions of those. The ability of the bacterial cultures to degrade phytate (Table 2) showed clearly that GPA and LAB were characterized by low Ftotal (maximal outcome 21) when compared to the other examined bacterial cultures (most of the outcomes over 60) and consequently low FINT-InsPs and FLD. Interestingly, some cultures of similar Ftotal (P-B cultures) differed in FINT-InsPs (two times higher in the BF group compared to C) and FLD (four times higher in C compared to the BF group). On the other hand, in anaerobe cultures of different Ftotal (over two times higher in C than in BF group), a comparable outcome for FINT-InsPs was obtained (Table 2). The analysis of the F factors provided an insight into the activity of the bacterial populations in terms of phytate degradation. Moreover, in the case of a high Ftotal value, the FINT-InsPs and FLD factors enable a more precise identification of the fact whether the cultures consisted mostly of bacteria with a specific phytase activity (FINT-InsPs > FLD) or whether there were also bacteria expressing a nonspecific phosphatase activity (or phytases of a broader substrate specificity). The latter case would lead to an accumulation of lower derivatives of phytate degradation (FLD > FINT-InsPs).

Table 2. Ability of the bacterial cultures to degrade phytate
Bacterial groupaDietbRatio of inositol phosphates in bacterial culturesc (LD: INT-InsPs NDPA)Ability of bacterial cultures to degrade phytated
F LD F INT-InsPs F total
  1. a

    GPA, Gram-positive anaerobes; LAB, lactic acid bacteria; P-B, Proteobacteria–Bacteroides.

  2. b

    Diets: C, conventional; V, vegetarian; BF, breastfeeding.

  3. c

    LD, lower derivatives of phytate degradation (InsP1, InsP2 and myo-inositol); INT-InsPs, intermediate derivatives of phytate degradation (the sum of InsP3, InsP4 and InsP5); NDPA, nondegraded phytic acid.

  4. d

    Ability to hydrolyse phytate to the lower inositol phosphates (FLD) or to generate intermediate inositol phosphates (FINT-InsPs); Ftotal – a general ability to degrade phytate.

GPAC1 : 1 : 7358
V2 : 1 : 78312
BF1 : 4 : 541721
LABC2 : 1 : 7437
V2 : 0 : 8516
BF2 : 1 : 8425
Escherichia coli C3 : 5 : 2274270
V0 : 9 : 148184
BF3 : 7 : 1236083
P-BC4 : 4 : 2333062
V2 : 6 : 2135064
BF1 : 7 : 286068
AnaerobesC6 : 2 : 2471864
V9 : 1 : 088795
BF1 : 4 : 572228


Although in vitro methods are burdened with some drawbacks and do not fully reflect in vivo conditions, they enable to investigate the specific metabolic activity of intestinal bacteria. In this study, microbiological media were used to examine populations of Gram-positive anaerobes, lactic acid bacteria, Enterobacteriaceae, Proteobacteria–Bacteroides and total anaerobes in three human groups that differed in age (adults-infants) or followed a different diet (conventional, vegetarian and breastfeeding). The media composition and the experimental design limited the changes observed during subculturing of complex bacterial populations. The PCR-DGGE approach with sequencing allowed verifying specificities of the applied microbiological media. The selectivity of the MRS and MacConkey media for both lactic acid bacteria and coliforms was confirmed. In the case of Garche's and NBGT media, the identification of predominating DGGE bands resulted in renaming of the cultures to Proteobacteria–Bacteroides and Gram-positive anaerobes (GPA), respectively. In the nonselective medium (Wilkins–Chalgren medium), representatives of the physiological intestinal microbiota were detected (Table 1), including Firmicutes (Lachnospiraceae, Clostridiaceae, Eubacteriaceae, Veillonellaceae), Bacteroidetes (Bacteroides vulgatus), Betaproteobacteria (Sutterellaceae) and Gammaproteobacteria (Enterobacteriaceae).

In this study, the coliforms proved to be the most efficient phytate degraders. Such high activity in phytate degradation might result from the sensitivity of that bacterial group to a deficiency of iron ions which, when bound with phytate, cannot be utilized by the bacteria. E. coli is equipped with large iron uptake systems and its metabolism is strongly regulated by iron depletion (McHugh et al. 2003). This may explain why E. coli degraded phytic acid mainly to InsP4 and InsP3 (phytate derivatives with no ability of binding minerals) and thus making the iron ions (as well as other minerals too) available to bacteria. Moreover, as it has been reported for Pseudomonas (member of the Gammaproteobacteria class), InsP3 mediates in transporting the iron into bacterial cells (Hirst et al. 1999).

Proteobacteria–Bacteroides were following coliforms as the second bacterial culture characterized by a high efficiency in phytate degradation. When compared to E. coli cultures, the differences observed in the profiles of phytate derivatives indicated that changes in the environment (different medium composition and anaerobic conditions) and differences in the bacterial composition may trigger considerable changes in the metabolic activity of bacterial consortia.

The results obtained in this study coincide with results of an experiment of Steer et al. (2004), where a medium enriched with phytate was inoculated with human adult microbiota and phytate-degrading strains were isolated. Among 63 identified strains as many as 27 belonged to Proteobacteria, 11 to Bacteroides and 2 to Fusobacterium, giving in total 64·5% of Gram-negative bacteria among phytate degraders. Such a high proportion of Proteobacteria in the total number of phytate degraders (Steer et al. 2004) correlated with a high effectiveness in phytate degradation in the Proteobacteria–Bacteroides cultures showed in this study.

The cultures of GPA and lactobacilli were characterized by their lowest ability to degrade phytic acid, however, among the examined GPA cultures those of infants' origin showed the highest phytase activity (although still low when compared to the other bacterial groups). In the studies by Steer et al. (2004), Gram-positive aerobic Enterococcus and Staphylococcus, and anaerobic clostridia formed about 35% of the total identified phytate degraders, whereas bifidobacteria and lactobacilli were not recovered from the chemostat effluent. In the present study, Staphylococcus and Enterococcus were not detected in the bacterial cultures, and most of the identified DGGE bands from the GPA profiles represented clostridia, which is partially consistent with the results obtained by Steer et al. (2004). Moreover, in the GPA cultures, bifidobacteria were also detected. According to Steer et al. (2004), bifidobacteria and lactobacilli are not involved in the process of phytate degradation in gastrointestinal tract. However, findings of Haros et al. (2005, 2007) showed that single Bifidobacterium strains manifest phytase activity to different extends, with Bif. pseudocatenulatum ATCC 27919 and Bif. longum subsp. infantis ATCC 15697 isolated from infant faeces being the most active strains. The species composition of Bifidobacterium populations is probably crucial for that activity because the genes coding phytase have been found in the genomes of Bif. longum subsp. infantis and Bif. pseudocatenulatum, whereas they are absent in most of the remaining bifidobacteria species (Tamayo-Ramos et al. 2012). The physiological role of phytase activity in microbiota of breastfed infants is not clear because we did not detect inositol phosphates (InsP3-InsP6) in the mothers' milk (at the micromolar level per gram of dry matter; the data not shown). The substrates for phytases are introduced in the infant's diet after weaning and that can be the time when phytase activity gives an advantage to phytate-degrading bacteria. Another issue that needs further investigation is whether the effects of phytase activity are important only for the bacteria or entail some consequences in the infant's health, such as an increased bioavailability of minerals and amino acids or an impact of phytate degradation products on physiology (development, maturation) of the infant's intestinal epithelium. In the case of lactic acid bacteria, the phytase activity is detected mostly in the strains isolated from sourdough and other naturally fermented plant material (De Angelis et al. 2003), which accounts for an adaptation to the conditions of an extremely high-phytate content.

The results obtained in the study show that Gammaproteobacteria are the main phytate degraders in the human intestine and those are probably them that initiate phytate degradation in the small intestine and continue it in the large intestine. Other groups of bacteria are also involved in this process, and they may both degrade phytates starting from InsP6 (Bif. longum subsp. infantis, Bif. pseudocatenulatum) and hydrolysed the intermediate derivatives of phytate degradation (InsP3, InsP4, InsP5) to the lower inositol phosphates and inorganic phosphorus. The effectiveness of the phytate degradation seems to be higher when the microbiota is already adapted to a higher-phytate concentration, as it is in the case of the vegetarians' microbiota. The results obtained in these examinations allow for the conclusion that more diverse microbiota (adults' vs infants' microbiota), shaped additionally by the diet (omnivores' vs vegetarians'), ensures a higher degree of phytate degradation. The substrate specificity for phytases of the intestinal anaerobes has not been investigated in detail yet. Therefore, the impact of particular bacterial groups or species on the phytate hydrolysis is difficult to evaluate. Whether the degradation of myo-inositol phosphates is a result of specific phytase activity or a result of the activity of nonspecific phosphatases acting on phytate, as those described for Lactobacillus pentosus (Palacios et al. 2005), is still to be elucidated.

Degradation of phytate in the stomach and small intestine is an activity effect of dietary phytases of plant or microbial (fungi) origin. In the intestine, the solubility of phytates plays a crucial role in terms of accessibility for bacterial phytases because phytates that reach the large intestine are mostly in an insoluble form (Schlemmer et al. 2009), which decreases their susceptibility for degradation. Moreover, the higher the concentration of InsP6 present in the small intestinal chyme, the stronger the InsP6 hydrolysis in the large intestine (Schlemmer et al. 2009). The degree of phytate solubilization may be the reason why approximately 50% of phytates remain nondegraded in vivo (Schlemmer et al. 2009), whereas in the in vitro tests, such as the one conducted in this study, the soluble phytate could be completely hydrolysed by the most diverse bacterial cultures (i.e. cultures in the Wilkins–Chalgren nonselective medium).

Studies on intact InsP6 and hydrolysates of InsP6 reported that both of them may strongly affect physiology of human and animal cells by manifesting an antitumor activity (Vucenik and Shamsuddinm 2003; Kumar et al. 2010). Although the mode of action of InsP6 and its hydrolysates is different (Suzuki and Hara 2010; Ishizuka et al. 2011), a suppression of cell proliferation and an induction of cell specialization, especially in conditions of colon carcinogenesis, is undoubtedly a desired effect triggered by these compounds. From this point of view, the complete degradation of phytate does not seem to be a desirable phenomenon. However, the microbiota investigated in this study represented bacteria living in the lumen of the intestine, and this part of microbiota differs from bacterial populations adhering to the mucus layer (Zoetendal et al. 2002). Moreover, bacterial isolates obtained from faeces express different abilities to adhere to epithelial cells and the mucus (Wasilewska et al. 2008). Thus, the ability to degrade phytate to different lower myo-inositol phosphates by bacterial populations adhered to the epithelium deserves further investigation.

Recapitulating, out of all the bacterial groups examined, the GPA and LAB showed the lowest potential for phytate degradation. The vegetarians' GPA and LAB were characterized by the lowest phytase activity when compared to both the omnivorous adults and the breastfed infants. Out of the bacterial populations examined, the highest phytase activity was determined for microbiota in the Wilkins–Chalgren nonselective medium and in the E. coli cultures. The results obtained from the nonselective medium showed that the intestinal bacteria cooperate in the gradual decompositions of phytate, which may reflect both the taxonomical and metabolic diversity (substrate specificity of phytases) of bacteria. Moreover, it has been shown that the microbiota from the environment of high-phytate content (vegetarians' intestine) was the most effective in degrading phytate, which suggests that microbiota adapt to such an environment and that the diet modulates metabolic activities of intestinal bacteria, which was reported in early enrichment studies.


The authors want to thank Danuta Rostek and Mirosław Obrębski for their excellent technical support and to Dr. Vicente Monedero for his critical comments to the manuscript.

The work was financed from the funds of the Polish Ministry of Sciences and Higher Education, grant number N N312 434337.