Journal of Functional Foods 35 (2017) 279–294

A pea (Pisum sativum L.) seed albumin extract prevents colonic DSS induced dysbiosis in mice

Isabel Aranda-Olmedo, Raquel Ruiz, M Jesús Peinado and Luis A Rubio

Physiology and Biochemistry of Animal Nutrition (EEZ, CSIC), Granada, Spain

Correspondence should be addressed to Luis A Rubio, Estación Experimental del Zaidín, 18100 Granada, Spain, Email:



This study investigates the effects of a pea (Pisum sativum) seed albumin extract (PSE) on the colonic microbiota in a model of experimental dextran sodium sulfate (DSS)-induced colitis in mice. Male C57BL/6J mice were assigned to three groups: one non-colitic and two colitic. Colitis was induced by incorporating DSS (3.5%) in the drinking water for four days, after which DSS was removed. The pre-treated group received orally PSE (15 g/kg•day) starting two weeks before colitis induction, and was maintained for nine days after. Mice pre-treated with PSE showed a recovery in colon length compared with non-treated DSS group. Both RT-qPCR and pyrosequencing analysis showed that DSS induced significant modifications in the microbiota composition of colonic contents and tissue. In conclusion, PSE modulated colonic microbiota in a model of experimental dextran sodium sulfate (DSS)-induced colitis in mice, and prevented colonic DSS induced dysbiosis in mice.

Full text



Inflammatory bowel disease (IBD) is a syndrome that comprises two major diseases, Crohn’s disease and ulcerative colitis, which are chronic relapsing inflammatory diseases characterized by chronic diarrhea, abdominal pain and rectal bleeding that impair the patients’ quality of life. Despite their efficacy, the main drugs used nowadays in the pharmacological therapy of intestinal inflammation (mainly aminosalycilates, immunosuppressants and biologicals) are limited in their long term use due to the frequent onset of adverse effects. A rational nutritional therapeutic strategy to treat IBD patients usually includes dietary modifications and nutritional supplementations to provide calories, reduce food antigenic stimulation, regulate inflammatory and immune response, and stimulate the mucosal regeneration (Griffiths, 2004), thus promoting the reduction of symptoms, and the induction and maintenance of clinical remission.

In this regard, the administration of selected vegetable extracts could have a positive effect on all these functions and be beneficial for the management of human IBD, or probably even more importantly, for its prevention. This could be the case of some legume seed protein extracts, as shown in a previous work by our group (Utrilla et al., 2015), where we reported that some pea seed fractions ameliorated the colonic mRNA expression of different pro-inflammatory markers as well as proteins involved in maintaining the epithelial barrier function. Nonetheless, little attention was paid to the role that the intestinal microbiota may play in this context, and the microbial analysis was limited to the caecal contents.

The implication of the intestinal microbiota composition in IBD pathogenesis is well documented. Indeed, the shifts usually found in both UC patients and DSS-treated mice (increased numbers of Enterobacteriaceae, including E. coli, and decreased Firmicutes, with selectively decreased Clostridium spp and Lactobacillus spp) (Sartor, 2008) were confirmed in the present work for the groups of the Firmicutes phylum here measured. Thus, in DSS mice lower counts of lactobacilli, B. coccoides/E. rectale, C. leptum (all of them Firmicutes) and bacteroides, together with higher counts of enterobacteria and Escherichia/Shigella were found by qPCR analysis at both contents and tissue colonic levels compared with control values (Figure 1). In addition, a recovery to control values in PSE treated animals was observed for lactobacilli, bifidobacteria, B. coccoides/E. rectale, C. leptum, enterobacteria, Escherichia/Shigella and bacteroides in the colon contents, while only Lactobacillus spp and Bacteroides spp showed a recovery at the tissue level. Pyrosequencing analysis of the microbiota composition at the tissue level in the colon confirmed the results obtained with qPCR for both Lactobacillaceae family and Lactobacillus spp proportions, which dropped in DSS mice but were not different from controls in PSE treated animals.

Works in vivo with defined dietary proteins are scarce, and in the case of legumes almost inexistent. In the current investigation, a recovery to control values was observed in PSE pre-treated mice in most bacterial groups. More specifically, lactobacilli and bifidobacteria counts at the contents level (Figure 1A), and lactobacilli counts and proportions at the tissue level (Figure 1B), were not different from healthy controls. We have previously found (Utrilla et al., 2015) that PSE ameliorated the colonic mRNA expression of different pro-inflammatory markers (cytokines, inducible enzymes, metalloproteinases, adhesion molecules, and toll-like receptors), as well as proteins involved in maintaining the epithelial barrier function (occludin, zonulae occludens). Here we show that pre-treatment with PSE restored microbial colonic counts, particularly those of some bacteria with probiotic effect (namely lactobacilli, bifidobacteria and the B. coccoides/E. rectale group at the contents level, and lactobacilli at the tissue level), and limited the growth of colitogenic microbes (enterobacteria and Escherichia/Shigella) (Figure 1). Therefore, the tissular changes at the histological and biochemical levels are most likely linked to shifts in the microbiota composition as described here.

It is interesting to note that in the current work while the DSS treatment induced significant increases in Bacteroidaceae and decreases in Prevotellaceae families, the opposite was found for the PSE pre-treated mice, whose values tended to return or returned to those of the healthy control group. In addition, Principal Components Analysis revealed that Prevotellaceae were associated with Porphyromonadaceae and Rikenellaceae, while Bacteroidaceae were associated with Enterobacteriaceae, Lachnospiraceae and Ruminococcaceae (Figure 2). Also, positive correlations were found among Prevotellaceae, Porphyromonadaceae and Rikenellaceae, and between Bacteroidaceae and Enterobacteriaceae.

From the literature it seems that Bacteroidaceae within the Bacteroidales is linked to high protein and animal fat diets, while the other Bacteroidales groups are linked to plant polysaccharides-rich diets. It has also been shown that oral intake of a multi fibre mix designed to match the fibre content of a healthy diet counteracts IBD-like intestinal inflammation and weight loss in DSS treated mice (Hartog et al., 2015). Also, the link between high proportions of Enterobacteriaceae and colitis via lipolysaccharide interaction with T cells interaction is well documented (Gronbach et al., 2014). Therefore, it is likely that the soluble NSP-rich PSE here used modulated the intestinal microbiota to the Prevotella enterotype in the DSS treated mice (closer to the Bacteroides/Enterobacteriaceae enterotype) which would have resulted in lower colon inflammation markers as previously reported (Utrilla et al., 2015). Quite interestingly, Monk et al. (2016) found both an increase in Prevotellaceae and enhanced multiple concurrent gut health promoting parameters that translated into reduced colitis severity in mice fed diets supplemented with Phaseolus vulgaris beans, and these effects were not linked to the polyphenols seed content.

With the information we have at present, it is not possible to establish the mode of action of PSE. However, the effect was more pronounced in the microbiota composition of the colon contents, and PSE consists of a soluble albumin extract containing 411.4 mg protein/g freeze-dried material, the remaining consisting of soluble NSP (Rubio et al., 2014). As NSP fractions are known to be metabolized almost exclusively by the intestinal microbiota giving place to SCFA (Flint et al., 2012), and fibre may induce other putative beneficial changes in the microbiota composition, we suggest a combined effect of the protein fraction on the inflammatory process itself with an effect on the microbiota probably due mainly to the NSP fraction.

In conclusion, pre-treatment with PSE prevented the DSS-induced dysbiosis in mice. This results, together with those previously reported on the colonic mRNA expression of different pro-inflammatory markers and of proteins involved in maintaining the epithelial barrier, strongly support the contention that pea seed meal or its protein extracts might represent valuable natural ingredients to prevent IBD incidence, and deserve closer research attention as tools to improve gut health and reduce the severity of relapse in mucosal damage-associated pathologies.





Figure 1. RT-qPCR analysis of the microbiota composition (log10 no. of copies of 16S RNA/mg sample) of the colonic contents (A) or tissue (B) of DSS induced colitic mice. Control: non-colitic control; DSS: 3.5% (w/v) DSS in the drinking water for four days; PSE: orally treated by gavage with pea seed extract (15 g/kg·day) over the whole experimental period, starting two weeks before colitis induction. Bars are means ± SEM of ten mice per group. Values are log10 no. of copies per milligram of freeze-dried sample. For each bacterial group, bars with different superscripts differ (p < 0.01).



Figure 2. Principal Components Analysis at Family level (after SIMPER analysis and VARIMAX rotation) of pyrosequencing analysis. Control: non-colitic control; DSS: 3.5% (w/v) DSS in the drinking water for four days; PSE: orally treated by gavage with pea seed extract (15 g/kg·day) over the whole experimental period, starting two weeks before colitis induction.



Flint, H. J., Scott, K. P., Duncan, S. H., Louis, P., & Forano, E. (2012). Microbial degradation of complex carbohydrates in the gut. Gut Microbes, 3:289-306. DOI: 10.4161/gmic.19897.

Griffiths, A. M. (2004). Enteral nutrition in the management of Crohn’s disease. Journal of Parenteral and Enteral Nutrition, 29 (4 Suppl.):S108-S112.

Gronbach, K., Flade, I., Holst, O., Lindner, B., Ruscheweyh, H. J.,… Frick, J. S. (2014). Endotoxicity of Lipopolysaccharide as a Determinant of T-Cell Mediated Colitis Induction in Mice. Gastroenterology, 146:765-75. DOI: 10.1053/j.gastro.2013.11.033.

Hartog, A., Belle, F. N., Bastiaans, J., de Graaff, P., Garssen, J., Harthoorn, L. F., Vos, A. P. (2015). A potential role for regulatory T-cells in the amelioration of DSS induced colitis by dietary non-digestible polysaccharides. The Journal of Nutritional Biochemistry, 26:227-33. DOI: 10.1016/j.jnutbio.2014.10.011.

Monk, J. M., Lepp, D., Zhang, C. P., Wu, W., … Power, K. A. (2016). Diets enriched with cranberry beans alter the microbiota and mitigate colitis severity and associated inflammation. The Journal of Nutritional Biochemistry, 28:129-39. DOI: 10.1016/j.jnutbio.2015.10.014.

Rubio, L. A., Pérez, A., Ruiz, R., Guzmán, M. A., Aranda-Olmedo, I., & Clemente, A. (2014). Characterization of pea (Pisum sativum) seed protein fractions. Journal of the Science of Food and Agriculture, 94:280-87. DOI: 10.1002/jsfa.6250.

Sartor, R. B. (2008). Microbial Influences in Inflammatory Bowel Diseases. Gastroenterology, 134:577–94. DOI: 10.1053/j.gastro.2007.11.059.

Utrilla, M. P., Peinado, M. J., Ruiz, R., Rodríguez-Nogales, A., Algieri, F., Rodríguez-Cabezas, M. E.,…, Rubio, L. A. (2015). Pea (Pisum sativum L.) seed albumin extracts show anti-inflammatory effect in a DSS model of mouse colitis. Molecular Nutrition & Food Research, 59:807-19. DOI: 10.1002/mnfr.201400630.