Study of the biosynthesis and functionality of polyphosphate in Bifidobacterium longum KABP042

Abstract

Polyphosphate (poly-P) biosynthesis in bacteria has been linked to many physiological processes and has been characterized as an interesting functional molecule involved in intestinal homeostasis. We determined the capacity for poly-P production of 18 probiotic strains mainly belonging to Bifidobacterium and former Lactobacillus genera, showing that poly-P synthesis varied widely between strains and is dependent on the availability of phosphate and the growth phase. Bifidobacteria were especially capable of poly-P synthesis and poly-P kinase (ppk) genes were identified in their genomes together with a repertoire of genes involved in phosphate transport and metabolism. In Bifidobacterium longum KABP042, the strain we found with highest poly-P production, variations in ppk expression were linked to growth conditions and presence of phosphate in the medium. Moreover, the strain produced poly-P in presence of breast milk and lacto-N-tetraose increased the amount of poly-P synthesized. Compared to KABP042 supernatants low in poly-P, exposure of Caco-2 cells to KABP042 supernatants rich in poly-P resulted in decreased epithelial permeability and increased barrier resistance, induction of epithelial protecting factors such as HSP27 and enhanced expression of tight junction protein genes. These results highlight the role of bifidobacteria-derived poly-P as a strain-dependent functional factor acting on epithelial integrity.

Figure 1

Poly-P accumulation in the studied strains after 6 and 16 h of growth in (a) MEI medium and (b) MRSc medium.

Figure 2

Schematic illustration of the presence and distribution of the phoRP, pstSCAB, pit, ppgk, ppx-gppA and ppk genes in studied Bifidobacterium spp strains. Each solid arrow indicates an open reading frame (ORF). The lengths of the arrows are proportional to the length of the predicted ORF.

Figure 3

Poly-P production in B. longum KABP042. (a) Growth in MEI and LP-MEI. (b) Levels of intracellular and (c) extracellular poly-P as a function of growth. The amount of poly-P (as nmol Pi) from bacteria present in one ml of culture volume and nmol Pi/ml of cell-free culture supernatant are depicted, respectively. Two-way ANOVA with Sidak's multiple comparisons test was employed (see text for p values). (d) Expression of ppk gene (3, 6 and 16 h from panel a) determined by qPCR. Data represent fold changes relative to reference conditions (growth in LP-MEI). The Pair Wise Fixed Reallocation Randomisation Test implemented in REST was employed and asterisks indicate p < 0.0001.

Figure 4

Effect of different components of the growth medium in poly-P synthesis in B. longum KABP042. (a) Growth in MEI medium with different carbon sources (glucose and LNT), human milk added at 1% v/v (no glucose added) and MEI plus a mixture of polyamines (polyA). (b) Poly-P synthesis in the conditions shown in a. Asterisks indicate a significant difference with p < 0.01 (Two-way ANOVA, Dunnett multiple comparisons test) with respect to glucose. (c) ppk expression determined by qPCR. Relative expression is expressed as fold changes in gene expression using growth in MEI (with glucose) as a reference.

Figure 5

Effects of B. longum KABP042 conditioned media on intestinal permeability in Caco-2 model. (a) Transepithelial electrical resistance (TEER) in Caco-2 exposed to conditioned MEI or LP-MEI (Cm; medium fermented by KABP042, filtered and neutralized), respectively, for 72 h. (b) Apparent permeability (P_aap) to Lucifer Yellow in Caco-2 cells monolayers exposed to MEI and LP-MEI conditioned or non-conditioned media for 72 h. One-way ANOVA with Sidak's multiple comparisons test was employed. (c) Relative expression of the genes (fold change) for tight junction proteins (ZO-1, OCLN, JAM-1) determined by qPCR. The reference conditions were those of cells cultivated in the presence of non-conditioned MEI or LP-MEI, respectively. The Pair Wise Fixed Reallocation Randomisation Test implemented in REST was employed.

Figure 6

Induction of HSP27 in Caco-2 cells by incubation with supernatants of B. longum KABP042. (a) Western blot with anti-HSP27; MEI and LP-MEI are non-fermented media; Cm, conditioned medium (i.e. fermented by KABP042, filtered and neutralized); β-actin was used as a loading control. Two representative samples of Caco-2 cells treated with Cm are seen. The gels were transferred and the membranes were cut in two pieces that were probed with anti-HSP27 or anti-β-actin, respectively. Complete gel images are provided in Supplementary Information. (b) Comparison of the relative expression of HSP27 between Caco-2 cells exposed to MEI or LP-MEI conditioned media. In each case the expression is normalized to that of the corresponding non-conditioned medium. Student's t-test was applied. Lowest expression was set to 1, n ≥ 3. (c) Correlation between the amount of poly-P in B. longum KABP042 supernatants and the relative expression of HSP27 in Caco-2 cells.