Inhibition of Candida albicans morphogenesis by chitinase from Lactobacillus rhamnosus GG

Abstract

Lactobacilli have been evaluated as probiotics against Candida infections in several clinical trials, but with variable results. Predicting and understanding the clinical efficacy of Lactobacillus strains is hampered by an overall lack of insights into their modes of action. In this study, we aimed to unravel molecular mechanisms underlying the inhibitory effects of lactobacilli on hyphal morphogenesis, which is a crucial step in C. albicans virulence. Based on a screening of different Lactobacillus strains, we found that the closely related taxa L. rhamnosus, L. casei and L. paracasei showed stronger activity against Candida hyphae formation compared to other Lactobacillus species tested. By exploring the activity of purified compounds and mutants of the model strain L. rhamnosus GG, the major peptidoglycan hydrolase Msp1, conserved in the three closely related taxa, was identified as a key effector molecule. We could show that this activity of Msp1 was due to its ability to break down chitin, the main polymer in the hyphal cell wall of C. albicans. This identification of a Lactobacillus-specific protein with chitinase activity having anti-hyphal activity will assist in better strain selection and improved application in future clinical trials for Lactobacillus-based Candida-management strategies.

Figure 1

Antihyphal activity and lactic acid production of specific Lactobacillus strains. (a) Hyphal induction of C. albicans (10⁶ cells/ml) during co-incubation with live Lactobacillus cells (10⁸ CFU/ml) and (b) D- and L-lactic acid production of the investigated Lactobacillus strains after growth into stationary phase. The results on hyphal inhibition were normalized to hyphal formation of C. albicans solely.

Figure 2

Inhibition of C. albicans hyphae by L. rhamnosus GG and its components. Hyphal induction of C. albicans (10⁶ cells/ml) during co-incubation with (a) live L. rhamnosus GG cells, cell-free supernatant (CFS), UV-inactivated cells and heat-killed cells (10⁸ cells/ml); (b) the isolated lectin-like proteins Llp1 and Llp2 (50 µg/mL) and purified EPS (200 µg/mL) from L. rhamnosus GG; (c) different concentrations of Msp1 from L. rhamnosus GG; (d) different concentrations of lactic acid (50% L-lactic acid and 50% D-lactic acid) and (e) the combination of lactic acid (mixed, D-lactic acid and L-lactic acid) and Msp1; (f) Biofilm formation during co-incubation with L. rhamnosus GG cell-free supernatant (CFS) (20%), lactic acid (0, 4%) and Msp1 (10 µg/mL), with MRS broth and water as respective controls. The results were normalized to hyphal formation and biofilm formation of C. albicans solely. Single and double asterisks indicate respectively p-values below 0.1 and 0.01, compared to C. albicans solely.

Figure 3

Mutant analysis supports key role for Msp1 in anti-hyphal activity. (a) Hyphal induction of C. albicans (10⁶ cells/ml) during co-incubation with L. rhamnosus GG mutant strains, lacking long galactose-rich EPS, Llp1, Llp2, SpaCBA pili or D-alanylation of the lipoteichoic acids (LTA) on their surface. The hyphal inhibition levels were normalized to inhibition level of L. rhamnosus GG wild-type. An asterisk indicate p-values below 0.001, compared to L. rhamnosus GG wild-type. (b) Visualization of Msp1 on the surface of wild-type (upper panel) and dltD mutant cells (lower panel) by indirect immunofluorescence using light microscopy. (c) Quantification of proteins in culture supernatant of L. rhamnosus GG WT and the dltD mutant using ELISA. (d) Hyphal induction of C. albicans (10⁶ CFU/ml) during co-incubation with cell-free supernatant from L. rhamnosus GG wild-type and from its dltD mutant strain. Single and double asterisks indicate respectively p-values below 0.001 and 0.0001, compared to C. albicans solely.

Figure 4

Enzymatic activity of Msp1. (a) Microscopic images of L. rhamnosus GG (left) and L. plantarum WCFS1 (right) after incubation with C. albicans hyphae. Arrows indicate sites where the poles of lactobacilli seem to interact with the hyphae. Representative images are shown. (b) Chemical deglycosylation of Msp1 does not influence its anti-hyphal activity. (c) Msp1 can break down chitin-azure, a chitin derivative. As a control, the sodium acetate buffer was used. (d) The chitinase inhibitor, Bisdionine C, can prevent hyphal inhibition partially. The results on hyphal inhibition were normalized to hyphal formation of C. albicans solely. Asterisks indicate p-values below 0.05, double asterisks indicate p-values below 0.01, compared to C. albicans solely.

Figure 5

A schematic representation of the proposed mode of action underlying the anti-C. albicans activity of L. rhamnosus GG and the other possible, investigated targets. On the surface of L. rhamnosus GG, several potential interaction partners for components on the cell wall of C. albicans cells can be found. The potential interactions between C. albicans and L. rhamnosus GG surface components that were tested in this manuscript are indicated with black arrows. In the hyphal cell wall, the proportion of chitin is much higher (b) than in unhyphenised cells (a), which makes the polymer available for the hydrolytic activity of Msp1 (c). Subsequent contact with Msp1 causes degrading and destabilizing the hyphal cell wall (d). The size proportions between C. albicans and the lactobacilli are not respected for clarity.