Extracellular DNA: A Nutritional Trigger of Mycoplasma bovis Cytotoxicity

Abstract

Microbial access to host nutrients is a key factor of the host-pathogen interplay. With their nearly minimal genome, wall-less bacteria of the class Mollicutes have limited metabolic capacities and largely depend on host nutrients for their survival. Despite these limitations, host-restricted mycoplasmas are widely distributed in nature and many species are pathogenic for humans and animals. Yet, only partial information is available regarding the mechanisms evolved by these minimal pathogens to meet their nutrients and the contribution of these mechanisms to virulence. By using the ruminant pathogen Mycoplasma bovis as a model system, extracellular DNA (eDNA) was identified as a limiting nutrient for mycoplasma proliferation under cell culture conditions. Remarkably, the growth-promoting effect induced by supplementation with eDNA was associated with important cytotoxicity for actively dividing host cells, but not confluent monolayers. To identify biological functions mediating M. bovis cytotoxicity, we produced a library of transposon knockout mutants and identified three critical genomic regions whose disruption was associated with a non-cytopathic phenotype. The coding sequences (CDS) disrupted in these regions pointed towards pyruvate metabolism as contributing to M. bovis cytotoxicity. Hydrogen peroxide was found responsible for eDNA-mediated M. bovis cytotoxicity, and non-cytopathic mutants were unable to produce this toxic metabolic compound. In our experimental conditions, no contact between M. bovis and host cells was required for cytotoxicity. Further analyses revealed important intra-species differences in eDNA-mediated cytotoxicity and H₂O₂ production, with some strains displaying a cytopathic phenotype despite no H₂O₂ production. Interestingly, the genome of strains PG45 and HB0801 were characterized by the occurrence of insertion sequences (IS) at close proximity to several CDSs found disrupted in non-cytopathic mutants. Since PG45 and HB0801 produced no or limited amount of H₂O₂, IS-elements might influence H₂O₂ production in M. bovis. These results confirm the multifaceted role of eDNA in microbial communities and further identify this ubiquitous material as a nutritional trigger of M. bovis cytotoxicity. M. bovis may thus take advantage of the multiple sources of eDNA in vivo to modulate its interaction with host cells, a way for this minimal pathogen to overcome its limited coding capacity.

Keywords: Mycoplasmas bovis; cytotoxicity; extracellular DNA; hydrogen peroxide; pyruvate metabolism; virulence factors.

Figure 1

The growth-promoting effect of eDNA on M. bovis. Comparative growth of M. bovis under axenic and cell culture conditions (A). RM16 proliferation was monitored in SP4 medium (SP4), cell culture medium (DMEM), and cell culture medium supplemented with 10 μg/ml calf thymus DNA (DMEMD). Mycoplasma titers (log CFU/ml) were determined by CFU titrations. The cytopathic effect induced by M. bovis upon co-incubation with host cells (B). EBL cells (10⁴ cells) were inoculated with RM16 at an MOI of 2 (RM16) or mock-infected (Mock). After 72 h of co-incubation, monolayers were stained with crystal violet and survival cells were estimated by measuring the optical density at 590 nm (OD 590). When indicated, DMEM-based medium was supplemented with 10 μg/ml calf thymus DNA (eDNA). Calf thymus DNA was subjected to the following enzymatic treatments: RNase A (RNase), Proteinase K (ProtK) and DNase I (DNase) digestion (see section “Materials and Methods”). The asterisk indicates that polynucleotides were removed from DNase I digestion products (DNase*). Infected and mock-infected samples were treated identically. Data are the means of at least three independent assays. Standard deviations are indicated by error bars. p-values were determined by using two-sided independent sample t tests and comparing OD590 values of RM16-infected samples to those of mock-infected samples (^***p < 0.001).

Figure 2

Phenotypic characterization of M. bovis knockout mutants (A) and M. bovis type and field strains (B). M. bovis mutants and strains are described in Tables 1, 2. Mycoplasmas were co-incubated with EBL cells in DMEM-based medium supplemented with 10 μg/ml calf thymus DNA (black bars) or DMEM-based medium without DNA (white bars). EBL cells (10⁴ cells) were inoculated at an MOI of 2. After 72 h of co-incubation, samples were used either for CFU titrations (log CFU/ml), analysis of EBL cell survival (% OD590 nm) or hydrogen peroxide production (H₂O₂). Infected and mock-infected (Mock) samples were treated identically. Data are the means of at least three independent assays. Standard deviations are indicated by error bars. p-values were determined by using two-sided independent sample t tests and comparing (1) mycoplasma titers (log CFU/ml) reached in the presence of DNA to those reached without supplementation, and (2) EBL cell survival (% OD590 nm) or (3) H2O2 production (H2O2) following co-incubation with M. bovis mutants or strains to values obtained with RM16 (ns, p > 0.05; ^**p < 0.01; ^***p < 0.001).

Figure 3

Analysis of M. bovis loci in non-cytopathic mutants. For each locus, CDS found disrupted in non-cytopathic mutants are indicated by black arrows (Table 2) and surrounding CDS by open arrows. The CDS annotations are indicated either by the gene name or the MBOVPG45 locus tag number. Putative functions or features of CDS products are indicated when available: MP, membrane protein; HP, hypothetical proteins; Lipo, lipoproteins. Green and red arrows indicate the position and name of IS transposase in the genome of M. bovis strains HB0801 and PG45, respectively. The occurrence of IS-elements in the draft genome sequence of RM16 was suggested by sequence homology at contig ends. The position of these IS-elements in RM16 is are indicated by dotted boxes.

Figure 4

Phenotypic characterization of M. bovis vaccine strain P150. EBL cell survival (% OD590 nm) after 72 h of co-incubation with HB0801 (P1) and P150 (P150) in DMEM-based medium supplemented with 10 μg/ml calf thymus DNA (A). Mycoplasma titers (log CFU/ml) reached by P1 and P150 after 48 h incubation in axenic DMEM-based medium without pyruvate and supplemented with 10 μg/ml calf thymus DNA and H₂O₂ production as determined by using MQuant^™ Peroxid-test strips (B). Data are the means of at least three independent assays. Standard deviations are indicated by error bars.

FIGUER 5

Main sources of eDNA in the host and its influence on M. bovis cytotoxicity. The main sources of endogenous and exogenous eDNA are indicated (Aucamp et al., 2018). M. bovis is able to induce a pronounced cytopathic effect when a large amount of eDNA is available.