Engineering a genome-reduced bacterium to eliminate Staphylococcus aureus biofilms in vivo

Abstract

Bacteria present a promising delivery system for treating human diseases. Here, we engineered the genome-reduced human lung pathogen Mycoplasma pneumoniae as a live biotherapeutic to treat biofilm-associated bacterial infections. This strain has a unique genetic code, which hinders gene transfer to most other bacterial genera, and it lacks a cell wall, which allows it to express proteins that target peptidoglycans of pathogenic bacteria. We first determined that removal of the pathogenic factors fully attenuated the chassis strain in vivo. We then designed synthetic promoters and identified an endogenous peptide signal sequence that, when fused to heterologous proteins, promotes efficient secretion. Based on this, we equipped the chassis strain with a genetic platform designed to secrete antibiofilm and bactericidal enzymes, resulting in a strain capable of dissolving Staphylococcus aureus biofilms preformed on catheters in vitro, ex vivo, and in vivo. To our knowledge, this is the first engineered genome-reduced bacterium that can fight against clinically relevant biofilm-associated bacterial infections.

Keywords: Mycoplasma; in vivo; bacterial therapy; biofilm; synthetic biology.

Figure EV1. Generation and characterization of M. pneumoniae mutant strains

1. Scheme depicting the protocol followed for the generation of ssDNA recombineering substrates employed to obtain the engineered strains. The illustration shows (from left to right) the amplification of dsDNA PCR products, their incubation with streptavidin‐coated magnetic beads, the NaOH‐mediated release of the strand of interest and the electrophoresis analysis of the products before and after ssDNA purification.

2. Top, electrophoresis analysis of the PCR products obtained at each edited loci for the indicated strains. The internal code of the oligos employed for the screening is shown in brackets, and their sequences can be found in Dataset EV6. Bottom, scheme depicting the expected sizes of the PCR products if the respective locus is edited or not.

3. Plot showing the results of the mass spectrometry analysis done for the mutant strains generated in this work. Bars represent the area under the curve (AUC) values for the three most abundant peptides of each protein in the proteome. Results are shown as the mean ± SD of two technical replicates (n = 2), except for ∆mpn453 strain for which only data for one technical replicate are available. The complete data of the MS analysis can be found in the Dataset EV1. Note that the ∆mpn051 strain was not included in this analysis, as its corresponding edited gene is disrupted by a transposon insertion, and not deleted.

4. Plot showing the estimated doubling times of the mutant strains after 48 h of growth calculated by total protein content increase. Results are shown as the mean ± SD of three biological replicates (n = 3). Complete data of this analysis can be found in Dataset EV2. Results from Fisher’s PLSD test are shown for those strains that showed a doubling time statistically different from that of the parental WTgp35 strain. ***P ≤ 0.0005; ****P ≤ 0.00005.

Figure 1. Assessment of the virulence of M. pneumoniae mutant strains in mouse mammary glands

1. Representative images of abdominal mammary glands found in mice at 4 days post‐inoculation of the indicated strains. Lesion score from negative (–) to maximum (+++) is shown below each picture.

2. Intensity and extension of the hemorrhagic lesions found in the abdominal (Ab) and thoracic (T) mammary glands of mice infected with WT bacteria.

3. Graph showing log₁₀ CFU/gland at 4 days post‐inoculation with the indicated strains (n ≥ 5). Each circle represents the values obtained in individual animals, whereas mean ± SD is represented with lines inside each group. No statistical differences were found between infection groups using a one‐sided ANOVA followed by the post‐hoc Fisher’s PLSD test.

Figure EV2. Histological score of mouse mammary glands infected with M. pneumoniae WT and CV2 strains

Plot showing the results of the blind histopathological analysis carried out at 4 days post‐infection of the abdominal mouse mammary glands. Samples were excised and processed by hematoxylin‐eosin staining and scored (0–3) for the presence of neutrophil infiltration and interstitial inflammation. Each circle represents the scored assigned in individual samples (n ≥ 5), whereas mean ± SD is represented with lines inside each group. No statistical differences were found between infection groups by Fisher’s PLSD test, although a strong statistical tendency (P = 0.07) was found for interstitial inflammation as shown in the graph.

Figure 2. Interleukin expression profile of animals treated with WT or CV2 strains

Plots showing the RT–qPCR analysis conducted to quantify relative expression of the indicated interleukins coding genes in the mammary glands. Results are expressed as the mean ± SD (n = 10 for PBS, n = 5 for WT, and n = 4 for CV2) of the 2^−ΔΔCt relative expression values of the indicated interleukins; the value from each individual animal was calculated from three technical replicates. The values obtained in the PBS group were used as control for normalization of gene expression (= 1). Statistical analysis was performed using a one‐sided ANOVA followed by the post‐hoc Fisher’s PLSD test: (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.005).

Figure EV3. Expression profile of other interleukins in animals treated with CV2 or WT strains

Plots showing the RT–qPCR analysis conducted to quantify relative expression of the indicated interleukins coding genes in the mammary glands. Results are expressed as mean ± SD (n = 10 for PBS, n = 5 for WT, and n = 4 for CV2) of the 2^−ΔΔCt relative expression values of the indicated interleukins; the value from each individual animal was calculated from three technical replicates. The values obtained in the PBS group were used as control for normalization of gene expression (= 1). Statistical analysis was performed using a one‐sided ANOVA followed by the post‐hoc Fisher’s PLSD test. No statistically significant differences were found between the different groups.

Figure EV4. Evaluation of adaptive immune response induced by M. pneumoniae strains in a subcutaneous mice model

1. Experimental design. CD1 female mice were inoculated subcutaneously with a single or repeated bacterial solution of WT or CV2 containing 1 × 10⁸ CFU/mouse at day 0 or day 0+ day 4 (referred to as group 1× and group 2×, respectively). On day 18, animals were sacrificed. The image was created with the help of BioRender.com.

2. Macroscopic evaluation of the subcutaneous tissue of animals subjected to one (1×) or two doses (2×) of PBS, WT or CV2. The ratio of animals showing relevant findings at the inoculation point is shown within each picture (L, left; R, right).

3. Determination of INF‐y, IL‐4, IgM, and IgG protein levels in serum samples measured by ELISA. Each circle represents the values obtained in individual animals (n ≥ 3), subjected to one (1×) or two doses (2×), whereas mean ± SD is represented with lines inside each group. Statistical analysis was performed using one‐way ANOVA and the Tukey post‐hoc test. *P < 0.05.

Figure 3. Validation of secretion signals with alginate lyase A1‐III

1. Plot showing results of the turbidimetric assay (OD 300 nm) conducted to evaluate alginate lyase activity present in the culture supernatants (SN) of the indicated strains carrying the alginate lyase A1‐III coding sequence fused to different secretion signals. Culture supernatants (n = 1) were collected at 0, 24, or 48 h post‐inoculation as indicated.

2. Comparison of the turbidimetric assay results obtained with mpn142 and mpn142Opt secretion signals. Culture supernatants (n = 1) were collected at 0, 5, or 24 h post‐inoculation as indicated. WT strain and medium (Hayflick) were added as controls.

Figure 4. In vitro dispersion assay of S. aureus mature biofilms formed in microplates

1. A, B

   Mature S. aureus biofilms were allowed to develop for 24 h in polystyrene plates and then treated for the indicated time intervals with cell suspensions, or culture supernatants of the CV2 or the CV2‐DispB strains (A) or the WT or the WT‐DispB strain (B). Biofilm presence was assessed by crystal violet staining and included a negative staining control (i.e., crystal violet without biofilm). The results are expressed as mean ± SD of OD 595 nm absorbance values obtained from three different biological replicates (n = 3) and statistically compared by one‐sided ANOVA followed by the post‐hoc Fisher’s PLSD test: *P ≤ 0.05; **P ≤ 0.005; ***P ≤ 0.0005; ****P ≤ 0.00005.

Figure 5. In vitro and ex vivo dispersion assays of S. aureus mature biofilms formed on sealed catheters

1. A

   Schematic representation of the experimental procedure. Catheters pre‐colonized with S. aureus were allowed to form biofilms by incubation at 37°C (for in vitro dispersion assay) or by subcutaneous implantation in mice (for ex vivo dispersion assay) before being treated in vitro with different mycoplasma strains and estimate biofilm dispersion by crystal violet staining.

2. B

   Representative pictures of catheters from the ex vivo dispersion assay with the indicated strains after crystal violet staining are shown. A staining control based on catheters in which no biofilm is formed was also included.

3. C, D

   Plots showing results obtained from the in vitro or ex vivo dispersion assays with the indicated strains and from the staining control. Each circle represents the OD 595 nm values obtained for individual catheters (n ≥ 4), whereas mean ± SD is represented with lines inside each group. Results from one‐sided ANOVA followed by Fisher’s PLSD test are shown for treatments statistically different from those based on strains not secreting dispersin B (i.e., the WT or CV2 strains). *P ≤ 0.05; **P ≤ 0.005; ***P ≤ 0.0005.

Figure 6. In vivo dispersion assay of S. aureus mature biofilms formed on catheters

1. A

   Schematic representation of the experimental procedure. Catheters pre‐colonized with S. aureus were allowed to form biofilms in an in vivo context by subcutaneous implantation in mice. 24 h post‐implantation mice were treated with a single subcutaneous injection of different mycoplasma strains and the effectiveness of each treatment was followed by positron tomography with [¹⁸F]‐FDG‐MicroPET.

2. B

   Representative images of longitudinal slices of [¹⁸F]‐FDG‐MicroPET uptake in mice carrying implanted catheters (red arrows) at day 1 or day 4 of the treatments. Micro‐PET images have been superimposed with CT‐3D images used as anatomical reference. Brain location is highlighted (b).

3. C, D

   Plots showing the SUV60 variation (%) between day 1 and day 4 of the different treatments. Each circle represents the SUV60 variation obtained for individual animals (n ≥ 4), whereas mean ± SD is represented with lines inside each group. Data below the dotted lines indicate that the SUV 60 values decreased at D4 post‐treatment. Results from one‐sided ANOVA followed by Fisher’s PLSD test are shown. *P≤ 0.05; **P ≤ 0.005; ***P ≤ 0.0005.

Figure EV5. Impact in S. aureus growth curves of lysostaphin production by M. pneumoniae driven by different promoter sequences

Plot showing a growth curve of S. aureus. After 6 h of growth, 20 µl of different treatments was added. Dotted lines represent treatments based on culture supernatants (n = 1) of different M. pneumoniae strains carrying the mpn142Opt‐Lysostaphin construct under control of the indicated promoter sequences. A treatment based on the supernatant of a culture of WT strain was added as control. Continuous lines represent treatments based on recombinant lysostaphin protein at the indicated concentrations in µg/µl.