Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 5 (11), 180810
eCollection

Mutations in Bacterial Genes Induce Unanticipated Changes in the Relationship Between Bacterial Pathogens in Experimental Otitis Media

Affiliations

Mutations in Bacterial Genes Induce Unanticipated Changes in the Relationship Between Bacterial Pathogens in Experimental Otitis Media

Vinal Lakhani et al. R Soc Open Sci.

Abstract

Otitis media (OM) is a common polymicrobial infection of the middle ear in children under the age of 15 years. A widely used experimental strategy to analyse roles of specific phenotypes of bacterial pathogens of OM is to study changes in co-infection kinetics of bacterial populations in animal models when a wild-type bacterial strain is replaced by a specific isogenic mutant strain in the co-inoculating mixtures. As relationships between the OM bacterial pathogens within the host are regulated by many interlinked processes, connecting the changes in the co-infection kinetics to a bacterial phenotype can be challenging. We investigated middle ear co-infections in adult chinchillas (Chinchilla lanigera) by two major OM pathogens: non-typeable Haemophilus influenzae (NTHi) and Moraxella catarrhalis (Mcat), as well as isogenic mutant strains in each bacterial species. We analysed the infection kinetic data using Lotka-Volterra population dynamics, maximum entropy inference and Akaike information criteria-(AIC)-based model selection. We found that changes in relationships between the bacterial pathogens that were not anticipated in the design of the co-infection experiments involving mutant strains are common and were strong regulators of the co-infecting bacterial populations. The framework developed here allows for a systematic analysis of host-host variations of bacterial populations and small sizes of animal cohorts in co-infection experiments to quantify the role of specific mutant strains in changing the infection kinetics. Our combined approach can be used to analyse the functional footprint of mutant strains in regulating co-infection kinetics in models of experimental OM and other polymicrobial diseases.

Keywords: Akkaike information criterion; Condorcet winner; Lotka–Volterra; maximum entropy estimation; otitis media; polymicrobial infection.

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1.
Figure 1.
Schematic representation of our framework to determine roles of mutant bacterial strains in regulating co-infection kinetics. (a) Inter- and intra-species interactions between two bacterial species NTHi and Mcat residing within a host can be both active (solid lines) or passive (dashed lines) in nature. These interactions can be simplified and described by LV interaction parameters ({αij}). α11 (greater than 0) and α22 (greater than 0) represent intra-species interactions for NTHi and Mcat, respectively. α12 and α21 represent the overall effect of Mcat on the growth of NTHi and NTHi on the growth of Mcat, respectively. α12 and α21 can be positive (competitive interaction), zero (neutral interaction) or negative (cooperative interaction). (b) Replacing a wild-type strain by a mutant strain in the co-infection experiments can change the LV interactions. These changes may be anticipated (blue ‘X’) or not anticipated (red ‘X’) based on the design of the experiment. Our framework uses data from co-infection experiments involving the wild-type strains to generate models that determine if these unanticipated changes are weak or strong regulators of bacterial kinetics. These models are compared to each other using AIC.
Figure 2.
Figure 2.
Application of the scheme on synthetic data. (a) Values of N1, N2 pairs (104 pairs) obtained from steady-state solutions of the ODEs corresponding to the LV model where {αij} were drawn from uniform distributions in the following ranges: 2.74× 10−3α11 ≤ 0.2, −200 ≤ α12 ≤ 5, −5 ≤ α21 ≤ 0.1 and 1.9 ≤ α22 ≤ 140. The solutions where either N1 or N2 went to zero values or became very large (N1 > 530 × 106 or N2 > 7 × 106) were not included in the synthetic dataset. (b) Synthetic data (105 data points) for a co-infection with the mixture wt + α(+)22 strain. The α(+)22 strain was generated by increasing the lower range of α22 to 120. (c) Synthetic data (105 data points) for the co-infection for the mixture wt + α(+)12 strain. The α(+)12 strain was generated by increasing the lower range of α12 to −2. (d) The percentage of the time a runner model won against an opponent model in head-to-head comparison of AICs for the models describing the synthetic data in (b). The m = 19 different models are indexed by integers. The percentages shown were obtained for t = 100 trials, each with a sample size of n′ = 1000. A bright row indicates the winning model. (e) Results in head-to-head comparisons between the models presented similar to the data in (c). (f,g) The probability pC for the Condorcet winner to win all the pairwise encounters in the t samples is shown for increasing sample size n′. pC for the Condorcet winner model (#i) is calculated using pC = ∏j(≠i) fij, where fij (greater than 1/2) denotes the fraction of the t samples where the Condorcet winner model #i was preferred over model #j. The product is calculated for all the m − 1 pairwise combinations where m number of models were considered. pC increased with the sample size (n′). The winning models are denoted in the second column by the changes in α11, α12, α21 and α22 for the wild-type+wild-type co-infection. O indicates no change, X indicates an increase and o indicates a decrease.
Figure 3.
Figure 3.
Mean populations of NTHi and Mcat strains in co-infection experiments. Shows the mean populations of NTHi and Mcat strains at day 7 and day 14 post inoculation calculated from counts in the bullae measured in greater than 5 chinchillas for each of the 12 different cases of co-inoculation. The combinations of the bacterial strains used for co-inoculating the chinchillas are listed.
Figure 4.
Figure 4.
Comparison between the Condorcet winner model and measurement at day 7 post inoculation with mutant strains. The probability distribution function p^(N1,N2) generated by the Condorcet winner model for a co-infection at day 7 post inoculation is shown using a heat map. The measured bacterial loads for the same co-infection for individual chinchillas are shown in red points. The anticipated changes in the LV parameters for a co-infection involving a specific mutant strain are shown in the first row of the table shown on the left of a sub-figure. The changes suggested by the Condorcet winning model are shown in the second row. A filled or a smaller empty circle indicates an increase or decrease of a specific parameter, respectively. The cases where the phenotype is uncertain, i.e. either an increase or decrease, are marked by a bull's-eye symbol (formula image). (a) NTHi (wt)-Mcat (hag), (b) NTHi (wt)-Mcat (mcaB), (c) NTHi (wt)-Mcat (aaa), (d) NTHi (luxS)-Mcat (wt), (e) NTHi (wt)-Mcat (mclR) and (f) NTHi (wt)-Mcat (dtgt). Note that we compared our prediction against the measured data in terms of average populations as bacterial measurements were available for only few animals. The individual data points shown on the graphs were not explicitly compared, thus some of the individual measurements (dots) can lie at the boundaries of the predicted distribution (coloured squares) and need not reflect quality of comparison between the average bacterial populations.
Figure 5.
Figure 5.
The prediction at day 14 post inoculation for the wild-type strains generated using the day 7 data. The data are displayed using the same visualization scheme as in figure 4.

Similar articles

See all similar articles

Cited by 1 article

References

    1. Bakaletz LO. 2010. Immunopathogenesis of polymicrobial otitis media. J. Leukoc. Biol. 87, 213–222. (10.1189/jlb.0709518) - DOI - PMC - PubMed
    1. Armbruster CE, Hong W, Pang B, Weimer KE, Juneau RA, Turner J, Swords WE. 2010. Indirect pathogenicity of Haemophilus influenzae and Moraxella catarrhalis in polymicrobial otitis media occurs via interspecies quorum signaling. MBio 1, e00102-10 (10.1128/mBio.00102-10) - DOI - PMC - PubMed
    1. Brogden KA, Guthmiller JM, Taylor CE. 2005. Human polymicrobial infections. Lancet 365, 253–255. (10.1016/S0140-6736(05)70155-0) - DOI - PMC - PubMed
    1. Duan K, Sibley CD, Davidson CJ, Surette MG. 2009. Chemical interactions between organisms in microbial communities. In Bacterial sensing and signaling (eds Collin M, Schuch R.), vol. 16, pp. 1–17. Basel, Switzerland: Karger; (10.1159/000219369) - DOI - PubMed
    1. Lysenko ES, Ratner AJ, Nelson AL, Weiser JN. 2005. The role of innate immune responses in the outcome of interspecies competition for colonization of mucosal surfaces. PLoS Pathog. 1, e1 (10.1371/journal.ppat.0010001) - DOI - PMC - PubMed

LinkOut - more resources

Feedback