Chlamydial infection and spatial ascension of the female genital tract: a novel hybrid cellular automata and continuum mathematical model

Abstract

Sexually transmitted chlamydial infection initially establishes in the endocervix in females, but if the infection ascends the genital tract, significant disease, including infertility, can result. Many of the mechanisms associated with chlamydial infection kinetics and disease ascension are unknown. We attempt to elucidate some of these processes by developing a novel mathematical model, using a cellular automata-partial differential equation model. We matched our model outputs to experimental data of chlamydial infection of the guinea-pig cervix and carried out sensitivity analyses to determine the relative influence of model parameters. We found that the rate of recruitment and action of innate immune cells to clear extracellular chlamydial particles and the rate of passive movement of chlamydial particles are the dominant factors in determining the early course of infection, magnitude of the peak chlamydial time course and the time of the peak. The rate of passive movement was found to be the most important factor in determining whether infection would ascend to the upper genital tract. This study highlights the importance of early innate immunity in the control of chlamydial infection and the significance of motility-diffusive properties and the adaptive immune response in the magnitude of infection and in its ascension.

Figure 1

The pseudo-three dimensional problem domain. A rectangular lattice with wrapped side boundaries is used to provide a cylindrical model.

Figure 2

(a) The number of chlamydial IFUs over the time-course of infection: black curves denote model-generated time-courses from our baseline parameter sets and the red data denotes the mean IFUs (with standard error) from 14 guinea pigs initially infected intravaginally with 10³ IFUs. (b) The model-predicted time-courses from simulations of the percentage of epithelial cells infected. The red curves represent the mean of the simulations and the blue curves represent the median simulation.

Figure 3

Pie charts of the relative contribution of each parameter to the variability in the 100 model simulations of the (a) peak chlamydial load, (b) maximum percentage of infected cells. The relative contribution was calculated by factor prioritization by reduction of variance sensitivity indices according to a regression model with interaction terms.

Figure 3

Pie charts of the relative contribution of each parameter to the variability in the 100 model simulations of the (a) peak chlamydial load, (b) maximum percentage of infected cells. The relative contribution was calculated by factor prioritization by reduction of variance sensitivity indices according to a regression model with interaction terms.

Figure 4

Spatial distributions from scenario 2 (10-fold increase in the diffusion coefficient compared with baseline) for the simulation with the median peak value for chlamydia particles. The upper panel shows the spatial distribution of chlamydial particles over the length and circumference of the endocervix for days 0, 3, 6, 9, and 12 post infection. Darker areas correspond with higher particle counts. The lower panel shows the density of chlamydial particles versus distance of endocervix ascension for days 0, 3, 6, 9, and 12 post infection.