Drosophila melanogaster Natural Variation Affects Growth Dynamics of Infecting Listeria monocytogenes

Abstract

We find that in a Listeria monocytogenes/Drosophila melanogaster infection model, L. monocytogenes grows according to logistic kinetics, which means we can measure both a maximal growth rate and growth plateau for the microbe. Genetic variation of the host affects both of the pathogen growth parameters, and they can vary independently. Because growth rates and ceilings both correlate with host survival, both properties could drive evolution of the host. We find that growth rates and ceilings are sensitive to the initial infectious dose in a host genotype-dependent manner, implying that experimental results differ as we change the original challenge dose within a single strain of host.

Keywords: Host-variation; bacterial growth dynamics; genetics of immunity.

Figure 1

Median time to death (MTD) and bacterial load following Listeria infection vary in Drosophila naturally derived variants. Median value of (A) survival measured in median time to death (in days) and (B) bacterial load 48 hr postinfection measured in median colony forming units (CFU) when infected with 1000 CFU of Listeria monocytogenes. Arrows and colored columns represent each of the 11 RAL lines used for further study (colors are consistent throughout the figures): RAL 359 (light green), RAL 821 (purple), RAL 375 (dark yellow), RAL 59 (red), RAL 309 (gray), RAL 382 (pink), RAL 136 (dark green), RAL 508 (dark red) RAL 732 (brown), RAL 787 (blue), RAL 73 (orange).

Figure 2

Logistic growth curves of Drosophila naturally derived variants differ primarily in the bacterial growth plateau. Microbial 11 growth curves showing the 95% confidence interval of 11 RAL-lines when infected with 100 L. monocytogenes. (A) All growth curves are on the same axis for comparison. We chose a reference line (B) RAL 309 and graphed the rest of the growth curves (C-L) as compared to RAL 309. The fit (R2) and survival (MTD in hr) for each curve is displayed. Bacterial load was LN transformed to fit logistic curve.

Figure 3

Maximal growth rate does not correlate with growth plateau. Correlating two microbial growth parameters, maximal growth rate and growth plateau, to each other, (A) using all 11 lines and (B) censoring the three lines with the lowest plateau and large 95% confidence intervals for maximal growth rate. Dashed lines represent the 95% confidence interval of the linear regression. n.s. signifies that the slope of the linear regression is not significantly different than zero. Bacterial load was LN transformed to get growth parameter measurements.

Figure 4

Survival negatively correlates with bacterial growth plateau and also with maximal growth rate when points with lowest K are removed: two microbial growth parameters are correlated with survival (MTD) as a measurement of health. The correlation with (A, C) all strains and (B, D) censored strains (without the three RAL lines with the lowest growth plateau and highest confidence intervals for the maximal growth rate). Growth plateau correlated with survival with all (A) and censored (B) data, while maximal growth rate is not correlated with survival with all (C) data, but it is with the censored (D) data. The 95% confidence interval is shown in the dashed-lines of the linear regression. Significance means the slope of the linear regression is significantly different than zero (*P < 0.05, **P < 0.01). Bacterial load was LN transformed to get bacterial parameter measurements.

Figure 5

The effect of initial infectious dose on bacterial growth curve depends on host genetics: Microbial growth curve with different initial doses (10¹, 10², 10³, 10⁴, 10⁵) of three Drosophila lines. (A) Lines were shown to have different growth dynamics with an initial dose of 10². The growth dynamics with different inocula of three lines: (B) w¹¹¹⁸, (C) RAL 309, and (D) RAL 508. All curves were fit to a logistic curve and the 95% confidence interval is shown with the exception of RAL 309 with an initial dose of 10⁵ because the data did not fit a logistic curve; the median bacterial growth of each time point were connected instead. The three Drosophila genotypes have a different dose where the reach their maximal growth plateau (K_M), which are shown with an asterisk (*) in the color of the first initial dose to reach the K_M. Data were LN transformed to fit a logistic growth curve.

Figure 6

Logistic growth curves of Drosophila immune mutants differ primarily in the bacterial growth plateau. Microbial growth curves showing the 95% confidence interval of w¹¹¹⁸ and two immune mutant lines when infected with 100 L. monocytogenes. (A) All growth curves are on the same axis for comparison, the rest are graphed with (B) w¹¹¹⁸, the lab wild type. The two immune mutants are (C) kenny and (D) CG2247. The fit (R²) and survival (MTD in hours) for each curve is displayed. Bacterial load was LN transformed to fit logistic curve.